(lively music) >> Hello, everyone. Welcome back to the Supercloud presentation here. I'm theCUBE, I'm John Furrier, your host. What a great segment here. We're going to unpack the networking aspect of the cloud, how that translates into what Supercloud architecture and platform deployment scenarios look like. And demystify multi-cloud, hybridcloud. We've got two great experts. Amir Khan, the Co-Founder and CEO of Alkira, Atif Khan, Co-Founder and CTO of Alkira. These guys been around since 2018 with the startup, but before that story, history in the tech industry. I mean, routing early days, multiple waves, multiple cycles. >> Welcome three decades. >> Welcome to Supercloud. >> Thanks. >> Thanks for coming on. >> Thank you so much for having us. >> So, let's get your take on Supercloud because it's been one of those conversations that really galvanized the industry because it kind of highlights almost this next wave, this next side of the street that everyone's going to be on that's going to be successful. The laggards on the legacy seem to be stuck on the old model. SaaS is growing up, it's ISVs, it's ecosystems, hyperscale, full hybrid. And then multi-cloud around the corners cause all this confusion, everyone's hand waving. You know, this is a solution, that solution, where are we? What do you guys see as this supercloud dynamic? >> So where we start from is always focusing on the customer problem. And in 2018 when we identified the problem, we saw that there were multiple clouds with many diverse ways of doing things from the network perspective, and customers were struggling with that. So we delved deeper into that and looked at each one of the cloud architectures completely independent. And there was no common solution and customers were struggling with that from the perspective. They wanted to be in multiple clouds, either through mergers and acquisitions or running an application which may be more cost effective to run in something or maybe optimized for certain reasons to run in a different cloud. But from the networking perspective, everything needed to come together. So that's, we are starting to define it as a supercloud now, but basically, it's a common infrastructure across all clouds. And then integration of high lift services like, you know, security or IPAM services or many other types of services like inter-partner routing and stuff like that. So, Amir, you agree then that multi-cloud is simply a default result of having whatever outcomes, either M&A, some productivity software, maybe Azure. >> Yes. >> Amazon has this and then I've got on-premise application, so it's kinds mishmash. >> So, I would qualify it with hybrid multi-cloud because everything is going to be interconnected. >> John: Got it. >> Whether it's on-premise, remote users or clouds. >> But have CTO perspective, obviously, you got developers, multiple stacks, got AWS, Azure and GCP, other. Not everyone wants to kind of like go all in, but yet they don't want to hedge too much because it's a resource issue. And I got to learn this stack, I got to learn that stack. So then now, you have this default multi-cloud, hybrid multi-cloud, then it's like, okay, what do I do? How do you spread that around? Is it dangerous? What's the the approach technically? What's some of the challenges there? >> Yeah, certainly. John, first, thanks for having us here. So, before I get to that, I'll just add a little bit to what Amir was saying, like how we started, what we were seeing and how it, you know, correlates with the supercloud. So, as you know, before this company, Alkira, we were doing, we did the SD-WAN company, which was Viptela. So there, we started seeing when people started deploying SD-WAN at like a larger scale. We started like, you know, customers coming to us and saying they needed connectivity into the cloud from the SD-WAN. They wanted to extend the SD-WAN fabric to the cloud. So we came up with an architecture, which was like later we started calling them Cloud onRamps, where we built, you know, a transit VPC and put like the virtual instances of SD-WAN appliances extended from there to the cloud. But before we knew, like it started becoming very complicated for the customers because it wasn't just connectivity, it also required, you know, other use cases. You had to instantiate or bring in security appliances in there. You had to secure all of that stuff. There were requirements for, you know, different regions. So you had to bring up the same thing in different regions. Then multiple clouds, what did you do? You had to replicate the same thing in multiple clouds. And now if there was was requirement between clouds, how were you going to do it? You had to route traffic from somewhere, and come up with all those routing controls and stuff. So, it was very complicated. >> Like spaghetti code, but on network. >> The games begin, in fact, one of our customers called it spaghetti mess. And so, that's where like we thought about where was the industry going and which direction the industry was going into? And we came up with the Alkira where what we are doing is building a common infrastructure across multiple clouds, across in, you know, on-prem locations, be it data centers or physical sites, branches sites, et cetera, with integrated security and network networking services inside. And, you know, nowadays, networking is not only about connectivity, you have to secure everything. So, security has to be built in. Redundancy, high availability, disaster recovery. So all of that needs to be built in. So that's like, you know, kind of a definition of like what we thought at that time, what is turning into supercloud now. >> Yeah. It's interesting too, you mentioned, you know, VPCs is not, configuration of loans a hassle. Nevermind the manual mistakes could be made, but as you decide to do something you got to, "Oh, we got to get these other things." A lot of the hyper scales and a lot of the alpha cloud players now, and cloud native folks, they're kind of in that mode of, "Wow, look at what we've built." Now, they're got to maintain, how do I refresh it? Like, how do I keep the talent? So they got this similar chaotic environment where it's like, okay, now they're already already through, so I think they're going to be okay. But then some people want to bypass it completely. So there's a lot of customers that we see out there that fit the makeup of, I'm cloud first, I've lifted and shifted, I move some stuff to the cloud. But I want to bypass all that learnings from all the people that are gone through the past three years. Can I just skip that and go to a multi-cloud or coherent infrastructure? What do you think about that? What's your view? >> So yeah, so if you look at these enterprises, you know, many of them just to find like the talent, which for one cloud as far as the IT staff is concerned, it's hard enough. And now, when you have multiple clouds, it's hard to find people the talent which is, you know, which has expertise across different clouds. So that's where we come into the picture. So our vision was always to simplify all of this stuff. And simplification, it cannot be just simplification because you cannot just automate the workflows of the cloud providers underneath. So you have to, you know, provide your full data plane on top of it, fed full control plane, management plane, policy and management on top of it. And coming back to like your question, so these nowadays, those people who are working on networking, you know, before it used to be like CLI. You used to learn about Cisco CLI or Juniper CLI, and you used to work on it. Nowadays, it's very different. So automation, programmability, all of that stuff is the key. So now, you know, Ops guys, the DevOps guys, so these are the people who are in high demand. >> So what do you think about the folks out there that are saying, okay, you got a lot of fragmentation. I got the stacks, I got a lot of stove pipes, if you will, out there on the stack. I got to learn this from Azure. Can you guys have with your product abstract the way that's so developers don't need to know the ins and outs of stack's, almost like a gateway, if you will, the old days. But like I'm a developer or team develop, why should I have to learn the management layer of Azure? >> That's exactly what we started, you know, out with to solve. So it's, what we have built is a platform and the platform sits inside the cloud. And customers are able to build their own network or a virtual network on top using that platform. So the platform has its own data plane, own control plane and management plane with a policy layer on top of it. So now, it's the platform which is sitting in different clouds, but from a customer's point of view, it's one way of doing networking. One way of instantiating or bringing in services or security services in the middle. Whether those are our security services or whether those are like services from our partners, like Palo Alto or Checkpoint or Cisco. >> So you guys brought the SD-WAN mojo and refactored it for the cloud it sounds like. >> No. >> No? (chuckles) >> We cannot said. >> All right, explain. >> It's way more than that. >> I mean, SD-WAN was wan. I mean, you're talking about wide area networks, talking about connected, so explain the difference. >> SD-WAN was primarily done for one major reason. MPLS was expensive, very strong SLAs, but very low speed. Internet, on the other hand, you sat at home and you could access your applications much faster. No SLA, very low cost, right? So we wanted to marry the two together so you could have a purely private infrastructure and a public infrastructure and secure both of them by creating a common secure fabric across all those environments. And then seamlessly tying it into your internal branch and data center and cloud network. So, it merely brought you to the edge of the cloud. It didn't do anything inside the cloud. Now, the major problem resides inside the clouds where you have to optimize the clouds themselves. Take a step back. How were the clouds built? Basically, the cloud providers went to the Ciscos and Junipers and the rest of the world, built the network in the data centers or across wide area infrastructure, and brought it all together and tried to create a virtualized layer on top of that. But there were many limitations of this underlying infrastructure that they had built. So number of routes per region, how inter region connectivity worked, or how many routes you could carry to the VPCs of V nets? That all those were becoming no common policy across, you know, these environments, no segmentation across these environments, right? So the networking constructs that the enterprise customers were used to as enterprise class carry class capabilities, they did not exist in the cloud. So what did the customer do? They ended up stitching it together all manually. And that's why Atif was alluding to earlier that it became a spaghetti mess for the customers. And then what happens is, as a result, day two operations, you know, troubleshooting, everything becomes a nightmare. So what do you do? You have to build an infrastructure inside the cloud. Cloud has enough raw capabilities to build the solutions inside there. Netflix's of the world. And many different companies have been born in the cloud and evolved from there. So why could we not take the raw capabilities of the clouds and build a network cloud or a supercloud on top of these clouds to optimize the whole infrastructure and seamlessly connecting it into the on-premise and remote user locations, right? So that's your, you know, hybrid multi-cloud solution. >> Well, great call out on the SD-WAN in common versus cloud. 'Cause I think this is important because you're building a network layer in the cloud that spans out so the customers don't have to get into the, there's a gap in the system that I'm used to, my operating environment, of having lockdown security and network. >> So yeah. So what you do is you use the raw capabilities like bandwidth or virtual machines, or you know, containers, or, you know, different types of serverless capabilities. And you bring it all together in a way to solve the networking problems, thereby creating a supercloud, which is an abstraction layer which hides all the complexity of the underlying clouds from the customer, right? And it provides a common infrastructure across all environments to that customer, right? That's the beauty of it. And it does it in a way that it looks like, if they have the networking knowledge, they can apply it to this new environment and carry it forward. One way of doing security across all clouds and hybrid environments. One way of doing routing. One way of doing large-scale network address translation. One way of doing IPAM services. So people are tired of doing individual things and individual clouds and on-premise locations, right? So now they're getting something common. >> You guys brought that, you brought all that to bear and flexible for the customer to essentially self-serve their network cloud. >> Yes, yeah. Is that the wave? >> And nowadays, from business perspective, agility is the key, right? You have to move at the pace of the business. If you don't, you are losing. >> So, would it be safe to say that you guys have a network supercloud? >> Absolutely, yeah. >> We, pretty much, yeah. Absolutely. >> What does that mean to our customer? What's in it for them? What's the benefit to the customer? I got a network supercloud, it connects, provides SLA, all the capabilities I need. What do they get? What's the end point for them? What's the end? >> Atif, maybe you can talk some examples. >> The IT infrastructure is all like distributed now, right? So you have applications running in data centers. You have applications running in one cloud. Other cloud, public clouds, enterprises are depending on so many SaaS applications. So now, these are, you can call these endpoints. So a supercloud or a network cloud, from our perspective, it's a cloud in the middle or a network in the middle, which provides connectivity from any endpoint to any endpoint. So, you are able to connect to the supercloud or network cloud in one way no matter where you are. So now, whichever cloud you are in, whichever cloud you need to connect to. And also, it's not just connecting to the cloud. So you need to do a lot of stuff, a lot of networking inside the cloud also. So now, as Amir was saying, every cloud has its own from a networking, you know, the concept perspective or the construct, they are different. There are limitations in there also. So this supercloud, which is sitting on top, basically, your platform is sitting into the cloud, but the supercloud is built on top of using your platform. So that abstracts all those complexities, all those limitations. So now your limitations are whatever the limitations of that platform are. So now your platform, that platform is in our control. So we can keep building it, we can keep scaling it horizontally. Because one of the things is that, you know, in this cloud era, one of the things is autoscaling these services. So why can't the network now autoscale also, just like your other services. >> Network autoscaling is a genius idea, and I think that's a killer. I want to ask the the follow on question because I think, first of all, I love what you guys are doing. So, I think it's a great example of this new innovation. It's not obvious until you see it, right? Geographical is huge. So, you know, single instance, global instances, multiple instances, you're seeing global. How do you guys look at that global equation? Because as companies expand their clouds into geos, and then ultimately, you know, it's obviously continent, region and locales. You're going to have geographic issues. So, this is an extension of your network cloud? >> Amir: It is the extension of the network cloud because if you look at this hyperscalers, they're sitting pretty much everywhere in the globe. So, wherever their regions are, the beauty of building a supercloud is that you can by definition, be available in those regions. It literally takes a day or two of testing for our stack to run in those regions, to make sure there are no nuances that we run into, you know, for that region. The moment we bring it up in that region, all customers can onboard into that solution. So literally, what used to take months or years to build a global infrastructure, now, you can configure it in 10 minutes basically, and bring it up in less than one hour. Since when did we see any solution- >> And by the way, >> that can come up with. >> when the edge comes out too, you're going to start to see more clouds get bolted on. >> Exactly. And you can expand to the edge of the network. That's why we call cloud the new edge, right? >> John: Yeah, it is. Now, I think you guys got a good solutions, network clouds, superclouds, good. So the question on the premise side, so I get the cloud play. It's very cool. You can expand out. It's a nice layer. I'm sure you manage the SLAs between latency and all kinds of things. Knowing when not to do things. Physics or physics. Okay. Now, you've got the on-premise. What's the on-premise equation look like? >> So on-premise, the kind of customers, we are working with large enterprises, mid-size enterprises. So they have on-prem networks, they have deployed, in many cases, they have deployed SD-WAN. In many cases, they have MPLS. They have data centers also. And a lot of these companies are, you know, moving the applications from the data center into the cloud. But we still have large enterprise- >> But for you guys, you can sit there too with non server or is it a box or what is it? >> It's a software stack, right? So, we are a software company. >> Okay, so no box. >> No box. >> Okay, got it. >> No box. >> It's even better. So, we can connect any, as I mentioned, any endpoint, whether it's data centers. So, what happens is usually these enterprises from the data centers- >> John: It's a cloud endpoint for you. >> Cloud endpoint for us. And they need highspeed connectivity into the cloud. And our network cloud is sitting inside the or supercloud is sitting inside the cloud. So we need highspeed connectivity from the data centers. This is like multi-gig type of connectivity. So we enable that connectivity as a service. And as Amir was saying, you are able to bring it up in minutes, pretty much. >> John: Well, you guys have a great handle on supercloud. I really appreciate you guys coming on. I have to ask you guys, since you have so much experience in the industry, multiple inflection points you've guys lived through and we're all old, and we can remember those glory days. What's the big deal going on right now? Because you can connect the dots and you can imagine, okay, like a Lambda function spinning up some connectivity. I need instant access to a new route, throw some, I need to send compute to an edge point for process data. A lot of these kind of ad hoc services are going to start flying around, which used to be manually configured as you guys remember. >> Amir: And that's been the problem, right? The shadow IT, that was the biggest problem in the enterprise environment. So that's what we are trying to get the customers away from. Cloud teams came in, individuals or small groups of people spun up instances in the cloud. It was completely disconnected from the on-premise environment or the existing IT environment that the customer had. So, how do you bring it together? And that's what we are trying to solve for, right? At a large scale, in a carrier cloud center (indistinct). >> What do you call that? Shift right or shift left? Shift left is in the cloud native world security. >> Amir: Yes. >> Networking and security, the two hottest areas. What are you shifting? Up or down? I mean, the network's moving up the stack. I mean, you're seeing the run times at Kubernetes later' >> Amir: Right, right. It's true we're end-to-end virtualization. So you have plumbing, which is the physical infrastructure. Then on top of that, now for the first time, you have true end-to-end virtualization, which the cloud-like constructs are providing to us. We tried to virtualize the routers, we try to virtualize instances at the server level. Now, we are bringing it all together in a truly end-to-end virtualized manner to connect any endpoint anywhere across the globe. Whether it's on-premise, home, multiple clouds, or SaaS type environments. >> Yeah. If you talk about the technical benefits beyond virtualizations, you kind of see in virtualization be abstracted away. So you got end-to-end virtualization, but you don't need to know virtualization to take advantage of it. >> Exactly. Exactly. >> What are some of the tech involved where, what's the trend around on top of virtual? What's the easy button for that? >> So there are many, many use cases from the customers and they're, you know, some of those use cases, they used to deliver out of their data centers before. So now, because you, know, it takes a long time to spend something up in the data center and stuff. So the trend is and what enterprises are looking for is agility. And to achieve that agility, they are moving those services or those use cases into the cloud. So another technical benefit of like something like a supercloud and what we are doing is we allow customers to, you know, move their services from existing data centers into the cloud as well. And I'll give you some examples. You know, these enterprises have, you know, tons of partners. They provide connectivity to their partners, to select resources. It used to happen inside the data center. You would bring in connectivity into the data center and apply like tons of ACLs and whatnot to make sure that you are able to only connect. And now those use cases are, they need to be enabled inside the cloud. And the customer's customers are also, it's not just coming from the on-prem, they're coming from the cloud as well. So, if they're coming from the cloud as well as from on-prem, so you need like an infrastructure like supercloud, which is sitting inside the cloud and is able to handle all these use cases. So all of these use cases have to be, so that requires like moving those services from the data center into the cloud or into the supercloud. So, they're, oh, as we started building this service over the last four years, we have come across so many use cases. And to deliver those use cases, you have to have a platform. So you have to have your own platform because otherwise you are depending on somebody else's, you know, capabilities. And every time their capabilities change, you have to change. >> John: I'm glad you brought up the platform 'cause I want to get your both reaction to this. So Bob Muglia just said on theCUBE here at Supercloud, that supercloud is a platform that provides programmatically consistent services hosted on heterogeneous cloud providers. So the question is, is supercloud a platform or an architecture in your view? >> That's an interesting view on things, you know? I mean, if you think of it, you have to design or architect a solution before we turn it into a platform. >> John: It's a trick question actually. >> So it's a, you know, so we look at it as that you have to have an architectural approach end to end, right? And then you build a solution based on that approach. So, I don't think that they are mutually exclusive. I think they go hand in hand. It's an architecture that you turn into a solution and provide that agility and high availability and disaster recovery capability that it built into that. >> It's interesting that these definitions might be actually redefined with this new configuration. >> Amir: Yes. >> Because architecture and platform used to mean something, like, aight here's a platform, you buy this platform. >> And then you architecture solution. >> Architect it via vendor. >> Right, right, right. >> Okay. And they have to deal with that architecture in the place of multiple superclouds. If you have too many stove pipes, then what's the purpose of supercloud? >> Right, right, right. And because, you know, historically, you built a router and you sold it to the customer. And the poor customer was supposed to install it all, you know, and interconnect all those things. And if you have 40, 50,000 router network, which we saw in our lifetime, 'cause there used to be many more branches when we were growing up in the networking industry, right? You had to create hierarchy and all kinds of things to figure out how to solve that problem. We are no longer living in that world anymore. You cannot deploy individual virtual instances. And that's what approach a lot of people are taking, which is a pure overly network. You cannot take that approach anymore. You have to evolve the architecture and then build the solution based on that architecture so that it becomes a platform which is readily available, highly scalable, and available. And at the same time, it's very, very easy to deploy. It's a SaaS type solution, right? >> So you're saying, do the architecture to get the solution for the platform that the customer has. >> Amir: Yes. >> They're not buying a platform, they end up with a platform- >> With the platform. >> as a result of Supercloud path. All right. So that's what's, so you mentioned, that's a great point. I want to double click on what you just said. 'Cause I like that what you said. What's the deployment strategy in your mind for supercloud? I'm an architect. I'm at an enterprise in the Midwest. I'm an insurance company, got some cloud action going on. I'm mostly on-premise. I've got the mandate to transform the company. We have apps. We'll be fully transformed in five years. What's my strategy? What do I do? >> Amir: The resources. >> What's the deployment strategy? Single global instance, code in every region, on every cloud? >> It needs to be a solution which is available as a SaaS service, right? So from the customer's perspective, they are onboarding into the supercloud. And then the supercloud is allowing them to do whatever they used to do, you know, historically and in the new world, right? That needs to come together. And that's what we have built is that, we have brought everything together in a way that what used to take months or years, and now taking an hour or two hours, and then people test it for a week or so and deploy it in production. >> I want to bring up something we were talking about before we were on camera about the TCP/IP, the OSI model. That was a concept that destroyed the proprietary narcissist. Work operating systems of the mini computers, which brought in an era of tech prosperity for generations. TCP/IP was kind of the magical moment that allowed for that kind of super networking connection. Inter networking is what's called as a category. It feels like something's going on here with supercloud. The way you describe it, it feels like there's this unification idea. Like the reality is we've got multiple stuff sitting around by default, you either clean it up or get rid of it, right? Or it's almost a, it's either a nuance, a new nuisance or chaos. >> Yeah. And we live in the new world now. We don't have the luxury of time. So we need to move as fast as possible to solve the business problems. And that's what we are running into. If we don't have automated solutions which scale, which solve our problems, then it's going to be a problem. And that's why SaaS is so important in today's world. Why should we have to deploy the network piecemeal? Why can't we have a solution? We solve our problem as we move forward and we accomplish what we need to accomplish and move forward. >> And we don't really need standards here, dude. It's not that we need a standards body if you have unification. >> So because things move so fast, there's no time to create a standards body. And that's why you see companies like ours popping up, which are trying to create a common infrastructure across all clouds. Otherwise if we vent the standardization path may take long. Eventually, we should be going in that direction. But we don't have the luxury of time. That's what I was trying to get to. >> Well, what's interesting is, is that to your point about standards and ratification, what ratifies a defacto anything? In the old days there was some technical bodies involved, but here, I think developers drive everything. So if you look at the developers and how they're voting with their code. They're instantly, organically defining everything as a collective intelligence. >> And just like you're putting out the paper and making it available, everybody's contributing to that. That's why you need to have APIs and terra form type constructs, which are available so that the customers can continue to improve upon that. And that's the Net DevOps, right? So that you need to have. >> What was once sacrilege, just sayin', in business school, back in the days when I got my business degree after my CS degree was, you know, no one wants to have a better mousetrap, a bad business model to have a better mouse trap. In this case, the better mouse trap, the better solution actually could be that thing. >> It is that thing. >> I mean, that can trigger, tips over the industry. >> And that that's where we are seeing our customers. You know, I mean, we have some publicly referenceable customers like Coke or Warner Music Group or, you know, multiple others and chart industries. The way we are solving the problem. They have some of the largest environments in the industry from the cloud perspective. And their whole network infrastructure is running on the Alkira infrastructure. And they're able to adopt new clouds within days rather than waiting for months to architect and then deploy and then figure out how to manage it and operate it. It's available as a service. >> John: And we've heard from your customer, Warner, they were just on the program. >> Amir: Yes. Okay, okay. >> So they're building a supercloud. So superclouds aren't just for tech companies. >> Amir: No. >> You guys build a supercloud for networking. >> Amir: It is. >> But people are building their own superclouds on top of all this new stuff. Talk about that dynamic. >> Healthcare providers, financials, high-tech companies, even startups. One of our startup customers, Tekion, right? They have these dealerships that they provide sales and support services to across the globe. And for them to be able to onboard those dealerships, it is 80% less time to production. That is real money, right? So, maybe Atif can give you a lot more examples of customers who are deploying. >> Talk about some of the customer activity. What are they like? Are they laggards, they innovators? Are they trying to hit the easy button? Are they coming in late or are you got some high customers? >> Actually most of our customers, all of our customers or customers in general. I don't think they have a choice but to move in this direction because, you know, the cloud has, like everything is quick now. So the cloud teams are moving faster in these enterprises. So now that they cannot afford the network nor to keep up pace with the cloud teams. So, they don't have a choice but to go with something similar where you can, you know, build your network on demand and bring up your network as quickly as possible to meet all those use cases. So, I'll give you an example. >> John: So the demand's high for what you guys do. >> Demand is very high because the cloud teams have- >> John: Yeah. They're going fast. >> They're going fast and there's no stopping. And then network teams, they have to keep up with them. And you cannot keep deploying, you know, networks the way you used to deploy back in the day. And as far as the use cases are concerned, there are so many use cases which our customers are using our platform for. One of the use cases, I'll give you an example of these financial customers. Some of the financial customers, they have their customers who they provide data, like stock exchanges, that provide like market data information to their customers out of data centers part. But now, their customers are moving into the cloud as well. So they need to come in from the cloud. So when they're coming in from the cloud, you cannot be giving them data from your data center because that takes time, and your hair pinning everything back. >> Moving data is like moving, moving money, someone said. >> Exactly. >> Exactly. And the other thing is like you have to optimize your traffic flows in the cloud as well because every time you leave the cloud, you get charged a lot. So, you don't want to leave the cloud unless you have to leave the cloud, your traffic. So, you have to come up or use a service which allows you to optimize all those traffic flows as well, you know? >> My final question to you guys, first of all, thanks for coming on Supercloud Program. Really appreciate it. Congratulations on your success. And you guys have a great positioning and I'm a big fan. And I have to ask, you guys are agile, nimble startup, smart on the cutting edge. Supercloud concept seems to resonate with people who are kind of on the front range of this major wave. While all the incumbents like Cisco, Microsoft, even AWS, they're like, I think they're looking at it, like what is that? I think it's coming up really fast, this trend. Because I know people talk about multi-cloud, I get that. But like, this whole supercloud is not just SaaS, it's more going on there. What do you think is going on between the folks who get it, supercloud, get the concept, and some are who are scratching their heads, whether it's the Ciscos or someone, like I don't get it. Why is supercloud important for the folks that aren't really seeing it? >> So first of all, I mean, the customers, what we saw about six months, 12 months ago, were a little slower to adopt the supercloud kind of concept. And there were leading edge customers who were coming and adopting it. Now, all of a sudden, over the last six to nine months, we've seen a flurry of customers coming in and they are from all disciplines or all very diverse set of customers. And they're starting to see the value of that because of the practical implications of what they're doing. You know, these shadow IT type environments are no longer working and there's a lot of pressure from the management to move faster. And then that's where they're coming in. And perhaps, Atif, if you can give a few examples of. >> Yeah. And I'll also just add to your point earlier about the network needing to be there 'cause the cloud teams are like, let's go faster. And the network's always been slow because, but now, it's been almost turbocharged. >> Atif: Yeah. Yeah, exactly. And as I said, like there was no choice here. You had to move in this industry. And the other thing I would add a little bit is now if you look at all these enterprises, most of their traffic is from, even from which is coming from the on-prem, it's going to the cloud SaaS applications or public clouds. And it's more than 50% of traffic, which is leaving your, you know, what you used to call, your network or the private network. So now it's like, you know, before it used to just connect sites to data centers and sites together. Now, it's a cloud as well as the SaaS application. So it's either internet bound or the public cloud bound. So now you have to build a network quickly, which caters to all these use cases. And that's where like something- >> And you guys, your solution to me is you eliminate all that work for the customer. Now, they can treat the cloud like a bag of Legos. And do their thing. Well, I oversimplify. Well, you know I'm talking about. >> Atif: Right, exactly. >> And to answer your question earlier about what about the big companies coming in and, you know, now they slow to adopt? And, you know, what normally happens is when Cisco came up, right? There used to be 16 different protocols suites. And then we finally settled on TCP/IP and DECnet or AppleTalk or X&S or, you know, you name it, right? Those companies did not adapt to the networking the way it was supposed to be done. And guess what happened, right? So if the companies in the networking space do not adopt this new concept or new way of doing things, I think some of them will become extinct over time. >> Well, I think the force and function too is the cloud teams as well. So you got two evolutions. You got architectural relevance. That's real as impact. >> It's very important. >> Cost, speed. >> And I look at it as a very similar disruption to what Cisco's the world, very early days did to, you know, bring the networking out, right? And it became the internet. But now we are going through the cloud. It's the cloud era, right? How does the cloud evolve over the next 10, 15, 20 years? Everything's is going to be offered as a service, right? So slowly data centers go away, the network becomes a plumbing thing. Very, you know, simple to deploy. And everything on top of that is virtualized in the cloud-like manners. >> And that makes the networks hardened and more secure. >> More secure. >> It's a great way to be secure. You remember the glory days, we'll go back 15 years. The Cisco conversation was, we got to move up to stack. All the manager would fight each other. Now, what does that actually mean? Stay where we are. Stay in your lane. This is kind of like the network's version of moving up the stack because not so much up the stack, but the cloud is everywhere. It's almost horizontally scaled. >> It's extending into the on-premise. It is already moving towards the edge, right? So, you will see a lot- >> So, programmability is a big program. So you guys are hitting programmability, compatibility, getting people into an environment they're comfortable operating. So the Ops people love it. >> Exactly. >> Spans the clouds to a level of SLA management. It might not be perfectly spanning applications, but you can actually know latencies between clouds, measure that. And then so you're basically managing your network now as the overall infrastructure. >> Right. And it needs to be a very intelligent infrastructure going forward, right? Because customers do not want to wait to be able to troubleshoot. They don't want to be able to wait to deploy something, right? So, it needs to be a level of automation. >> Okay. So the question for you guys both on we'll end on is what is the enablement that, because you guys are a disruptive enabler, right? You create this fabric. You're going to enable companies to do stuff. What are some of the things that you see and your customers might be seeing as things that they're going to do as a result of having this enablement? So what are some of those things? >> Amir: Atif, perhaps you can talk through the some of the customer experience on that. >> It's agility. And we are allowing these customers to move very, very quickly and build these networks which meet all these requirements inside the cloud. Because as Amir was saying, in the cloud era, networking is changing. And if you look at, you know, going back to your comment about the existing networking vendors. Some of them still think that, you know, just connecting to the cloud using some concepts like Cloud OnRamp is cloud networking, but it's changing now. >> John: 'Cause there's apps that are depending upon. >> Exactly. And it's all distributed. Like IT infrastructure, as I said earlier, is all distributed. And at the end of the day, you have to make sure that wherever your user is, wherever your app is, you are able to connect them securely. >> Historically, it used to be about building a router bigger and bigger and bigger and bigger, you know, and then interconnecting those routers. Now, it's all about horizontal scale. You don't need to build big, you need to scale it, right? And that's what cloud brings to the customer. >> It's a cultural change for Cisco and Juniper because they have to understand that they're still could be in the game and still win. >> Exactly. >> The question I have for you, what are your customers telling you that, what's some of the anecdotal, like, 'cause you guys have a good solution, is it, "Oh my god, you guys saved my butt." Or what are some of the commentary that you hear from the customers in terms of praise and and glory from your solution? >> Oh, some even say, when we do our demo and stuff, they say it's too hard to believe. >> Believe. >> Like, too hard. It's hard, you know, it's >> I dont believe you. They're skeptics. >> I don't believe you that because now you're able to bring up a global network within minutes. With networking services, like let's say you have APAC, you know, on-prem users, cloud also there, cloud here, users here, you can bring up a global network with full routed connectivity between all these endpoints with security services. You can bring up like a firewall from a third party or our services in the middle. This is a matter of minutes now. And this is all high speed connectivity with SLAs. Imagine like before connecting, you know, Singapore to U.S. East or Hong Kong to Frankfurt, you know, if you were putting your infrastructure in columns like E-connects, you would have to go, you know, figure out like, how am I going to- >> Seal line In, connect to it? Yeah. A lot of hassles, >> If you had to put like firewalls in the middle, segmentation, you had to, you know, isolate different entities. >> That's called heavy lifting. >> So what you're seeing is, you know, it's like customer comes in, there's a disbelief, can you really do that? And then they try it out, they go, "Wow, this works." Right? It's deployed in a small environment. And then all of a sudden they start taking off, right? And literally we have seen customers go from few thousand dollars a month or year type deployments to multi-million dollars a year type deployments in very, very short amount of time, in a few months. >> And you guys are pay as you go? >> Pay as you go. >> Pay as go usage cloud-based compatibility. >> Exactly. And it's amazing once they get to deploy the solution. >> What's the variable on the cost? >> On the cost? >> Is it traffic or is it. >> It's multiple different things. It's packaged into the overall solution. And as a matter of fact, we end up saving a lot of money to the customers. And not only in one way, in multiple different ways. And we do a complete TOI analysis for the customers. So it's bandwidth, it's number of connections, it's the amount of compute power that we are using. >> John: Similar things that they're used to. >> Just like the cloud constructs. Yeah. >> All right. Networking supercloud. Great. Congratulations. >> Thank you so much. >> Thanks for coming on Supercloud. >> Atif: Thank you. >> And looking forward to seeing more of the demand. Translate, instant networking. I'm sure it's going to be huge with the edge exploding. >> Oh yeah, yeah, yeah, yeah. >> Congratulations. >> Thank you so much. >> Thank you so much. >> Okay. So this is Supercloud 2 event here in Palo Alto. I'm John Furrier. The network Supercloud is here. Checkout Alkira. I'm John Furry, the host. Thanks for watching. (lively music)

(upbeat music) (upbeat music) >> Hi everybody, welcome back to our coverage of the Cloud Native Security Con. I'm Dave Vellante, here in our Boston studio. We're connecting today with Palo Alto, with John Furrier and Lisa Martin. We're also live from the show floor in Seattle. But right now, I'm here with Andy Thurai who's from Constellation Research, friend of theCUBE, and we're going to discuss the intersection of AI and security, the potential of AI, the risks and the future. Andy, welcome, good to see you again. >> Good to be here again. >> Hey, so let's get into it, can you talk a little bit about, I know this is a passion of yours, the ethical considerations surrounding AI. I mean, it's front and center in the news, and you've got accountability, privacy, security, biases. Should we be worried about AI from a security perspective? >> Absolutely, man, you should be worried. See the problem is, people don't realize this, right? I mean, the ChatGPT being a new shiny object, it's all the craze that's about. But the problem is, most of the content that's produced either by ChatGPT or even by others, it's an access, no warranties, no accountability, no whatsoever. Particularly, if it is content, it's okay. But if it is something like a code that you use for example, one of their site projects that GitHub's co-pilot, which is actually, open AI + Microsoft + GitHub's combo, they allow you to produce code, AI writes code basically, right? But when you write code, problem with that is, it's not exactly stolen, but the models are created by using the GitHub code. Actually, they're getting sued for that, saying that, "You can't use our code". Actually there's a guy, Tim Davidson, I think he's named the professor, he actually demonstrated how AI produces exact copy of the code that he has written. So right now, it's a lot of security, accountability, privacy issues. Use it either to train or to learn. But in my view, it's not ready for enterprise grade yet. >> So, Brian Behlendorf today in his keynotes said he's really worried about ChatGPT being used to automate spearfishing. So I'm like, okay, so let's unpack that a little bit. Is the concern there that it just, the ChatGPT writes such compelling phishing content, it's going to increase the probability of somebody clicking on it, or are there other dimensions? >> It could, it's not necessarily just ChatGPT for that matter, right? AI can, actually, the hackers are using it to an extent already, can use to individualize content. For example, one of the things that you are able to easily identify when you're looking at the emails that are coming in, the phishing attack is, you look at some of the key elements in it, whether it's a human or even if it's an automated AI based system. They look at certain things and they say, "Okay, this is phishing". But if you were to read an email that looks exact copy of what I would've sent to you saying that, "Hey Dave, are you on for tomorrow? Or click on this link to do whatever. It could individualize the message. That's where the volume at scale to individual to masses, that can be done using AI, which is what scares me. >> Is there a flip side to AI? How is it being utilized to help cybersecurity? And maybe you could talk about some of the more successful examples of AI in security. Like, are there use cases or are there companies out there, Andy, that you find, I know you're close to a lot of firms that are leading in this area. You and I have talked about CrowdStrike, I know Palo Alto Network, so is there a positive side to this story? >> Yeah, I mean, absolutely right. Those are some of the good companies you mentioned, CrowdStrike, Palo Alto, Darktrace is another one that I closely follow, which is a good company as well, that they're using AI for security purposes. So, here's the thing, right, when people say, when they're using malware detection systems, most of the malware detection systems that are in today's security and malware systems, use some sort of a signature and pattern scanning in the malware. You know how many identified malwares are there today in the repository, in the library? More than a billion, a billion. So, if you are to check for every malware in your repository, that's not going to work. The pattern based recognition is not going to work. So, you got to figure out a different way of identification of pattern of usage, not just a signature in a malware, right? Or there are other areas you could use, things like the usage patterns. For example, if Andy is coming in to work at a certain time, you could combine a facial recognition saying, that should he be in here at that time, and should he be doing things, what he is supposed to be doing. There are a lot of things you could do using that, right? And the AIOps use cases, which is one of my favorite areas that I work, do a lot of work, right? That it has use cases for detecting things that are anomaly, that are not supposed to be done in a way that's supposed to be, reducing the noise so it can escalate only the things what you're supposed to. So, AIOps is a great use case to use in security areas which they're not using it to an extent yet. Incident management is another area. >> So, in your malware example, you're saying, okay, known malware, pretty much anybody can deal with that now. That's sort of yesterday's problem. >> The unknown is the problem. >> It's the unknown malware really trying to understand the patterns, and the patterns are going to change. It's not like you're saying a common signature 'cause they're going to use AI to change things up at scale. >> So, here's the problem, right? The malware writers are also using AI now, right? So, they're not going to write the old malware, send it to you. They are actually creating malware on the fly. It is possible entirely in today's world that they can create a malware, drop in your systems and it'll it look for the, let me get that name right. It's called, what are we using here? It's called the TTPs, Tactics, Techniques and procedures. It'll look for that to figure out, okay, am I doing the right pattern? And then malware can sense it saying that, okay, that's the one they're detecting. I'm going to change it on the fly. So, AI can code itself on the fly, rather malware can code itself on the fly, which is going to be hard to detect. >> Well, and when you talk about TTP, when you talk to folks like Kevin Mandia of Mandiant, recently purchased by Google or other of those, the ones that have the big observation space, they'll talk about the most malicious hacks that they see, involve lateral movement. So, that's obviously something that people are looking for, AI's looking for that. And of course, the hackers are going to try to mask that lateral movement, living off the land and other things. How do you see AI impacting the future of cyber? We talked about the risks and the good. One of the things that Brian Behlendorf also mentioned is that, he pointed out that in the early days of the internet, the protocols had an inherent element of trust involved. So, things like SMTP, they didn't have security built in. So, they built up a lot of technical debt. Do you see AI being able to help with that? What steps do you see being taken to ensure that AI based systems are secure? >> So, the major difference between the older systems and the newer systems is the older systems, sadly even today, a lot of them are rules-based. If it's a rules-based systems, you are dead in the water and not able, right? So, the AI-based systems can somewhat learn from the patterns as I was talking about, for example... >> When you say rules-based systems, you mean here's the policy, here's the rule, if it's not followed but then you're saying, AI will blow that away, >> AI will blow that away, you don't have to necessarily codify things saying that, okay, if this, then do this. You don't have to necessarily do that. AI can somewhat to an extent self-learn saying that, okay, if that doesn't happen, if this is not a pattern that I know which is supposed to happen, who should I escalate this to? Who does this system belong to? And the other thing, the AIOps use case we talked about, right, the anomalies. When an anomaly happens, then the system can closely look at, saying that, okay, this is not normal behavior or usage. Is that because system's being overused or is it because somebody's trying to access something, could look at the anomaly detection, anomaly prevention or even prediction to an extent. And that's where AI could be very useful. >> So, how about the developer angle? 'Cause CNCF, the event in Seattle is all around developers, how can AI be integrated? We did a lot of talk at the conference about shift-left, we talked about shift-left and protect right. Meaning, protect the run time. So, both are important, so what steps should be taken to ensure that the AI systems are being developed in a secure and ethically sound way? What's the role of developers in that regard? >> How long do you got? (Both laughing) I think it could go for base on that. So, here's the problem, right? Lot of these companies are trying to see, I mean, you might have seen that in the news that Buzzfeed is trying to hire all of the writers to create the thing that ChatGPT is creating, a lot of enterprises... >> How, they're going to fire their writers? >> Yeah, they replace the writers. >> It's like automated automated vehicles and automated Uber drivers. >> So, the problem is a lot of enterprises still haven't done that, at least the ones I'm speaking to, are thinking about saying, "Hey, you know what, can I replace my developers because they are so expensive? Can I replace them with AI generated code?" There are a few issues with that. One, AI generated code is based on some sort of a snippet of a code that has been already available. So, you get into copyright issues, that's issue number one, right? Issue number two, if AI creates code and if something were to go wrong, who's responsible for that? There's no accountability right now. Or you as a company that's creating a system that's responsible, or is it ChatGPT, Microsoft is responsible. >> Or is the developer? >> Or the developer. >> The individual developer might be. So, they're going to be cautious about that liability. >> Well, so one of the areas where I'm seeing a lot of enterprises using this is they are using it to teach developers to learn things. You know what, if you're to code, this is a good way to code. That area, it's okay because you are just teaching them. But if you are to put an actual production code, this is what I advise companies, look, if somebody's using even to create a code, whether with or without your permission, make sure that once the code is committed, you validate that the 100%, whether it's a code or a model, or even make sure that the data what you're feeding in it is completely out of bias or no bias, right? Because at the end of the day, it doesn't matter who, what, when did that, if you put out a service or a system out there, it is involving your company liability and system, and code in place. You're going to be screwed regardless of what, if something were to go wrong, you are the first person who's liable for it. >> Andy, when you think about the dangers of AI, and what keeps you up at night if you're a security professional AI and security professional. We talked about ChatGPT doing things, we don't even, the hackers are going to get creative. But what worries you the most when you think about this topic? >> A lot, a lot, right? Let's start off with an example, actually, I don't know if you had a chance to see that or not. The hackers used a bank of Hong Kong, used a defect mechanism to fool Bank of Hong Kong to transfer $35 million to a fake account, the money is gone, right? And the problem that is, what they did was, they interacted with a manager and they learned this executive who can control a big account and cloned his voice, and clone his patterns on how he calls and what he talks and the whole name he has, after learning that, they call the branch manager or bank manager and say, "Hey, you know what, hey, move this much money to whatever." So, that's one way of kind of phishing, kind of deep fake that can come. So, that's just one example. Imagine whether business is conducted by just using voice or phone calls itself. That's an area of concern if you were to do that. And imagine this became an uproar a few years back when deepfakes put out the video of Tom Cruise and others we talked about in the past, right? And Tom Cruise looked at the video, he said that he couldn't distinguish that he didn't do it. It is so close, that close, right? And they are doing things like they're using gems... >> Awesome Instagram account by the way, the guy's hilarious, right? >> So, they they're using a lot of this fake videos and fake stuff. As long as it's only for entertainment purposes, good. But imagine doing... >> That's right there but... >> But during the election season when people were to put out saying that, okay, this current president or ex-president, he said what? And the masses believe right now whatever they're seeing in TV, that's unfortunate thing. I mean, there's no fact checking involved, and you could change governments and elections using that, which is scary shit, right? >> When you think about 2016, that was when we really first saw, the weaponization of social, the heavy use of social and then 2020 was like, wow. >> To the next level. >> It was crazy. The polarization, 2024, would deepfakes... >> Could be the next level, yeah. >> I mean, it's just going to escalate. What about public policy? I want to pick your brain on this because I I've seen situations where the EU, for example, is going to restrict the ability to ship certain code if it's involved with critical infrastructure. So, let's say, example, you're running a nuclear facility and you've got the code that protects that facility, and it can be useful against some other malware that's outside of that country, but you're restricted from sending that for whatever reason, data sovereignty. Is public policy, is it aligned with the objectives in this new world? Or, I mean, normally they have to catch up. Is that going to be a problem in your view? >> It is because, when it comes to laws it's always miles behind when a new innovation happens. It's not just for AI, right? I mean, the same thing happened with IOT. Same thing happened with whatever else new emerging tech you have. The laws have to understand if there's an issue and they have to see a continued pattern of misuse of the technology, then they'll come up with that. Use in ways they are ahead of things. So, they put a lot of restrictions in place and about what AI can or cannot do, US is way behind on that, right? But California has done some things, for example, if you are talking to a chat bot, then you have to basically disclose that to the customer, saying that you're talking to a chat bot, not to a human. And that's just a very basic rule that they have in place. I mean, there are times that when a decision is made by the, problem is, AI is a black box now. The decision making is also a black box now, and we don't tell people. And the problem is if you tell people, you'll get sued immediately because every single time, we talked about that last time, there are cases involving AI making decisions, it gets thrown out the window all the time. If you can't substantiate that. So, the bottom line is that, yes, AI can assist and help you in making decisions but just use that as a assistant mechanism. A human has to be always in all the loop, right? >> Will AI help with, in your view, with supply chain, the software supply chain security or is it, it's always a balance, right? I mean, I feel like the attackers are more advanced in some ways, it's like they're on offense, let's say, right? So, when you're calling the plays, you know where you're going, the defense has to respond to it. So in that sense, the hackers have an advantage. So, what's the balance with software supply chain? Are the hackers have the advantage because they can use AI to accelerate their penetration of the software supply chain? Or will AI in your view be a good defensive mechanism? >> It could be but the problem is, the velocity and veracity of things can be done using AI, whether it's fishing, or malware, or other security and the vulnerability scanning the whole nine yards. It's scary because the hackers have a full advantage right now. And actually, I think ChatGPT recently put out two things. One is, it's able to direct the code if it is generated by ChatGPT. So basically, if you're trying to fake because a lot of schools were complaining about it, that's why they came up with the mechanism. So, if you're trying to create a fake, there's a mechanism for them to identify. But that's a step behind still, right? And the hackers are using things to their advantage. Actually ChatGPT made a rule, if you go there and read the terms and conditions, it's basically honor rule suggesting, you can't use this for certain purposes, to create a model where it creates a security threat, as that people are going to listen. So, if there's a way or mechanism to restrict hackers from using these technologies, that would be great. But I don't see that happening. So, know that these guys have an advantage, know that they're using AI, and you have to do things to be prepared. One thing I was mentioning about is, if somebody writes a code, if somebody commits a code right now, the problem is with the agile methodologies. If somebody writes a code, if they commit a code, you assume that's right and legit, you immediately push it out into production because need for speed is there, right? But if you continue to do that with the AI produced code, you're screwed. >> So, bottom line is, AI's going to speed us up in a security context or is it going to slow us down? >> Well, in the current version, the AI systems are flawed because even the ChatGPT, if you look at the the large language models, you look at the core piece of data that's available in the world as of today and then train them using that model, using the data, right? But people are forgetting that's based on today's data. The data changes on a second basis or on a minute basis. So, if I want to do something based on tomorrow or a day after, you have to retrain the models. So, the data already have a stale. So, that in itself is stale and the cost for retraining is going to be a problem too. So overall, AI is a good first step. Use that with a caution, is what I want to say. The system is flawed now, if you use it as is, you'll be screwed, it's dangerous. >> Andy, you got to go, thanks so much for coming in, appreciate it. >> Thanks for having me. >> You're very welcome, so we're going wall to wall with our coverage of the Cloud Native Security Con. I'm Dave Vellante in the Boston Studio, John Furrier, Lisa Martin and Palo Alto. We're going to be live on the show floor as well, bringing in keynote speakers and others on the ground. Keep it right there for more coverage on theCUBE. (upbeat music) (upbeat music) (upbeat music) (upbeat music)

>>Okay, welcome back to everyone. Cube's live coverage of VMware Explorer, 2022. I'm John fur, host of the cube. Two sets three days of live coverage. Dave Ante's here. Lisa Martin, Dave Nicholson, all host of the cube 12 interviews today, just we're with Rocklin and rolling, getting down to the end of the show. As we wind down and look back and look at the future. We've got Steven Jones. Here's the general manager of the VMware cloud on AWS. He's with Amazon web service. Steven Jones. Welcome to the cube. >>Thanks John. >>Welcome back cube alumni. I've been on many times going back to 2015. Yeah. >>Pleasure to be here. Great >>To see you again. Thanks for coming on. Obviously 10 years at AWS, what a ride is that's been, come on. That's fantastic. Tell me it's been crazy. >>Wow. Learned a lot of stuff along the way, right? I mean, we, we, we knew that there was a lot of opportunity, right? Customers wanting the agility and flexibility of, of the cloud and, and we, we still think it's early days, right? I mean, you'll hear Andy say that animals say that, but it really is. Right. If you look at even just the amount of spend that's being spent on, on clouds, it's in the billions, right. And the amount of, of spend in it is still in the trillion. So there's, there's a long way to go and customers are pushing us hard. Obviously >>It's been interesting a lot going on with VM. We're obviously around with them, obviously changing the strategy with their, their third generation and their narrative. Obviously the Broadcom thing is going on around them. And 10 years at abs, we've been, we've been, this'll be our ninth year, no 10th year at reinvent coming up for us. So, but it's 10 years of everything at Amazon, 10 years of S three, 10 years of C two. So if you look at the, the marks of time, now, the history books are starting to be written about Amazon web services. You know, it's about 10 years of full throttle cube hyperscaler in action. I mean, I'm talking about real growth, like >>Hardcore, for sure. I'll give you just one anecdote. So when I first joined, I think we had maybe two EC two instances back in the day and the maximum amount of memory you could conversion into one of these machines was I think 128 gig of Ram fast forward to today. You literally can get a machine with 24 terabytes of Ram just in insane amounts. Right? My, my son who's a gamer tells me he's got 16 gig in his, in his PC. You need to, he thinks that's a lot. >>Yeah. >>That's >>Excited about that. That's not even on his graphics card. I mean, he's, I know it's coming next. The GPU, I mean, just all >>The it's like, right? >>I mean, all the hardware innovation that you guys have done, I mean, look at every it's changed. Everyone's changed their strategy to copy AWS nitro, Dave ante. And I talk about this all the time, especially with James Hamilton and the team over there, Peter DeSantos, these guys have, are constantly going at the atoms and innovating at the, at the level. I mean that, that's how hardcore it is over there right now. I mean, and the advances on the Silicon graviton performance wise is crazy. I mean, so what does that enabling? So given that's continuing, you guys are continuing to do great work there on the CapEx side, we think that's enabling another set of new net new applications because we're starting to see new things emerge. We saw snowflake come on, customer of AWS refactor, the data warehouse, they call it a data cloud. You're starting to see Goldman Sachs. You see capital one, you see enterprise customers building on top of AWS and building a cloud business without spending the CapEx >>Is exactly right. And Ziggy mentioned graviton. So graviton is one of our fastest growing compute families now. And you know, you mentioned a couple of ISVs and partners of ours who are leaning in heavily on porting their own software. Every event Adam announced that we're working with SAP to, to help them port their HANA cloud, which is a, a database of service offering HANA flagship to graviton as well. So it's, it's definitely changing. >>And I think, you know, one of the, and we're gonna circle back to VMware is kind of a point to this. This conversation is that, is that if you look at the trends, right, okay. VMware really tried hard to do cloud and they had a good shot at it V cloud air, but it just, they didn't have the momentum that you guys had at AWS. We saw a lot, lot of other stragglers try to do cloud. They fell off the road, OpenStack, HP, and the list goes on and on. I don't wanna get into that, but the point is, as you guys become more powerful and you're open, right? So you have open ecosystem, you have people now coming back, taking advantage and refactoring and picking up where they left off. VMware was the one of the first companies that actually said, you know what pat Gelsinger said? And I was there, let's clear up the positioning. Let's go all in with AWS. That's >>Right >>At that time, 2016. >>Yeah. This was new for us, for >>Sure. And then now that's set the standard. Now everybody else is kind of doing it. Where is the VMware cloud relationship right now? How is that going out? State's worked. >>It's working well very well. It's I mean, we're celebrating, I think we made the announcement what, five years ago at this conference. Yeah. 2016. So, I mean, it's, it's been a tremendous ride. The best part are the customers who were coming and adopting and proving to us that our vision back then was the right vision. And, and, and what's been different. I think about this relationship. And it was new for us was that we, we purposely went after a jointly engineered solution. This wasn't a, we've got a, a customer or a partner that's just going to run and build something on us. This is something where we both bring muscle and we actually build a, a joint offering together. Talk about, about the main difference. >>Yeah. And that, and that's been working, but now here at this show, if you look at, if you squint through the multi-cloud thing, which is like just, I think positioning for, you know, what could happen in, in a post broad Broadcom world, the cloud native has traction they're Tansu where, where customers were leaning in. So their enterprise customer is what I call the classic. It, you know, mainstream enterprise, which you guys have been doing a lot of business with. They're now thinking, okay, I'm gonna go on continu, accelerate on, in the public cloud, but I'm gonna have hybrid on premise as well. You guys have that solution. Now they're gonna need cloud native. And we were speculating that VMware is probably not gonna be able to get 'em all of it. And, and that there's a lot more cloud native options as customers want more cloud native. How do you see that piece on Amazon side? Because there's a lot of benefits between the VMware cloud on AWS and the services that you guys have natively in your cloud. So we see customers really taking advantage of the AWS goodness, as well as expanding the cloud side at VMware cloud on AWS. >>Yeah. There's probably two ways I would look at this. Right? So, so one is the combination of VMware cloud on AWS. And then both native services just generally brings more options to customers. And so typically what we're seeing now is customers are just able to move much faster, especially as it comes to data center, evacuations, migrating all their assets, right? So it used to be that, and still some customers they're like, I I've gotta think through my entire portfolio of applications and decide what to refactor. And the only way I can move it to cloud is to actually refactor it into some net new application, more and more. We're actually seeing customers. They've got their assets. A lot of them are still on premises in a VMware state, right. They can move those super quick and then modernize those. And so I think where you'll see VMware and AWS very aligned is on this, this idea of migrate. Now you need to get the benefits of TCO and, and the agility that comes with being in the cloud and then modernize. We took a step further, which is, and I think VMware would agree here too, but all of the, the myriad of services, I think it's 200 plus now AWS native services are for use right alongside any that a customer wants to run in VMware. And so we have examples of customers that are doing just, >>And that's, that's how you guys see the native and, and VMware cloud integrating in. Yeah, that's, that's important because this, I mean, if I always joke about, you know, we've been here 12 years listening in the hallways and stuff, you know, on the bus to the event last night, walking the parties and whatnot, listening in the streets, there's kind of two conversations that rise right to the top. And I wanna get your reaction to this Steven, because this seems to be representative of this demographic here at VMware conference, there's conversations around ransomware and storage and D dub and recovery. It's all, a lot of those happen. Yeah. Clearly a big crowd here that care about, you know, Veeam and NetApp and storage and like making sure stuff's secure and air gapped. And a lot of that kind of, I call nerdy conversations and then the other one is, okay, I gotta get the cloud story. >>Right. So there's kind of the operational security. And then there's like, okay, what's my path to true cloud. I need to get this moving. I need to have better applications. My company is the application now not it serves some sort of back office function. Yeah. It's like, my company is completely using technology as its business. So the app is the business. So that means everything's technology driven, not departmental siloed. So there's a, that's what I call the true cloud conversation. How do you, how do you see that evolving because VMware customers are now going there. And I won't say, I won't say they're behind, but they're certainly going there faster than ever before. >>I think, I think, I mean, it's an interesting con it's an interesting way to put it and I, I would completely agree. I think it's, it's very clear that I think a lot of customer companies are actually being disrupted. Right. And they have to move fast and reinvent themselves. You said the app is now becoming the company. Right. I mean, if, if you look at where not too many years back, there were, you know, big companies like Netflix that were born in the cloud. Right. Airbnb they're disruptors. >>There's, that's the >>App, right? That's the app. Yeah. So I, I would exactly agree. And, and that's who other companies are competing with. And so they have to move quickly. You talked about some, some technology that allows them to do that, right? So this week we announced the general availability of a NetApp on tap solution. It's been available on AWS for some time as a fully managed FSX storage solution. But now customers can actually leverage it with, with VMC. Now, why is that important? Well, there's tens of thousands of customers running VMware. On-premises still, there's thousands of them that are actually using NetApp filers, right? NetApp, NetApp filers, and the same enterprise features like replication. D do you were talking about and Snapp and clone. Those types of things can be done. Now within the V VMware state on AWS, what's even better is they can actually move faster. So consider replicating all this, you know, petabytes and petabytes of data that are in these S from on-premises into AWS, this, this NetApp service, and then connected connecting that up to the BMC option. So it just allows customers much, much. >>You guys, you guys have always been customer focus. Every time I sat down with the Andy jazzy and then last year with Adam, same thing we worked back from, I know it's kind of a canned answer on some of the questions from media, but, but they do really care. I've had those conversations. You guys do work backwards from the customer, actually have documents called working backwards. But one of the things that I observed, we talked about here yesterday on the cube was the observations of reinvent versus say, VM world. Now explore is VM world's ecosystem was very partner-centric in the sense of the partners needed to rely on VMware. And the customers came here for both more of the partners, not so much VMware in the sense there wasn't as much, many, many announcements can compare that to the past, say eight years of reinvent, where there's so much Amazon action going on the partners, I won't say take as a second, has a backseat to Amazon, but the, the attendees go there generally for what's going on with AWS, because there's always new stuff coming out. >>And it's, it's amazing. But this year it starts to see that there's an overlap or, or change between like the VMware ecosystem. And now Amazon there's, a lot of our interviews are like, they're on both ecosystems. They're at Amazon's show they're here. So you start to see what I call the naturalization of partners. You guys are continuing to grow, and you'll probably still have thousands of announcements at the event this year, as you always do, but the partners are much more part of the AWS equation, not just we're leasing all these new services and, and oh, for sure. Look at us, look at Amazon. We're growing. Cause you guys were building out and look, the growth has been great. But now as you guys get to this next level, the partners are integral to the ecosystem. How do you look at that? How has Amazon thinking about that? I know there's been some, some, a lot of active reorgs around AWS around solving this problem or no solve the problem, addressing the need and this next level of growth. What's your reaction to >>That? Well, I mean, it's, it's a, it's a good point. So I have to be honest with you, John. I, I, I spent eight of my 10 years so far at AWS within the partner organization. So partners are very near and dear to my heart. We've got tens of thousands of partners and you are you're right. You're starting to see some overlap now between the VMware partner ecosystem and what we've built now in AWS and partners are big >>By the way, you sell out every reinvent. So it's, you have a lot of partners. I'm not suggesting that you, that there's no partner network there, but >>Partners are critical. I mean, absolutely naturally we want a relationship with a customer, but in order to scale the way we need to do to meet the, the needs of customers, we need partners. Right. We, we can't, we can't interact with every single customer as much as we would like to. Right. And so partners have long built teams and expertise that, that caters to even niche workloads or opportunity areas. And, and we love partners >>For that. Yeah. I know you guys do. And also we'll point out just to kind of give props to you guys on the partner side, you don't, you keep that top of the stack open on Amazon. You've done some stuff for end to end where customers want all Amazon, but for the most part, you let competition come in, even on, so you guys are definitely partner friendly. I'm just observing more the maturization of partners within the reinvent ecosystem, cuz we're there every year. I mean, it's, I mean, first of all, they're all buzzing. I mean, it's not like there's no action. There's a lot of customers there it's sold out as big numbers, but it just seems that the partners are much more integrated into the value proposition of at a AWS because of the, the rising tide and, and now their enablement, cuz now they're part of the, of the value proposition. Even more than ever before >>They, they really are. And they, and they're building a lot of capabilities and services on us. And so their customers are our customers. And like you say, it's rising tide, right. We, we all do better together. >>Okay. So let's talk about the VMware cloud here. What's the update here in terms of the show, what's your, what's your main focus cuz a lot of people here are doing, doing sessions. What's been some of the con content that you guys are producing here. >>Yeah. So the best part obviously is a always the customer conversations to partner conversations. So a, a lot of, a lot of sessions there, we did keynote yesterday in Ryan and I, where we talked about a number of announcements that are, I think pretty material now to the offering a joint announcement with NetApp yesterday as well around the storage solution I was talking about. And then some, some really good technical deep dives on how the offering works. Customers are still interested in like how, how do I take what I've got on premises and easily move into AWS and technology like HSX H CX solution with VMware makes it really easy without having to re IP applications. I mean, you know, it is super difficult sometimes to, to move an application. If you've got figure out where all the firewall rules are and re iPing those, those things source. But yeah, it's, it's been fantastic. >>A lot of migrations to the cloud too. A lot of cloud action, new cloud action. You guys have probably seen an uptake on services right on the native side. >>Yes. Yes. For sure. So maybe I just outlined some of the, some of the assets we made this week. So absolutely >>Go ahead. >>We, we announced a new instance family as a, a major workhorse underneath the VMware cloud offering called I, I, you mentioned nitro earlier, this is on, based on our latest generation of nitro, which allows us to offer as you know, bare metal instances, which is, which is what VMware actually VMware was our first partnership and customer that I would say actually drove us to really get Nira done and out the door. And we've continued to iterate on that. And so this I four, I instance, it's based on the, the latest Intel isolate processor with more than double the Ram double the compute, a whopping 75 gigabytes per second network. So it's a real powerhouse. The cool thing is that with the, with the NetApp storage solution that we, we discussed, we're now disaggregating the need to provision, compute and storage at the same time. It used to be, if you wanted to add more storage to your VSAN array, that was on a V VMware cloud. Yeah. You'd add another note. You might not need more compute for memory. You'd have to add another note. And so now customers can simply start adding chunks of storage. And so this opens up customers. I had a customer come to me yesterday and said, there's no reason for us not to move. Now. We were waiting for something that like this, that allowed us to move our data heavy workloads yeah. Into VMware cloud. It's >>Like, it's like the, the alignment. You mentioned alignment earlier. You know, I would say that VMware customers are lined up now almost perfectly with the hybrid story that's that's seamless or somewhat seems it's never truly seamless. But if you look at like what Deepak's doing with Kubernetes and open source, you, you guys have that there talking that big here, you got vs a eight vSphere, eight out it's all cloud native. So that's lined up with what you guys are doing on your services and the horsepower. They have their stuff, you have yours that works better together. So it seems like it's more lined up than ever before. What's your take on that? Do you agree? And, and if so, what folks watching here that are VMware customers, what's, what's the motivation now to go faster? >>Look, it is, it is absolutely lined up. We are, as, as I mentioned earlier, we are jointly engineering and developing this thing together. And so that includes not just the nuts and bolts underneath, but kind of the vision of where it's going. And so we're, we're collectively bringing in customer feedback. >>What is that vision real quick? >>So that vision has to actually help an under help meet even the most demanding customer workloads. Okay. So you've got customer workloads that are still locked in on premises. And why is that? Well, it used to be, there was big for data and migration, right? And the speed. And so we continue to iterate this and that again is a joint thing. Instead of say, VMware, just building on AWS, it really is a, a tight partnership. >>Yeah. The lift and shift is a, an easy thing to do. And, and, and by the way, that could be a hassle too. But I hear most people say the reason holding us back on the workloads is it's just a lot of work, a hassle making it easier is what they want. And you guys are doing that. >>We are doing that. Absolutely. And by the way, we've got not just engineering teams, but we've got customer support teams on both sides working together. We also have flexible commercial options, right? If a customer wants to buy from AWS because they've negotiated some kind of deal with us, they can do that. They wanna buy from VMware for a similar reason. They could buy from VMware. So are >>They in the marketplace? >>They are in the market. There, there are some things in the marketplace. So you talked about Tansu, there's a Tansu offering in the marketplace. So yes. Customers can >>Contract. Yeah. Marketplaces. I'm telling you that's very disruptive. I'm Billy bullish on the market AIOS marketplace. I think that's gonna be a transformative way. People have what they procure and fully agree, deploy and how, and channel relationships are gonna shift. I think that's gonna be a disruptive enabler to the partner equation and, and we haven't even seen it yet. We're gonna be up there in September for their inaugural event. I think it's a small group, but we're gonna be documenting that. So even final question for you, what's next for you? What's on the agenda. You got reinvent right around the corner. Your P ones are done. Right? I know. Assuming all that, I turn that general joke. That's an internal Amazon joke. FYI. You've got your plan. What's next for the world. Obviously they're gonna go this, take this, explore global. No matter what happens with Broadcom, this is gonna be a growth wave with hybrid. What's next for you and your team with AWS and VMware's relationship? >>Yeah. So both of us are hyper focused on adding additional options, both from a, an instance compute perspective. You know, VMware announced some, some, some additional offerings that we've got. We've got a fully complete, like, so they're, they announce things like VMware flex compute V VMware flex storage. You mentioned earlier, there was a conversation around ransomware. There's a new ransomware based offering. So we're hyper focused on rounding out, continuing to round out the offering and giving customers even more choice >>Real quick. Jonathan made me think about the ransomware we were at reinforce Steven Schmidtz now the CSO. Now you got a CSO. AJ's the CSO. You got a whole focus, huge emphasis on security right now. I know you always have, but now it's much more public. It's PO more positive, I think, than some of the other events I've been to. It's been more Lum and doom. What's the security tie in here with VMware. Can you share a little bit real quick on the security piece update around this relationship? >>Yeah, you bet. So as you know, security for us is job zero. Like you don't have anything of security. And so what are the things that, that we're excited about specifically with VMware is, is the latest offering that, that we put together and it's called this, this ransomware offering. And it's, it's a little bit different than other ransomware. I mean, a lot of people have ransomware offerings today, just >>Air gap. >>Right, right, right. Exactly. No, that's easy. No, this one is different. So on the back end, so within VMC, there's this, this option where CU we can be to be taking iterative snapshots of a customer environment. Now, if an event were to occur, right. And a customer is like, I have to know if I'm compromised, we can actually spin up super easy. This is cloud. Remember? Yeah. We can spin up a, a copy of this environment, throw a switch, pick a snapshot with NSX. So VMware NSX firewall it off and then use some custom tooling from VMware to actually see if it's been compromised or not. And then iterate through that until you actually know you're clean. And that's different than just tools that do maybe a >>Little bit of scam. We had Tom gills on yesterday and, and one of the things Dave ante had to leave is taking the sun to college is last one in the house and B nester now, but Tom Gill was on. We were talking about how good their security story is ware. And they really weren't showboating it as much as they could have here. I thought they could have done a better job, but this is an example of kind of them really leaning in with you guys. That's the key part of the relationship. >>Yeah, it really is. And I think this is something is materially different than what you can get elsewhere. And it's exciting for, >>Okay. Now the, the real question I want to know is what's your plans for AWS reinvent the blockbuster end of the year, Amazon surf show that gets bigger and bigger. I know it's still hybrid now, but it's looking be hybrid, but people are back in person last year. You guys were the first event really come back and still had massive numbers. AWS summit, New York at 19,000. I heard last week in Chicago, big numbers. So we're expecting reinvent to be pretty large this year. What are you, what are you gonna do there? What's your role there? >>We are expecting, well, I'll be there. I cover multiple businesses. Obviously. We're, we're planning on some additional announcements, obviously in the VMware space as well. And one of the other businesses I run is around SAP. And you should look for some things there as well. Yeah. Really looking forward to reinvent, except for the fact that it's right after Thanksgiving. But I think it >>Always ruins my, I always get an article out. I like, why are you we're having, we're having Thanksgiving dinner. I gotta write this article. It's gotta get Adam, Adam. Leski exclusive. We, every year we do a, a CEO sit down with Andy was the CEO and then now Adam. But yeah, it's a great event to me. I think it sets the tone. And it's gonna be very interesting to see the big clouds are coming to the big cloud. You guys, and you guys are now called hyperscalers. Now, multiple words. It's interesting. You guys are providing the CapEx goodness for everybody else now. And that relationship seems to be the new, the new industry standard of you guys provide the enablement and then everyone you get paid, cuz it's a service. A whole nother level of cloud is emerging in the partner network, GSI other companies. Yeah. >>Yeah. I mean we're really scaling. I mean we continue to iterate and release regions at a fast clip. We just announced support for VMware in Hong Kong. Yeah. So now we're up to 21 regions for this service, >>The sovereign clouds right around the corner. Let's we'll talk about that soon. Steven. Thanks for coming. I know you gotta go. Thank you for your valuable time. Coming in. Put Steven Jones. Who's the general manager of the VMware cloud on AWS business. Four AWS here inside the cube day. Three of cube coverage. I'm John furrier. Thanks for watching. We'll be right back.

(upbeat music) >> We're back at VeeamON 2022. We're here at the Aria in Las Vegas Dave Vellante with Dave Nicholson. Bill Andrews is here. He's the president and CEO of ExaGrid, mass boy. Bill, thanks for coming on theCUBE. >> Thanks for having me. >> So I hear a lot about obviously data protection, cyber resiliency, what's the big picture trends that you're seeing when you talk to customers? >> Well, I think clearly we were talking just a few minutes ago, data's growing like crazy, right This morning, I think they said it was 28% growth a year, right? So data's doubling almost just a little less than every three years. And then you get the attacks on the data which was the keynote speech this morning as well, right. All about the ransomware attacks. So we've got more and more data, and that data is more and more under attack. So I think those are the two big themes. >> So ExaGrid as a company been around for a long time. You've kind of been the steady kind of Eddy, if you will. Tell us about ExaGrid, maybe share with us some of the differentiators that you share with customers. >> Sure, so specifically, let's say in the Veeam world you're backing up your data, and you really only have two choices. You can back that up to disc. So some primary storage disc from a Dell, or a Hewlett Packard, or an NetApp or somebody, or you're going to back it up to what's called an inline deduplication appliance maybe a Dell Data Domain or an HPE StoreOnce, right? So what ExaGrid does is we've taken the best of both those but not the challenges of both those and put 'em together. So with disc, you're going to get fast backups and fast restores, but because in backup you keep weekly's, monthly's, yearly retention, the cost of this becomes exorbitant. If you go to a deduplication appliance, and let's say the Dell or the HPs, the data comes in, has to be deduplicated, compare one backup to the next to reduce that storage, which lowers the cost. So fixes that problem, but the fact that they do it inline slows the backups down dramatically. All the data is deduplicated so the restores are slow, and then the backup window keeps growing as the data grows 'cause they're all scale up technologies. >> And the restores are slow 'cause you got to rehydrate. >> You got to rehydrate every time. So what we did is we said, you got to have both. So our appliances have a front end disc cache landing zone. So you're right directed to the disc., Nothing else happens to it, whatever speed the backup app could write at that's the speed we take it in at. And then we keep the most recent backups in that landing zone ready to go. So you want to boot a VM, it's not an hour like a deduplication appliance it's a minute or two. Secondly, we then deduplicate the data into a second tier which is a repository tier, but we have all the deduplicated data for the long term retention, which gets the cost down. And on top of that, we're scale out. Every appliance has networking processor memory end disc. So if you double, triple, quadruple the data you double, triple, quadruple everything. And if the backup window is six hours at 100 terabyte it's six hours at 200 terabyte, 500 terabyte, a petabyte it doesn't matter. >> 'Cause you scale out. >> Right, and then lastly, our repository tier is non-network facing. We're the only ones in the industry with this. So that under a ransomware attack, if you get hold of a rogue server or you hack the media server, get to the backup storage whether it's disc or deduplication appliance, you can wipe out all the backup data. So you have nothing to recover from. In our case, you wipe it out, our landing zone will be wiped out. We're no different than anything else that's network facing. However, the only thing that talks to our repository tier is our object code. And we've set up security policies as to how long before you want us to delete data, let's say 10 days. So if you have an attack on Monday that data doesn't get deleted till like a week from Thursday, let's say. So you can freeze the system at any time and do restores. And then we have immutable data objects and all the other stuff. But the culmination of a non-network facing tier and the fact that we do the delayed deletes makes us the only one in the industry that can actually truly recover. And that's accelerating our growth, of course. >> Wow, great description. So that disc cache layer is a memory, it's a flash? >> It's disc, it's spinning disc. >> Spinning disc, okay. >> Yeah, no different than any other disc. >> And then the tiered is what, less expensive spinning disc? >> No, it's still the same. It's all SaaS disc 'cause you want the quality, right? So it's all SaaS, and so we use Western Digital or Seagate drives just like everybody else. The difference is that we're not doing any deduplication coming in or out of that landing zone to have fast backups and fast restores. So think of it like this, you've got disc and you say, boy it's too expensive. What I really want to do then is put maybe a deduplication appliance behind it to lower the cost or reverse it. I've got a deduplication appliance, ugh, it's too slow for backups and restores. I really want to throw this in front of it to have fast backups first. Basically, that's what we did. >> So where does the cost savings, Bill come in though, on the tier? >> The cost savings comes in the fact that we got deduplication in that repository. So only the most recent backup >> Ah okay, so I get it. >> are the duplicated data. But let's say you had 40 copies of retention. You know, 10 weekly's, 36 monthly's, a few yearly. All of that's deduplicated >> Okay, so you're deduping the stuff that's not as current. >> Right. >> Okay. >> And only a handful of us deduplicate at the layer we do. In other words, deduplication could be anywhere from two to one, up to 50 to one. I mean it's all over the place depending on the algorithm. Now it's what everybody's algorithms do. Some backup apps do two to one, some do five to one, we do 20 to one as well as much as 50 to one depending on the data types. >> Yeah, so the workload is going to largely determine the combination >> The content type, right. with the algos, right? >> Yeah, the content type. >> So the part of the environment that's behind the illogical air gap, if you will, is deduped data. >> Yes. >> So in this case, is it fair to say that you're trading a positive economic value for a little bit longer restore from that environment? >> No, because if you think about backup 95% of the customers restores are from the most recent data. >> From the disc cache. >> 95% of the time 'cause you think about why do you need fast restores? Somebody deleted a file, somebody overwrote a file. They can't go work, they can't open a file. It's encrypted, it's corrupted. That's what IT people are trying to keep users productive. When do you go for longer-term retention data? It's an SEC audit. It's a HIPAA audit. It's a legal discovery, you don't need that data right away. You have days and weeks to get that ready for that legal discovery or that audit. So we found that boundary where you keep users productive by keeping the most recent data in the disc cache landing zone, but anything that's long term. And by the way, everyone else is long term, at that point. >> Yeah, so the economics are comparable to the dedupe upfront. Are they better, obviously get the performance advance? >> So we would be a lot looped. The thing we replaced the most believe it or not is disc, we're a lot less expensive than the disc. I was meeting with some Veeam folks this morning and we were up against Cisco 3260 disc at a children's hospital. And on our quote was $500,000. The disc was 1.4 million. Just to give you an example of the savings. On a Data Domain we're typically about half the price of a Data Domain. >> Really now? >> The reason why is their front end control are so expensive. They need the fastest trip on the planet 'cause they're trying to do inline deduplication. >> Yeah, so they're chasing >> They need the fastest memory >> on the planet. >> this chips all the time. They need SSD on data to move in and out of the hash table. In order to keep up with inline, they've got to throw so much compute at it that it drives their cost up. >> But now in the case of ransomware attack, are you saying that the landing zone is still available for recovery in some circumstances? Or are you expecting that that disc landing zone would be encrypted by the attacker? >> Those are two different things. One is deletion, one is encryption. So let's do the first scenario. >> I'm talking about malicious encryption. >> Yeah, absolutely. So the first scenario is the threat actor encrypts all your primary data. What's does he go for next? The backup data. 'Cause he knows that's your belt and suspend is to not pay the ransom. If it's disc he's going to go in and put delete commands at the disc, wipe out the disc. If it's a data domain or HPE StoreOnce, it's all going to be gone 'cause it's one tier. He's going to go after our landing zone, it's going to be gone too. It's going to wipe out our landing zone. Except behind that we have the most recent backup deduplicate in the repository as well as all the other backups. So what'll happen is they'll freeze the system 'cause we weren't going to delete anything in the repository for X days 'cause you set up a policy, and then you restore the most recent backup into the landing zone or we can restore it directly to your primary storage area, right? >> Because that tier is not network facing. >> That's right. >> It's fenced off essentially. >> People call us every day of the week saying, you saved me, you saved me again. People are coming up to me here, you saved me, you saved me. >> Tell us a story about that, I mean don't give me the names but how so. >> I'll actually do a funnier story, 'cause these are the ones that our vendors like to tell. 'Cause I'm self-serving as the CEO that's good of course, a little humor. >> It's your 15 minutes of job. >> That is my 15 minutes of fame. So we had one international company who had one ExaGrid at one location, 19 Data Domains at the other locations. Ransomware attack guess what? 19 Data Domains wiped out. The one ExaGrid, the only place they could restore. So now all 20 locations of course are ExaGrids, China, Russia, Mexico, Germany, US, et cetera. They rolled us out worldwide. So it's very common for that to occur. And think about why that is, everyone who's network facing you can get to the storage. You can say all the media servers are buttoned up, but I can find a rogue server and snake my way over the storage, I can. Now, we also of course support the Veeam Data Mover. So let's talk about that since we're at a Veeam conference. We were the first company to ever integrate the Veeam Data Mover. So we were the first actually ever integration with Veeam. And so that Veeam Data Mover is a protocol that goes from Veeam to the ExaGrid, and we run it on both ends. So that's a more secure protocol 'cause it's not an open format protocol like SaaS. So with running the Veeam Data Mover we get about 30% more performance, but you do have a more secure protocol layer. So if you don't get through Veeam but you get through the protocol, boom, we've got a stronger protocol. If you make it through that somehow, or you get to it from a rogue server somewhere else we still have the repository. So we have all these layers so that you can't get at it. >> So you guys have been at this for a while, I mean decade and a half plus. And you've raised a fair amount of money but in today's terms, not really. So you've just had really strong growth, sequential growth. I understand it, and double digit growth year on year. >> Yeah, about 25% a year right now >> 25%, what's your global strategy? >> So we have sales offices in about 30 countries already. So we have three sales teams in Brazil, and three in Germany, and three in the UK, and two in France, and a lot of individual countries, Chile, Argentina, Columbia, Mexico, South Africa, Saudi, Czech Republic, Poland, Dubai, Hong Kong, Australia, Singapore, et cetera. We've just added two sales territories in Japan. We're adding two in India. And we're installed in over 50 countries. So we've been international all along the way. The goal of the company is we're growing nicely. We have not raised money in almost 10 years. >> So you're self-funding. You're cash positive. >> We are cash positive and self-funded and people say, how have you done that for 10 years? >> You know what's interesting is I remember, Dave Scott, Dave Scott was the CEO of 3PAR, and he told me when he came into that job, he told the VCs, they wanted to give him 30 million. He said, I need 80 million. I think he might have raised closer to a hundred which is right around what you guys have raised. But like you said, you haven't raised it in a long time. And in today's terms, that's nothing, right? >> 100 is 500 in today's terms. >> Yeah, right, exactly. And so the thing that really hurt 3PAR, they were public companies so you could see all this stuff is they couldn't expand internationally. It was just too damn expensive to set up the channels, and somehow you guys have figured that out. >> 40% of our business comes out of international. We're growing faster internationally than we are domestically. >> What was the formula there, Bill, was that just slow and steady or? >> It's a great question. >> No, so what we did, we said let's build ExaGrid like a McDonald's franchise, nobody's ever done that before in high tech. So what does that mean? That means you have to have the same product worldwide. You have to have the same spares model worldwide. You have to have the same support model worldwide. So we early on built the installation. So we do 100% of our installs remotely. 100% of our support remotely, yet we're in large enterprises. Customers racks and stacks the appliances we get on with them. We do the entire install on 30 minutes to about three hours. And we've been developing that into the product since day one. So we can remotely install anywhere in the world. We keep spares depots all over the world. We can bring 'em up really quick. Our support model is we have in theater support people. So they're in Europe, they're in APAC, they're in the US, et cetera. And we assign customers to the support people. So they deal with the same support person all the time. So everything is scalable. So right now we're going to open up India. It's the same way we've opened up every other country. Once you've got the McDonald's formula we just stamp it all over the world. >> That's amazing. >> Same pricing, same product same model, same everything. >> So what was the inspiration for that? I mean, you've done this since day one, which is what like 15, 16 years ago. Or just you do engineering or? >> No, so our whole thought was, first of all you can't survive anymore in this world without being an international company. 'Cause if you're going to go after large companies they have offices all over the world. We have companies now that have 17, 18, 20, 30 locations. And there were in every country in the world, you can't go into this business without being able to ship anywhere in the world and support it for a single customer. You're not going into Singapore because of that. You're going to Singapore because some company in Germany has offices in the U.S, Mexico Singapore and Australia. You have to be international. It's a must now. So that was the initial thing is that, our goal is to become a billion dollar company. And we're on path to do that, right. >> You can see a billion. >> Well, I can absolutely see a billion. And we're bigger than everybody thinks. Everybody guesses our revenue always guesses low. So we're bigger than you think. The reason why we don't talk about it is we don't need to. >> That's the headline for our writers, ExaGrid is a billion dollar company and nobody's know about it. >> Million dollar company. >> On its way to a billion. >> That's right. >> You're not disclosing. (Bill laughing) But that's awesome. I mean, that's a great story. I mean, you kind of are a well kept secret, aren't you? >> Well, I dunno if it's a well kept secret. You know, smaller companies never have their awareness of big companies, right? The Dells of the world are a hundred billion. IBM is 70 billion, Cisco is 60 billion. Easy to have awareness, right? If you're under a billion, I got to give a funny story then I think we got to close out here. >> Oh go ahead please. >> So there's one funny story. So I was talking to the CIO of a super large Fortune 500 company. And I said to him, "Just so who do you use?" "I use IBM Db2, and I use, Cisco routers, and I use EMC primary storage, et cetera. And I use all these big." And I said, "Would you ever switch from Db2?" "Oh no, the switching costs would kill me. I could never go to Oracle." So I said to him, "Look would you ever use like a Pure Storage, right. A couple billion dollar company." He says, "Who?" >> Huh, interesting. >> I said to him, all right so skip that. I said, "VMware, would you ever think about going with Nutanix?" "Who?" Those are billion dollar plus companies. And he was saying who? >> Public companies. >> And he was saying who? That's not uncommon when I talk to CIOs. They see the big 30 and that's it. >> Oh, that's interesting. What about your partnership with Veeam? Tell us more about that. >> Yeah, so I would actually, and I'm going to be bold when I say this 'cause I think you can ask anybody here at the conference. We're probably closer first of all, to the Veeam sales force than any company there is. You talk to any Veeam sales rep, they work closer with ExaGrid than any other. Yeah, we are very tight in the field and have been for a long time. We're integrated with the Veeam Data Boomer. We're integrated with SOBR. We're integrated with all the integrations or with the product as well. We have a lot of joint customers. We actually do a lot of selling together, where we go in as Veeam ExaGrid 'cause it's a great end to end story. Especially when we're replacing, let's say a Dell Avamar to Dell Data Domain or a Dell Network with a Dell Data Domain, very commonly Veeam ExaGrid go in together on those types of sales. So we do a lot of co-selling together. We constantly train their systems engineers around the world, every given week we're training either inside sales teams, and we've trained their customer support teams in Columbus and Prague. So we're very tight with 'em we've been tight for over a decade. >> Is your head count public? Can you share that with us? >> So we're just over 300 employees. >> Really, wow. >> We have 70 open positions, so. >> Yeah, what are you looking for? Yeah, everything, right? >> We are looking for engineers. We are looking for customer support people. We're looking for marketing people. We're looking for inside sales people, field people. And we've been hiring, as of late, major account reps that just focus on the Fortune 500. So we've separated that out now. >> When you hire engineers, I mean I think I saw you were long time ago, DG, right? Is that true? >> Yeah, way back in the '80s. >> But systems guy. >> That's how old I am. >> Right, systems guy. I mean, I remember them well Eddie Castro and company. >> Tom West. >> EMV series. >> Tom West was the hero of course. >> The EMV 4000, the EMV 20,000, right? >> When were kids, "The Soul of a New Machine" was the inspirational book but anyway, >> Yeah Tracy Kidder, it was great. >> Are you looking for systems people, what kind of talent are you looking for in engineering? >> So it's a lot of Linux programming type stuff in the product 'cause we run on a Linux space. So it's a lot of Linux programs so its people in those storage. >> Yeah, cool, Bill, hey, thanks for coming on to theCUBE. Well learned a lot, great story. >> It's a pleasure. >> That was fun. >> Congratulations. >> Thanks. >> And good luck. >> All right, thank you. >> All right, and thank you for watching theCUBE's coverage of VeeamON 2022, Dave Vellante for Dave Nicholson. We'll be right back right after this short break, stay with us. (soft beat music)

(upbeat music) >> Welcome back to Las Vegas everybody. You're watching theCUBE coverage of AWS reinvent 2021. I tell you this place is packed. It's quite amazing here, over 20,000 people, I'd say it's closer to 25, maybe 27,000, and it's whole overflow, lots going on in the evenings. It's quite remarkable and we're really happy to be part of this. Jay Turner is here, he's the Vice President of Development and Operations, at PCCW Global. He's joined by Garrett Lowell, Vice President of Ecosystem Partnerships for the Americas at PCCW Global. Guys, welcome to theCUBE. Thanks for coming on. >> Thank you. >> Thank you so much. Jay, maybe you could take us through, for those people who aren't familiar with your company, what do you guys do, what are you all about? >> PCCW Global is the international operating wing of Hong Kong telecom. If it's outside of Hong Kong, it's our network. We've got about 695,000 kilometers of diverse cable, we've got about 43, 44 terabit of capacity came into business in 2005, if my brain is serving me correctly right now. We have a very diverse and vast portfolio ranging all the way from satellite teleports, all the way to IP transit. We're a Tier 1 service provider from that perspective as well. We do one of everything when it comes to networking and that's really, what was the basis of Console Connect, was inventing a platform to really enable our users to capitalize on our network and our assets. >> Okay. 2005, obviously you predated Cloud, you laid a bunch of fibers struck it in the ocean, I mean, global networks. There was a big trend to do that you had to think, you had to go bigger, go home in that business, (laughing) all right. Console Connect is your platform, is that right? >> Jay: Yes. >> So explain- >> Yeah, sorry, Console Connect is a software defined interconnection platform. We built a user self-service portal. Users can allocate ports, they get the LOAs issued to them directly from the platform. And then once they've got an active port or they've come in via one of our partnerships, they can then provision connectivity across our platform. That may be extending to their data centers or extending to their branch office, or it could be building a circuit into the Cloud via direct connect, could be building a circuit into an internet exchange. All of those circuits are going to be across that 685,000 kilometers of diverse fiber rather than going across the public internet. >> When you started, it took some time obviously to build out that infrastructure and then the Cloud came into play, but it was still early days, but it sounds like you're taking the AWS Cloud model and applying that to your business, eliminate all that undifferentiated heavy lifting, if you will, like the visioning in management. >> Yeah, we've heard many people, and that's kind of the impetus of this was, I want to be directly connected to my end point. And how do I do that? AWS, yes, they had direct connect, but figuring out how to do that as an enterprise was challenging. So we said, hey, we'll automate that for you. Just tell us what region you want to connect to. And we'll do all the heavy lifting and we'll just hand you back a villain tag. You're good to go. So it's a classic case, okay. AWS has direct connect. People will go, oh, that's directly competitive, but it's now you're adding value on top of that. Right? >> Yeah. >> Describe where you fit, Garrett, inside of the AWS ecosystem. You look around this hall and it's just a huge growing ecosystem, where you fit inside of that ecosystem and then your ecosystem. What's that like? >> Where we fit into the AWS ecosystem, as Jay alluded to, we're adding value to our partners and customers where they can come in, not only are they able to access the AWS platform as well as other Cloud platforms, but they're also able to access each other. We have a marketplace in our platform, which allows our customers and partners to put a description of their services on the marketplace and advertise their capabilities out to the rest of the ecosystem of PCCW Global and Console Connect. >> And you're doing that inside of AWS, is that right or at least in part? >> No, that's not inside of AWS. >> So your platform is your platform. >> Yes. >> Your relationship with AWS is to superpower direct connect. Is that right or? >> So we're directly connected to AWS throughout the globe. And this allows our customers and partners to be able to utilize not only the PCCW global network, but also to expand that capability to the AWS platform in Cloud. >> So wherever there's a Cloud, you plug into it, okay? >> Garrett: That's correct. >> Jay: Yeah. And then another advantage, the customer, obviously doesn't have to be directly co-located with AWS. They don't have to be in the same geographical region. If for some reason you need to be connected to U.S. west, but you're in Frankfurt, fine, we'll back all the traffic for you. >> Dave: Does that happen a lot? >> It actually does. >> How come? What's the use case there. >> Global diversity is certainly one of them just being able to have multiple footprints. But the other thing that we're seeing more of late is these Cloud-based companies are beginning to be attracted to where their customers are located. So they'll start seeing these packets of views and they'll go, well, we're going to go into that region as well, stand up a VPC there. We want our customers then being able to directly connect to that asset that's closest to them. And then still be able to back call that traffic if necessary or take it wherever. >> What's the big macro trends in your business? Broadly you see cost per bit coming down, you see data consumption and usage going through the roof. How does that affect you? What are some of the big trends that you see? >> I think one of the biggest ones and one that we targeted with Console Connect, we were hearing a lot of customers going, the world's changing so dynamically. We don't know how to do a one-year forecast of bandwidth, much less a three-year, which is what a lot of contracts are asking us for. So we said, hey, how about one day? Can you do one day? (Dave laughs) Because that's what our granularity is. We allow for anything from one day up to three years right now, and then even within that term, we're dynamic. If something happens, if suddenly some product goes through the roof and you've suddenly got a spike in traffic, if a ship drags its anchor through a sub sea cable, and suddenly you're having to pivot, you just come into the platform, you click a couple of buttons, 20 seconds later, we've modified your bandwidth for you or we've provisioned a new circuit for you, we've got your backup going, whatever. Really at the end of the day, it's the customer paying for their network, so the customer should be the one making those decisions. >> How's that affect pricing? I presume or so, I can have one day to a three-year term, for example if I commit to three years, I get a better deal. Is that right, or? >> You do, but at the end of the day, it's actually pretty much a moderate, a better deal. We don't want to force the hand of the customer. If you signed a 12 month contract with us, we're going to give you a 3% discount. >> So it's not really, that's not a motivation to do it. It's just (indistinct) reduce the transaction complexity. And that's why you will sign up for a longer term not to get the big discount. >> Correct. And then, like I said, even within a longer contract, we're still going to allow you to flex and flow and modify if you need to, because it's your network. >> What kind of constraints do you put on that? Do I have to commit to a flow? And then everything above that is, I can flex up. Is that how it works? >> Yeah. >> Okay. And then, the more I commit to, the better the deal is, or not necessarily? >> No, it's pretty much flat rate. >> Okay, I'm going to commit and I'm going to say, all right, I know I'm going to use X, or sign up for that and anything over it, you're pretty flexible, I might get a few points if I sign up for more, somebody might want to optimize that if they're big enough. >> And another really neat advantage, the other complaint we heard from customers, they go, I need three different direct connect, I need to be connected to three different parties, but I don't want to run three different cross-connects and I don't want to have three different ports. That's just an expense and I don't want. And we, fine, take your one gig port run one gig of services on it. If that's 20 different services, we're fine. We allow you to multiplex your port and provision as- >> So awesome. I love that model. I know some software companies who I would recommend to take a look at that pricing model. So Garrett, how do you segment the ecosystem? How do you look at that? Maybe you could draw and paint a picture of the idea of partners and what they look like. I know there's not just one category, but, >> Sure. Our ideal partners are internet exchangers, Cloud partners and SAS providers, because a big piece of our business is migration to the Cloud, and the flexibility of our platform allows and encourages our SAS providers and SI partners to perform migration to the Cloud much easier in a flexible format for their customers. >> What can you tell us, any kind of metrics you can give us around your business to give a sense of the scope, the scale? >> Well, of our business, (Dave laughs) one of the driving factors here, Gardner says that about 2023, I think, 40% of the enterprise workloads will be deployed in the Cloud, which is all fine and dandy, except in my head, you're just trading one set of complexities for another. Instead of having everything in a glass house and being able to understand that, now you're going, it's in the Cloud, now I need to manage my connectivity there. wait a minute, are my security policies still the same? Do they apply if I'm going across the public internet? What exposure have I just bought into myself to try to run this? The platform really aims at normalizing that as much as possible. If you're directly connected to AWS, at the end of the day, that's a really long ethernet cable. So your a glass house just got a lot bigger, but you're still able to maintain and use the exact same policies and procedures that you've been using. That's really one of our guiding principles, is to reduce that complexity and make it very simple for the user. >> I understand that, cause in the early days of Cloud, a lot of enterprises, the CIOs, they were concerned about security, then I think they realized, ah, AWS has pretty good security. CIA is using it. But still people would say to me, it's not that it's best security, it's just different. You know, we move slow, Dave. How do you accommodate, there's that diversity, I mean, AWS is obviously matured, but are you suggesting that you can take my security edicts in my glass house and bring those into your networks and ultimately into the Cloud? Is that how it works? >> That's the goal. It's not going to be a panacea more than likely, but the more edicts that we can allow you to bring across and not have to go back and revamp and, the better for you as a customer and the better really for us, because it normalizes things, it makes it much easier for us to accommodate more and more users. >> And is it such now in the eco, is all the diversity in the ecosystem, is it such that there's enough common patterns you guys can accommodate most of those use cases? >> Yeah, absolutely. One of the key components is the fact that the platform runs on our MPLS network, which is inherently secure. It's not on the public internet anywhere. We do have internet on demand capability. So in the event that a customer wants access to the internet, no problem. We can accommodate this. And we also have 5G capability built into the platform to allow flexibility of location and flexibility of, I would say, standing up new customer locations. And then the other component of the security is the fact that the customers can bring their own security and apply anywhere. We're not blocking, we don't have any port filters or anything of this nature. >> If would think 5G actually, I could see people arguing both sides, but my sense is 5G is going to be a huge driver for your business cause it's going to just create so much more demand for your services, I think. I can see somebody arguing the counter about it. What's your point of view on that? >> No, I think that's a fair assessment. I think it's going to drive business for everyone here on the show floor and it's pushing those workloads more toward the edge, which is not an area that people were typically concerned with. The edge was just the door that they walked through. That's becoming much different now. We're also going to start seeing, and we're already seeing it, huge trends of moving that data at the edge rather than bringing it all the way back to a central warehouse and help ending it. The ability to have a dynamic platform where you can see exactly what your network's doing and in the push of a button, modify that, or provision new connectivity in response to how your business is performing. >> Yeah, ultimately it's all about the applications that are going to be driving demand for more data. That's just a tailwind for you guys. >> Yeah. You look at, some of the car companies are coming on, Tesla, you're drive around with like eight CPUs and I think communicating back over the air. >> Dave: Yeah, right. >> You start scaling that and you start getting into some some real bottlenecks. >> Amazing business you guys having obviously capital intensive, but once you get in there, you got a big moat. That is a matter of getting on a flywheel and innovating. Guys, congratulations on all the progress and so much for coming on theCUBE. >> Thanks for the time. >> Thank you very much. >> Great to meet you guys. Good luck. All right, thank you for watching. This is Dave Vellante for theCUBE, the leader in High-Tech Coverage. We'll be right back. (upbeat music)

(upbeat music) >> Welcome back to Las Vegas everybody. You are watching theCube's coverage of AWS reinvent 2021. I'll tell you this place is packed. It's quite amazing here over 20,000 people, I'd say it's closer to 25, maybe 27,000. And there's a little overflow, lots going on in the evenings. It's quite remarkable. And we're really happy to be part of this. Jay Turner is here, he's the vice president of development and ops at PCCW Global. He's joined by Garrett Lowell, vice-president of ecosystem partnerships for the Americas at PCCW Global. Guys, welcome to theCube. Thanks for coming on. >> Thank you so much. >> So, Jay, maybe you could take us through for those people who aren't familiar with your company, what do you guys do? What do you all about? >> Yes, so PCCW Global is the international operating wing of Hong Kong Telecom. So if it's outside of Hong Kong, it's our network. We've got about 695,000 kilometers of diverse cable. We've got about 43, 44 terabit of capacity. Came into business in 2005 if my brain is serving me correctly right now. So we have a very diverse and vast portfolio ranging all the way from satellite teleports, all the way to IP transit. We're a tier one service provider from that perspective as well. So we do one of everything when it comes to networking and that's really what was the basis of Console Connect, was inventing a platform to really enable our users to capitalize on that our network and our assets. >> Okay, so 2005, obviously you predated cloud, you laid a bunch of fibers, it's getting in the ocean, I mean, global networks, I mean, there was a big trend to do that and you had to think, you had to go bigger or go home and that business. >> Jay: Yes you had to do. >> So and Console Connect is your platform, is that right? So explain. >> Yeah, sorry. Yeah, Console Connect is our software defined interconnection platform. So we built a user self-service portal. Users can allocate ports, they get the LOAs issue to them directly from the platform. And then once they've got an active port or they've come in via one of our partnerships, they can then provision connectivity across our platform. And that may be extending to their data centers or extending or their branch office, or it could be building a circuit into the cloud via direct connect, could be building a circuit into an internet exchange. And all of those circuits are going to be across that 685,000 kilometers of diverse fiber rather than going across the public internet. >> So, when you started, it took some time obviously to build out that infrastructure and then the cloud came into play, but it was still early days, but it sounds like you're taking the cloud model, AWS Cloud model and applying that to your business, eliminate all that undifferentiated, heavy lifting, if you will, the visioning and management. >> Yeah, we've heard many people and that's kind of the impetus of this was, I want to be directly connected to my end point. And how do I do that? And AWS, yes, they had direct connect, but figuring out how to do that as an enterprise was challenging. So we said, hey, we'll automate that for you. Just tell us what region you want to connect to. And we'll do all the heavy lifting, and we'll just hand you back a villain tag. You're good to go. >> So it's a classic case of, okay, AWS has direct connect, people they go, "Ah, that's directly competitive, but it's not, you're adding value on top of that." Right. So describe where you fit Garrett inside of the AWS ecosystem. You look around this hall and it's just a huge growing ecosystem, where you fit inside of that ecosystem and then your ecosystem, what's that like? >> Okay, so where we fit into the AWS ecosystem, as Jay alluded to, we're adding value to our partners and customers where they can come in, not only are they able to access the AWS platform as well as other cloud platforms, but they're also able to access each other. So we have a marketplace in our platform, which allows our customers and partners to put a description of their services on the marketplace and advertise their capabilities out to the rest of the ecosystem of PCCW Global and Console Connect. >> Okay, so and you're doing that inside of AWS? I that right? Or at least in part? >> No, that's not inside of AWS. >> Okay, so your platform is your platform. >> Yes. >> And then, so your relationship with AWS is to sort of superpower direct connect, is that right or? >> So we're directly connected to AWS throughout the globe. And this allows our customers and partners to be able to utilize not only the PCCW Global network, but also to expand that capability to the AWS platform in clouds. >> Wherever there's a cloud you plug into it? Okay. >> That's correct. >> And then another advantage there is the customer, obviously doesn't have to be directly co-located with AWS. They don't have to be in the same geographic region. If for some reason you need to be connected to US West, but you're in Frankfurt, fine, we'll back all the traffic for you. >> Does that happen a lot? >> It actually does. >> How come? Why, what's the use case there? >> Global diversity is certainly one of them, just being able to have multiple footprints. But the other thing that we're seeing more of late is these cloud-based companies are beginning to kind of be attracted to where their customers are located. So they'll start seeing these pockets of use and they'll go, well, okay, we're going to go into that region as well, stand up a VPC there. And so then we want to our customers then being able to directly connect to that asset, that's closest to them. And then still be able to back call that traffic if necessary or take it wherever. >> What are the big, sort of macro trends in your business? I mean, broadly you see cost per bit coming down, you see data consumption and usage going through the roof. How does that affect you? What are some of the big trends that you see? >> I think one of the biggest ones and one that we targeted with Console Connect, we were hearing a lot of customers going, the world's changing so dynamically. We don't know how to do a one-year forecast of bandwidth, much less a three-year, which is what a lot of contracts are asking us for. So we said, hey, how about one day? Can you do one day? (Dave laughing) Because that's what our granularity is. So we allow for anything from one day up to three years right now, and then even within that term, we're dynamic. So if something happens, suddenly some product goes through the roof and you've suddenly got a spike in traffic. If a ship drags its anchor through a sub sea cable, and suddenly you're having to pivot, you just come into the platform, you click a couple of buttons, 20 seconds later, we've modified your bandwidth for you, or we've provisioned a new circuit for you. We've got your backup going whatever. Really at the end of the day, it's the customer paying for their network, so the customer should be the one making those decisions. >> How's that affect pricing? I presume, so I can have one date or a three-year term. Presume if I commit to three years, I get a better deal, is that right or? >> You do, but I mean, at the end of the day, it's actually pretty much a moderate, a better deal. We don't want to force the hand of the customer. So yeah, if you signed a 12 month contract with us, we're going to give you a 3% discount. >> Yeah, so it's not really, that's not a motivation to do it. Is just you want to reduce the transaction complexity. And that's why you would sign up for a longer term not to get the big discount. >> Correct. And then, like I said, even within a longer contract, we're still going to allow you to flex and flow and modify if you need to because it's your network. >> What kind of constraints do you put on that? Do I have to commit to a floor and then everything above that is I can flex up? Is that how it works? Okay. And then the more I commit to the better the deal is, or not necessarily? >> No, it's pretty much flat, right. >> So, okay. So I'm going to come in and I'm going to say, all right, I know I'm going to use X, I'll sign up for that and anything over it. You're pretty flexible, I might get a few points if I sign up for more, somebody might want to optimize that if they're big enough. >> And another really neat advantage, and the other complaint we heard from customers, they go, I need three different direct connect, or I need to be connected to three different parties, but I don't want to run three different cross-connects and I don't want to have three different ports. That's just an expense I don't want. And we say, fine, take your one gig port, run one gig of services on it, if that's 20 different services, we're fine. So we allow you to multiplex your port and provision- >> It's awesome. I love that model. I know some software companies who I would recommend take a look at that pricing model. So, Garrett, how do you segment the ecosystem? How do you look at that way? Maybe you could draw paint a picture sort of the, the ideal partners and what they look like. I know there's not just one category, but. >> Sure, so our ideal partners are internet exchanges, cloud partners, and SAS providers, because a big piece of our business is migration to the cloud. And the flexibility of our platform allows and encourages our SAS providers and SI partners to perform migration to the cloud much easier and flexible in a flexible format for their customers. >> Yeah, so what can you tell us, any kind of metrics you can give us around your business to give a sense of the the scope, the scale. >> Well, of our business, kind of one of the driving factors here, Gardner says that about 2023, I think 40% of the enterprise workloads will be deployed in the cloud, which is all fine and dandy, except in my head, you're just trading one set of complexities for another. So now, instead of having everything in a glass house and being able to kind of understand that now you're going, well, okay, so it's in the cloud now I need to manage my connectivity there. And, oh, well, wait a minute, are my security policies still the same? Do they apply if I'm going across the public internet? What exposure have I just, bought into myself to try to run this? So the platform really aims at normalizing that as much as possible. If you're directly connected to AWS, at the end of the day, that's a really long ethernet cable. So you're a glass house just got a lot bigger, but you're still able to maintain and use the exact same policies and procedures that you've been using. So that's really one of our guiding principles is to reduce that complexity and make it very simple for the user. >> Well, I don't understand, 'cause in the early days of cloud, a lot of enterprises, CIO they were concerned about security. And I think they realized that AWS has pretty good security, well, CIA is using it. But still people would say to me, it's not that it's bad security, it's just different. We move slow, Dave. So how do you accommodate, now I don't know, does that diversity, I mean, AWS has obviously matured, but are you suggesting that you can take my security edicts in my glass house and bring those into your networks and ultimately into the cloud? Is that kind of how it works? >> That's the goal. It's not going to be a panacea more than likely, but the more edicts that we can allow you to bring across and not have to go back and revamp and the better for you as a customer and the better really for us, because it normalizes things, it makes it much easier for us to accommodate more and more users. >> It is such now in the eco, it was all the diversity in the ecosystem. Is it such that there's enough common patterns that you you guys can kind of accommodate most of those use cases? >> Yeah, absolutely. I think the, one of the key components is the fact that the platform runs on our MPLS network, which is inherently secure. It's not on the public internet anywhere. Now we do have internet on demand capability. So in the event that a customer wants access to the internet, no problem, we can accommodate this. And we also have 5G capability built into the platform to allow flexibility of location and flexibility of... I would say, standing up new customer locations. And then the other component of the security is the fact that the customers can bring their own security and apply anywhere. So we're not blocking, we don't have any port filters or anything of this nature. >> Well, I would think 5G actually, I mean, I could see people arguing both sides, but my sense is 5G is going to be a huge driver for your business, 'cause it's going to just create so much more demand for your services I think, I could see somebody arguing the counter, but what's your point of view on that? >> No. I think that's a fair assessment. I think it's going to drive business for everyone here on the show floor. And it's pushing those workloads more toward the edge, which is not an area that people were typically concerned with. The edge was just the door that they walked through. That's becoming much different now. And we're also going to start seeing, and we're already seeing it, huge trends of moving that data at the edge, rather than bringing it all the way back to a central warehouse in Hare pending it. So, again, the ability to have a dynamic platform where you can see exactly what your network's doing and in the push of a button, modify that, or provision new connectivity in response to how your business is performing. >> Yeah, and ultimately it's all about the applications that are going to be driving demand for more data. And that's just a tailwind for you guys. >> Yeah, yeah and then you look at some of the car companies are coming on, you know, Tesla, you're driving around with like eight CPU's in that thing, communicating back over the air. >> Dave: Yeah right. >> You start scaling that, and you start getting into some real bottleneck. >> Amazing business you guys having, obviously capital intensive, but once you get in there, you've got a big moat, and then it's a matter of getting on a flywheel and innovating. Guys, congratulations on all the progress and thanks so much for coming on theCube. >> Yeah. No, thanks for the time. >> Thank you very much. >> Yeah, great to meet you guys. Good luck. All right. Thank you for watching. This is Dave Vellante for theCube, the leader in high-tech coverage, right back. (upbeat music)

from the cube studios in palo alto in boston bringing you data driven insights from the cube and etr this is breaking analysis with dave vellante semiconductors are the heart of technology innovation for decades technology improvements have marched the cadence of silicon advancements in performance cost power and packaging in the past 10 years the dynamics of the semiconductor industry have changed dramatically soaring factory costs device volume explosions fabulous chip companies greater programmability compressed time to tape out a lot more software content the looming presence of china these and other factors have changed the power structure of the semiconductor business chips today power every aspect of our lives and have led to a global semiconductor shortage that's been well covered but we've never seen anything like it before we believe silicon's success in the next 20 years will be determined by volume manufacturing capabilities design innovation public policy geopolitical dynamics visionary leadership and innovative business models that can survive the intense competition in one of the most challenging businesses in the world hello and welcome to this week's wikibon cube insights powered by etr in this breaking analysis it's our pleasure to welcome daniel newman in one of the leading analysts in the technology business and founder of futurum research daniel welcome to the program thanks so much dave great to see you thanks for having me big topic yeah i'll say i'm really looking forward to this and so here's some of the topics that we want to cover today if we have time changes in the semiconductor industry i've said they've been dramatic the shift to nofap companies we're going to talk about volume manufacturing those shifts that have occurred largely due to the arm model we want to cover intel and dig into that and what it has to do to to survive and thrive these changes and then we want to take a look at how alternative processors are impacting the world people talk about is moore's law dead is it alive and well daniel you have strong perspectives on all of this including nvidia love to get your thoughts on on that plus talk about the looming china threat as i mentioned in in the intro but daniel before we get into it do these topics they sound okay how do you see the state of the semiconductor industry today where have we come from where are we and where are we going at the macro level there are a lot of different narratives that are streaming alongside and they're not running in parallel so much as they're running and converging towards one another but it gradually different uh you know degrees so the last two years has welcomed a semiconductor conversation that we really hadn't had and that was supply chain driven the covid19 pandemic brought pretty much unprecedented desire demand thirst or products that are powered by semiconductors and it wasn't until we started running out of laptops of vehicles of servers that the whole world kind of put the semiconductor in focus again like it was just one of those things dave that we as a society it's sort of taken for granted like if you need a laptop you go buy a laptop if you needed a vehicle there'd always be one on the lot um but as we've seen kind of this exponentialism that's taken place throughout the pandemic what we ended up realizing is that semiconductors are eating the world and in fact the next industrial the entire industrial itself the complex is powered by semiconductor technology so everything we we do and we want to do right you went from a vehicle that might have had 50 or 100 worth of semiconductors on a few different parts to one that might have 700 800 different chips in it thousands of dollars worth of semi of semiconductors so you know across the board though yes you're dealing with the dynamics of the shortage you're dealing with the dynamics of innovation you're dealing with moore's law and sort of coming to the end which is leading to new process we're dealing with the foundry versus fab versus invention and product development uh situation so there's so many different concurrent semiconductor narratives that are going on dave and we can talk about any of them and all of them and i'm sure as we do we'll overlap all these different themes you know maybe you can solve this mystery for me there's this this this chip shortage and you can't invent vehicle inventory is so tight but yet when you listen to uh the the ads if the the auto manufacturers are pounding the advertising maybe they're afraid of tesla they don't want to lose their brand awareness but anyway so listen it's by the way a background i want to get a little bit academic here but but bear with me i want to introduce actually reintroduce the concept of wright's law to our audience we know we all know about moore's law but the earlier instantiation actually comes from theodore wright t.p wright he was this engineer in the airplane industry and the math is a little bit abstract to apply but roughly translated says as the cumulative number of units produced doubles your cost per unit declines by a fixed percentage now in airplanes that was around 15 percent in semiconductors we think that numbers more like 20 25 when you add the performance improvements you get from silicon advancements it translates into something like 33 percent cost cost declines when you can double your cumulative volume so that's very important because it confers strategic advantage to the company with the largest volume so it's a learning curve dynamic and it's like andy jassy says daniel there's no compression algorithm for experience and it definitely applies here so if you apply wright's law to what's happening in the industry today we think we can get a better understanding of for instance why tsmc is dominating and why intel is struggling any quick thoughts on that well you have to take every formula like that in any sort of standard mathematics and kind of throw it out the window when you're dealing with the economic situation we are right now i'm not i'm not actually throwing it out the window but what i'm saying is that when supply and demand get out of whack some of those laws become a little bit um more difficult to sustain over the long term what i will say about that is we have certainly seen this found um this fabulous model explode over the last few years you're seeing companies that can focus on software frameworks and innovation that aren't necessarily getting caught up in dealing with the large capital expenditures and overhead the ability to as you suggested in the topics here partner with a company like arm that's developing innovation and then and then um you know offering it uh to everybody right and for a licensee and then they can quickly build we're seeing what that's doing with companies like aws that are saying we're going to just build it alibaba we're just going to build it these aren't chip makers these aren't companies that were even considered chip makers they are now today competing as chip makers so there's a lot of different dynamics going back to your comment about wright's law like i said as we normalize and we figure out this situation on a global scale um i do believe that the who can manufacture the most will certainly continue to have significant competitive advantages yeah no so that's a really interesting point that you're bringing up because one of the things that it leads me to think is that the chip shortage could actually benefit intel i think will benefit intel so i want to introduce this some other data and then get your thoughts on this very simply the chart on the left shows pc shipments which peaked in in 2011 and then began at steady decline until covid and they've the pcs as we know have popped up in terms of volume in the past year and looks like they'll be up again this year the chart on the right is cumulative arm shipments and so as we've reported we think arm wafer volumes are 10x those of x86 volumes and and as such the arm ecosystem has far better cost structure than intel and that's why pat gelsinger was called in to sort of save the day so so daniel i just kind of again opened up this this can of worms but i think you're saying long term volume is going to be critical that's going to confer low cost advantages but in the in in the near to mid-term intel could actually benefit from uh from this chip shortage well intel is the opportunity to position itself as a leader in solving the repatriation crisis uh this will kind of carry over when we talk more about china and taiwan and that relationship and what's going on there we've really identified a massive gap in our uh in america supply chain in the global supply chain because we went from i don't have the stat off hand but i have a rough number dave and we can validate this later but i think it was in like the 30-ish high 30ish percentile of manufacturing of chips were done here in the united states around 1990 and now we're sub 10 as of 2020. so we we offshored almost all of our production and so when we hit this crisis and we needed more manufacturing volume we didn't have it ready part of the problem is you get people like elon musk that come out and make comments to the media like oh it'll be fixed later this year well you can't build a fab in a year you can't build a fab and start producing volume and the other problem is not all chips are the same so not every fab can produce every chip and when you do have fabs that are capable of producing multiple chips it costs millions of dollars to change the hardware and to actually change the process so it's not like oh we're going to build 28 today because that's what ford needs to get all those f-150s out of the lot and tomorrow we're going to pump out more sevens for you know a bunch of hp pcs it's a major overhaul every time you want to retool so there's a lot of complexity here but intel is the one domestic company us-based that has basically raised its hand and said we're going to put major dollars into this and by the way dave the arm chart you showed me could have a very big implication as to why intel wants to do that yeah so right because that's that's a big part of of foundry right is is get those volumes up so i want to hold that thought because i just want to introduce one more data point because one of the things we often talk about is the way in which alternative processors have exploded onto the scene and this chart here if you could bring that up patrick thank you shows the way in which i think you're pointing out intel is responding uh by leveraging alternative fat but once again you know kind of getting getting serious about manufacturing chips what the chart shows is the performance curve it's on a log scale for in the blue line is x86 and the orange line is apple's a series and we're using that as a proxy for sort of the curve that arm is on and it's in its performance over time culminating in the a15 and it measures trillions of operations per second so if you take the traditional x86 curve of doubling every 18 to 24 months that comes out roughly to about 40 percent improvement per year in performance and that's diminishing as we all know to around 30 percent a year because the moore's law is waning the orange line is powered by arm and it's growing at over a hundred percent really 110 per year when you do the math and that's when you combine the cpu the the the neural processing unit the the the xpu the dsps the accelerators et cetera so we're seeing apple use arm aws to you to your point is building chips on on graviton and and and tesla's using our list is long and this is one reason why so daniel this curve is it feels like it's the new performance curve in the industry yeah we are certainly in an era where companies are able to take control of the innovation curve using the development using the open ecosystem of arm having more direct control and price control and of course part of that massive arm number has to do with you know mobile devices and iot and devices that have huge scale but at the same time a lot of companies have made the decision either to move some portion of their product development on arm or to move entirely on arm part of why it was so attractive to nvidia part of the reason that it's under so much scrutiny that that deal um whether that deal will end up getting completed dave but we are seeing an era where we want we i said lust for power i talked about lust for semiconductors our lust for our technology to do more uh whether that's software-defined vehicles whether that's the smartphones we keep in our pocket or the desktop computer we use we want these machines to be as powerful and fast and responsive and scalable as possible if you can get 100 where you can get 30 improvement with each year and generation what is the consumer going to want so i think companies are as normal following the demand of consumers and what's available and at the same time there's some economic benefits they're they're able to realize as well i i don't want to i don't want to go too deep into nvidia arm but what do you handicap that that the chances that that acquisition actually happens oh boy um right now there's a lot of reasons it should happen but there are some reasons that it shouldn't i still kind of consider it a coin toss at this point because fundamentally speaking um you know it should create more competition but there are some people out there that believe it could cause less and so i think this is going to be hung up with regulators a little bit longer than we thought we've already sort of had some previews into that dave with the extensions and some of the timelines that have already been given um i know that was a safe answer and i will take credit for being safe this one's going to be a hard one to call but it certainly makes nvidia an amazing uh it gives amazing prospects to nvidia if they're able to get this deal done yeah i i agree with you i think it's 50 50. okay my i want to pose the question is intel too strategic to fail in march of this year we published this article where we posed that question uh you and i both know pat pretty well we talked about at the time the multi-front war intel is waging in a war with amd the arm ecosystem tsmc the design firms china and we looked at the company's moves which seemed to be right from a strategy standpoint the looking at the potential impact of the u.s government intel's partnership with ibm and what that might portend the us government has a huge incentive to make sure intel wins with onshore manufacturing and that looming threat from china but daniel is intel too strategic to fail and is pat gelsinger making the right moves well first of all i do believe at this current juncture where the semiconductor and supply chain shortage and crisis still looms that intel is too strategic to fail i also believe that intel's demise is somewhat overstated not to say intel doesn't have a slate of challenges that it's going to need to address long term just with the technology adoption curve that you showed being one of them dave but you have to remember the company still has nearly 90 of the server cpu market it still has a significant market share in client and pc it is seeing market share erosion but it's not happened nearly as fast as some people had suggested it would happen with right now with the demand in place and as high as it is intel is selling chips just about as quickly as it can make them and so we right now are sort of seeing the tam as a whole the demand as a whole continue to expand and so intel is fulfilling that need but where are they really too strategic to fail i mean we've seen in certain markets in certain uh process in um you know client for instance where amd has gained of course that's still x86 we've seen uh where the m1 was kind of initially thought to be potentially a pro product that would take some time it didn't take nearly as long for them to get that product in good shape um but the foundry and fab side is where i think intel really has a chance to flourish right now one it can play in the arm space it can build these facilities to be able to produce and help support the production of volumes of chips using arm designs so that actually gives intel and inroads two is it's the company that has made the most outspoken commitment to invest in the manufacturing needs of the united states both here in the united states and in other places across the world where we have friendly ally relationships and need more production capabilities if not in intel b and there is no other logical company that's us-based that's going to meet the regulator and policymakers requirements right now that is also raising their hand and saying we have the know-how we've been doing this we can do more of this and so i think pat is leaning into the right area and i think what will happen is very likely intel will support manufacturing of chips by companies like qualcomm companies like nvidia and if they're able to do that some of the market share losses that they're potentially facing with innovation challenges um and engineering challenges could be offset with growth in their fab and foundry businesses and i think i think pat identified it i think he's going to market with it and you know convincing the street that's going to be a whole nother thing that this is exciting um but i think as the street sees the opportunity here this is an area that intel can really lean into so i think i i think people generally would recognize at least the folks i talk to and it'll be interested in your thoughts who really know this business that intel you know had the best manufacturing process in in the world obviously that's coming to question but but but but for instance people say well intel's 10 nanometer you know is comparable to tsm seven nanometer and that's sort of overstated their their nanometer you know loss but but so so they they were able to point as they were able to sort of hide some of the issues maybe in design with great process and and i i believe that comes down to volume so the question i have then is and i think so i think patrick's pat is doing the right thing because he's going after volume and that's what foundry brings but can he get enough volume or does he need for inst for instance i mean one of the theories i've put out there is that apple could could save the day for intel if the if the us government gets apple in a headlock and says hey we'll back off on break up big tech but you got to give pat some of your foundry volume that puts him on a steeper learning curve do you do you worry sometimes though daniel that intel just even with like qualcomm and broadcom who by the way are competitors of theirs and don't necessarily love them but even even so if they could get that those wins that they still won't have the volume to compete on a cost basis or do you feel like even if they're numbered a number three even behind samsung it's good enough what are your thoughts on that well i don't believe a company like intel goes into a business full steam and they're not new to this business but the obvious volume and expansion that they're looking at with the intention of being number two or three these great companies and you know that's same thing i always say with google cloud google's not out to be the third cloud they're out to be one well that's intel will want to to be stronger if the us government and these investments that it's looking at making this 50 plus billion dollars is looking to pour into this particular space which i don't think is actually enough but if if the government makes these commitments and intel being likely one of the recipients of at least some of these dollars to help expedite this process move forward with building these facilities to make increased manufacturing very likely there's going to be some precedent of law a policy that is going to be put in place to make sure that a certain amount of the volume is done here stateside with companies this is a strategic imperative this is a government strategic imperative this is a putting the country at risk of losing its technology leadership if we cannot manufacture and control this process of innovation so i think intel is going to have that as a benefit that the government is going to most likely require some of this manufacturing to take place here um especially if this investment is made the last thing they're going to want to do is build a bunch of foundries and build a bunch of fabs and end up having them not at capacity especially when the world has seen how much of the manufacturing is now being done in taiwan so i think we're concluding and i i i correctly if i'm wrong but intel is too strategic to fail and and i i sometimes worry they can go bankrupt you know trying to compete with the likes of tsmc and that's why the the the public policy and the in the in the partnership with the u.s government and the eu is i think so important yeah i don't think bankruptcy is an immediate issue i think um but while i follow your train of thought dave i think what you're really looking at more is can the company grow and continue to get support where i worry about is shareholders getting exhausted with intel's the merry-go-round of not growing fast enough not gaining market share not being clearly identified as a leader in any particular process or technology and sort of just playing the role of the incumbent and they the company needs to whether it's in ai whether it's at the edge whether it's in the communications and service provider space intel is doing well you look at their quarterly numbers they're making money but if you had to say where are they leading right now what what which thing is intel really winning uh consistently at you know you look at like ai and ml and people will point to nvidia you look at you know innovation for um client you know and even amd has been super disruptive and difficult for intel uh of course you we've already talked about in like mobile um how impactful arm has been and arm is also playing a pretty big role in servers so like i said the market share and the technology leadership are a little out of skew right now and i think that's where pat's really working hard is identifying the opportunities for for intel to play market leader and technology leader again and for the market to clearly say yes um fab and foundry you know could this be an area where intel becomes the clear leader domestically and i think that the answer is definitely yes because none of the big chipmakers in the us are are doing fabrication you know they're they're all outsourcing it to overseas so if intel can really lead that here grow that large here then it takes some of the pressure off of the process and the innovation side and that's not to say that intel won't have to keep moving there but it does augment the revenue creates a new profit center and makes the company even more strategic here domestically yeah and global foundry tapped out of of sub 10 nanometer and that's why ibm's pseudonym hey wait a minute you had a commitment there the concern i have and this is where again your point is i think really important with the chip shortage you know to go from you know initial design to tape out took tesla and apple you know sub sub 24 months you know probably 18 months with intel we're on a three-year design to tape out cycle maybe even four years so they've got to compress that but that as you well know that's a really hard thing to do but the chip shortage is buying them time and i think that's a really important point that you brought out early in this segment so but the other big question daniel i want to test with you is well you mentioned this about seeing arm in the enterprise not a lot of people talk about that or have visibility on that but i think you're right on so will arm and nvidia be able to seriously penetrate the enterprise the server business in particular clearly jensen wants to be there now this data from etr lays out many of the enterprise players and we've superimposed the semiconductor giants in logos the data is an xy chart it shows net score that's etr's measure of spending momentum on the vertical axis and market share on the horizontal axis market share is not like idc market share its presence in the data set and as we reported before aws is leading the charge in enterprise architecture as daniel mentioned they're they're designing their own chips nitro and graviton microsoft is following suit as is google vmware has project monterey cisco is on the chart dell hp ibm with red hat are also shown and we've superimposed intel nvidia china and arm and now we can debate the position of the logos but we know that one intel has a dominant position in the data center it's got to protect that business it cannot lose ground as it has in pcs because the margin pressure it would face two we know aws with its annapurna acquisition is trying to control its own destiny three we know vmware has project monterey and is following aws's lead to support these new workloads beyond x86 general purpose they got partnerships with pansando and arm and others and four we know cisco they've got chip design chops as does hpe maybe to a lesser extent and of course we know ibm has excellent semiconductor design expertise especially when it comes to things like memory disaggregation as i said jensen's going hard after the data center you know him well daniel we know china wants to control its own destiny and then there's arm it dominates mobile as you pointed out in iot can it make a play for the data center daniel how do you see this picture and what are your thoughts on the future of enterprise in the context of semiconductor competition it's going to take some time i believe but some of the investments and products that have been brought to market and you mentioned that shorter tape out period that shorter period for innovation whether it's you know the graviton uh you know on aws or the aiml chips that uh with trainium and inferentia how quickly aws was able to you know develop build deploy to market an arm-based solution that is being well received and becoming an increasing component of the services and and uh products that are being offered from aws at this point it's still pretty small and i would i would suggest that nvidia and arm in the spirit of trying to get this deal done probably don't necess don't want the enterprise opportunity to be overly inflated as to how quickly the company's going to be able to play in that space because that would somewhat maybe slow or bring up some caution flags that of the regulators that are that are monitoring this at the same time you could argue that arm offering additional options in competition much like it's doing in client will offer new form factors new designs um new uh you know new skus the oems will be able to create more customized uh hardware offerings that might be able to be unique for certain enterprises industries can put more focus you know we're seeing the disaggregation with dpus and how that technology using arm with what aws is doing with nitro but what what these different companies are doing to use you know semiconductor technology to split out security networking and storage and so you start to see design innovation could become very interesting on the foundation of arm so in time i certainly see momentum right now the thing is is most companies in the enterprise are looking for something that's fairly well baked off the shelf that can meet their needs whether it's sap or whether it's you know running different custom applications that the business is built on top of commerce solutions and so intel meets most of those needs and so arm has made a lot of sense for instance with these cloud scale providers but not necessarily as much sense for enterprises especially those that don't want to necessarily look at refactoring all the workloads but as software becomes simpler as refactoring becomes easier to do between different uh different technologies and processes you start to say well arm could be compelling and you know because the the bottom line is we know this from mobile devices is most of us don't care what the processor is the average person the average data you know they look at many of these companies the same in enterprise it's always mattered um kind of like in the pc world it used to really matter that's where intel inside was born but as we continue to grow up and you see these different processes these different companies nvidia amd intel all seen as very worthy companies with very capable technologies in the data center if they can offer economics if they can offer performance if they can offer faster time to value people will look at them so i'd say in time dave the answer is arm will certainly become more and more competitive in the data center like it was able to do at the edge in immobile yeah one of the things that we've talked about is that you know the software-defined data center is awesome but it also created a lot of wasted overhead in terms of offloading storage and and networking security and that much of that is being done with general purpose x86 processors which are more expensive than than for instance using um if you look at what as you mentioned great summary of what aws is doing with graviton and trainium and other other tooling what ampere is doing um in in in oracle and you're seeing both of those companies for example particularly aws get isvs to write so they can run general purpose applications on um on arm-based processors as well it sets up well for ai inferencing at the edge which we know arms dominating the edge we see all these new types of workloads coming into the data center if you look at what companies like nebulon and pensando and and others are doing uh you're seeing a lot of their offloads are going to arm they're putting arm in even though they're still using x86 in a lot of cases but but but they're offloading to arm so it seems like they're coming into the back door i understand your point actually about they don't want to overplay their hand there especially during these negotiations but we think that that long term you know it bears watching but intel they have such a strong presence they got a super strong ecosystem and they really have great relationships with a lot of the the enterprise players and they have influence over them so they're going to use that the the the chip shortage benefits them the uh the relationship with the us government pat is spending a lot of time you know working that so it's really going to be interesting to see how this plays out daniel i want to give you the last word your final thoughts on what we talked about today and where you see this all headed i think the world benefits as a whole with more competition and more innovation pressure i like to see more players coming into the fray i think we've seen intel react over the last year under pat gelsinger's leadership we've seen the technology innovation the angstrom era the 20a we're starting to see what that roadmap is going to look like we've certainly seen how companies like nvidia can disrupt come into market and not just using hardware but using software to play a major role but as a whole as innovation continues to take form at scale we all benefit it means more intelligent software-defined vehicles it puts phones in our hands that are more powerful it gives power to you know cities governments and enterprises that can build applications and tools that give us social networks and give us data-driven experiences so i'm very bullish and optimistic on as a whole i said this before i say it again i believe semiconductors will eat the world and then you know you look at the we didn't even really talk about the companies um you know whether it's in ai uh like you know grok or grav core there are some very cool companies building things you've got qualcomm bought nuvia another company that could you know come out of the blue and offer us new innovations in mobile and personal computing i mean there's so many cool companies dave with the scale of data the uh the the growth and demand and desire for connectivity in the world um it's never been a more interesting time to be a fan of technology the only thing i will say as a whole as a society as i hope we can fix this problem because it does create risks the supply chain inflation the economics all that stuff ties together and a lot of people don't see that but if we can't get this manufacturing issue under control we didn't really talk about china dave and i'll just say taiwan and china are very physically close together and the way that china sees taiwan and the way we see taiwan is completely different we have very little control over what can happen we've all seen what's happened with hong kong so there's just so many as i said when i started this conversation we've got all these trains on the track they're all moving but they're not in parallel these tracks are all converging but the convergence isn't perpendicular so sometimes we don't see how all these things interrelate but as a whole it's a very exciting time love being in technology and uh love having the chance to come out here and talk with you i love the optimism and you're right uh that competition in china that's going to come from china as well xi has made it a part of his legacy i think to you know re-incorporate taiwan that's going to be interesting to see i mean taiwan ebbs and flows with regard to you know its leadership sometimes they're more pro i guess i should say less anti-china maybe that's the better way to say it uh and and and you know china's putting in big fab capacity for nand you know maybe maybe people look at that you know some of that is the low end of the market but you know clay christensen would say well to go take a look at the steel industry and see what happened there so so we didn't talk much about china and that was my oversight but but they're after self-sufficiency it's not like they haven't tried before kind of like intel has tried foundry before but i think they're really going for it this time but but now what are your do you believe that china will be able to get self-sufficiency let's say within the next 10 to 15 years with semiconductors yes i would never count china out of anything if they put their mind to it if it's something that they want to put absolute focus on i think um right now china vacillates between wanting to be a good player and a good steward to the world and wanting to completely run its own show the the politicization of what's going on over there we all saw what happened in the real estate market this past week we saw what happened with tech ed over the last few months we've seen what's happened with uh innovation and entrepreneurship it is not entirely clear if china wants to give the more capitalistic and innovation ecosystem a full try but it is certainly shown that it wants to be seen as a world leader over the last few decades it's accomplished that in almost any area that it wants to compete dave i would say if this is one of gigi ping's primary focuses wanting to do this it would be very irresponsible to rule it out as a possibility daniel i gotta tell you i i love collaborating with you um we met face to face just recently and i hope we could do this again i'd love to have you you back on on the program thanks so much for your your time and insights today thanks for having me dave so daniel's website futuram research that's three use in futurum uh check that out for termresearch.com uh the the this individual is really plugged in he's forward thinking and and a great resource at daniel newman uv is his twitter so go follow him for some great stuff and remember these episodes are all available as podcasts wherever you listen all you do is search for breaking analysis podcast we publish each week on wikibon.com and siliconangle.com and by the way daniel thank you for contributing your your quotes to siliconangle the writers there love you uh you can always connect on twitter i'm at divalanto you can email me at david.velante at siliconangle.com appreciate the comments on linkedin and don't forget to check out etr.plus for all the survey data this is dave vellante for the cube insights powered by etr be well and we'll see you next time you

>> From theCUBE studios in Palo Alto, in Boston bringing you data-driven insights from theCUBE and ETR. This is Breaking Analysis with Dave Vellante >> The latest Arm Neoverse announcement further cements our opinion that it's architecture business model and ecosystem execution are defining a new era of computing and leaving Intel in it's dust. We believe the company and its partners have at least a two year lead on Intel and are currently in a far better position to capitalize on a major waves that are driving the technology industry and its innovation. To compete our view is that Intel needs a new strategy. Now, Pat Gelsinger is bringing that but they also need financial support from the US and the EU governments. Pat Gelsinger was just noted as asking or requesting from the EU government $9 billion, sorry, 8 billion euros in financial support. And very importantly, Intel needs a volume for its new Foundry business. And that is where Apple could be a key. Hello, everyone. And welcome to this week's weekly bond Cube insights powered by ETR. In this breaking analysis will explain why Apple could be the key to saving Intel and America's semiconductor industry leadership. We'll also further explore our scenario of the evolution of computing and what will happen to Intel if it can't catch up. Here's a hint it's not pretty. Let's start by looking at some of the key assumptions that we've made that are informing our scenarios. We've pointed out many times that we believe Arm wafer volumes are approaching 10 times those of x86 wafers. This means that manufacturers of Arm chips have a significant cost advantage over Intel. We've covered that extensively, but we repeat it because when we see news reports and analysis and print it's not a factor that anybody's highlighting. And this is probably the most important issue that Intel faces. And it's why we feel that Apple could be Intel's savior. We'll come back to that. We've projected that the chip shortage will last no less than three years, perhaps even longer. As we reported in a recent breaking analysis. Well, Moore's law is waning. The result of Moore's law, I.e the doubling of processor performance every 18 to 24 months is actually accelerating. We've observed and continue to project a quadrupling of performance every two years, breaking historical norms. Arm is attacking the enterprise and the data center. We see hyperscalers as the tip of their entry spear. AWS's graviton chip is the best example. Amazon and other cloud vendors that have engineering and software capabilities are making Arm-based chips capable of running general purpose applications. This is a huge threat to x86. And if Intel doesn't quickly we believe Arm will gain a 50% share of an enterprise semiconductor spend by 2030. We see the definition of Cloud expanding. Cloud is no longer a remote set of services, in the cloud, rather it's expanding to the edge where the edge could be a data center, a data closet, or a true edge device or system. And Arm is by far in our view in the best position to support the new workloads and computing models that are emerging as a result. Finally geopolitical forces are at play here. We believe the U S government will do, or at least should do everything possible to ensure that Intel and the U S chip industry regain its leadership position in the semiconductor business. If they don't the U S and Intel could fade to irrelevance. Let's look at this last point and make some comments on that. Here's a map of the South China sea in a way off in the Pacific we've superimposed a little pie chart. And we asked ourselves if you had a hundred points of strategic value to allocate, how much would you put in the semiconductor manufacturing bucket and how much would go to design? And our conclusion was 50, 50. Now it used to be because of Intel's dominance with x86 and its volume that the United States was number one in both strategic areas. But today that orange slice of the pie is dominated by TSMC. Thanks to Arm volumes. Now we've reported extensively on this and we don't want to dwell on it for too long but on all accounts cost, technology, volume. TSMC is the clear leader here. China's president Xi has a stated goal of unifying Taiwan by China's Centennial in 2049, will this tiny Island nation which dominates a critical part of the strategic semiconductor pie, go the way of Hong Kong and be subsumed into China. Well, military experts say it was very hard for China to take Taiwan by force, without heavy losses and some serious international repercussions. The US's military presence in the Philippines and Okinawa and Guam combined with support from Japan and South Korea would make it even more difficult. And certainly the Taiwanese people you would think would prefer their independence. But Taiwanese leadership, it ebbs and flows between those hardliners who really want to separate and want independence and those that are more sympathetic to China. Could China for example, use cyber warfare to over time control the narrative in Taiwan. Remember if you control the narrative you can control the meme. If you can crawl the meme you control the idea. If you control the idea, you control the belief system. And if you control the belief system you control the population without firing a shot. So is it possible that over the next 25 years China could weaponize propaganda and social media to reach its objectives with Taiwan? Maybe it's a long shot but if you're a senior strategist in the U S government would you want to leave that to chance? We don't think so. Let's park that for now and double click on one of our key findings. And that is the pace of semiconductor performance gains. As we first reported a few weeks ago. Well, Moore's law is moderating the outlook for cheap dense and efficient processing power has never been better. This slideshows two simple log lines. One is the traditional Moore's law curve. That's the one at the bottom. And the other is the current pace of system performance improvement that we're seeing measured in trillions of operations per second. Now, if you calculate the historical annual rate of processor performance improvement that we saw with x86, the math comes out to around 40% improvement per year. Now that rate is slowing. It's now down to around 30% annually. So we're not quite doubling every 24 months anymore with x86 and that's why people say Moore's law is dead. But if you look at the (indistinct) effects of packaging CPU's, GPU's, NPUs accelerators, DSPs and all the alternative processing power you can find in SOC system on chip and eventually system on package it's growing at more than a hundred percent per annum. And this means that the processing power is now quadrupling every 24 months. That's impressive. And the reason we're here is Arm. Arm has redefined the core process of model for a new era of computing. Arm made an announcement last week which really recycle some old content from last September, but it also put forth new proof points on adoption and performance. Arm laid out three components and its announcement. The first was Neoverse version one which is all about extending vector performance. This is critical for high performance computing HPC which at one point you thought that was a niche but it is the AI platform. AI workloads are not a niche. Second Arm announced the Neoverse and two platform based on the recently introduced Arm V9. We talked about that a lot in one of our earlier Breaking Analysis. This is going to performance boost of around 40%. Now the third was, it was called CMN-700 Arm maybe needs to work on some of its names, but Arm said this is the industry's most advanced mesh interconnect. This is the glue for the V1 and the N2 platforms. The importance is it allows for more efficient use and sharing of memory resources across components of the system package. We talked about this extensively in previous episodes the importance of that capability. Now let's share with you this wheel diagram underscores the completeness of the Arm platform. Arms approach is to enable flexibility across an open ecosystem, allowing for value add at many levels. Arm has built the architecture in design and allows an open ecosystem to provide the value added software. Now, very importantly, Arm has created the standards and specifications by which they can with certainty, certify that the Foundry can make the chips to a high quality standard, and importantly that all the applications are going to run properly. In other words, if you design an application, it will work across the ecosystem and maintain backwards compatibility with previous generations, like Intel has done for years but Arm as we'll see next is positioning not only for existing workloads but also the emerging high growth applications. To (indistinct) here's the Arm total available market as we see it, we think the end market spending value of just the chips going into these areas is $600 billion today. And it's going to grow to 1 trillion by 2030. In other words, we're allocating the value of the end market spend in these sectors to the marked up value of the Silicon as a percentage of the total spend. It's enormous. So the big areas are Hyperscale Clouds which we think is around 20% of this TAM and the HPC and AI workloads, which account for about 35% and the Edge will ultimately be the largest of all probably capturing 45%. And these are rough estimates and they'll ebb and flow and there's obviously some overlap but the bottom line is the market is huge and growing very rapidly. And you see that little red highlighted area that's enterprise IT. Traditional IT and that's the x86 market in context. So it's relatively small. What's happening is we're seeing a number of traditional IT vendors, packaging x86 boxes throwing them over the fence and saying, we're going after the Edge. And what they're doing is saying, okay the edge is this aggregation point for all these end point devices. We think the real opportunity at the Edge is for AI inferencing. That, that is where most of the activity and most of the spending is going to be. And we think Arm is going to dominate that market. And this brings up another challenge for Intel. So we've made the point a zillion times that PC volumes peaked in 2011. And we saw that as problematic for Intel for the cost reasons that we've beat into your head. And lo and behold PC volumes, they actually grew last year thanks to COVID and we'll continue to grow it seems for a year or so. Here's some ETR data that underscores that fact. This chart shows the net score. Remember that's spending momentum it's the breakdown for Dell's laptop business. The green means spending is accelerating and the red is decelerating. And the blue line is net score that spending momentum. And the trend is up and to the right now, as we've said this is great news for Dell and HP and Lenovo and Apple for its laptops, all the laptops sellers but it's not necessarily great news for Intel. Why? I mean, it's okay. But what it does is it shifts Intel's product mix toward lower margin, PC chips and it squeezes Intel's gross margins. So the CFO has to explain that margin contraction to wall street. Imagine that the business that got Intel to its monopoly status is growing faster than the high margin server business. And that's pulling margins down. So as we said, Intel is fighting a war on multiple fronts. It's battling AMD in the core x86 business both PCs and servers. It's watching Arm mop up in mobile. It's trying to figure out how to reinvent itself and change its culture to allow more flexibility into its designs. And it's spinning up a Foundry business to compete with TSMC. So it's got to fund all this while at the same time propping up at stock with buybacks Intel last summer announced that it was accelerating it's $10 billion stock buyback program, $10 billion. Buy stock back, or build a Foundry which do you think is more important for the future of Intel and the us semiconductor industry? So Intel, it's got to protect its past while building his future and placating wall street all at the same time. And here's where it gets even more dicey. Intel's got to protect its high-end x86 business. It is the cash cow and funds their operation. Who's Intel's biggest customer Dell, HP, Facebook, Google Amazon? Well, let's just say Amazon is a big customer. Can we agree on that? And we know AWS is biggest revenue generator is EC2. And EC2 was powered by microprocessors made from Intel and others. We found this slide in the Arm Neoverse deck and it caught our attention. The data comes from a data platform called lifter insights. The charts show, the rapid growth of AWS is graviton chips which are they're custom designed chips based on Arm of course. The blue is that graviton and the black vendor A presumably is Intel and the gray is assumed to be AMD. The eye popper is the 2020 pie chart. The instant deployments, nearly 50% are graviton. So if you're Pat Gelsinger, you better be all over AWS. You don't want to lose this customer and you're going to do everything in your power to keep them. But the trend is not your friend in this account. Now the story gets even gnarlier and here's the killer chart. It shows the ISV ecosystem platforms that run on graviton too, because AWS has such good engineering and controls its own stack. It can build Arm-based chips that run software designed to run on general purpose x86 systems. Yes, it's true. The ISV, they got to do some work, but large ISV they have a huge incentives because they want to ride the AWS wave. Certainly the user doesn't know or care but AWS cares because it's driving costs and energy consumption down and performance up. Lower cost, higher performance. Sounds like something Amazon wants to consistently deliver, right? And the ISV portfolio that runs on our base graviton and it's just going to continue to grow. And by the way, it's not just Amazon. It's Alibaba, it's Oracle, it's Marvell. It's 10 cents. The list keeps growing Arm, trotted out a number of names. And I would expect over time it's going to be Facebook and Google and Microsoft. If they're not, are you there? Now the last piece of the Arm architecture story that we want to share is the progress that they're making and compare that to x86. This chart shows how Arm is innovating and let's start with the first line under platform capabilities. Number of cores supported per die or, or system. Now die is what ends up as a chip on a small piece of Silicon. Think of the die as circuit diagram of the chip if you will, and these circuits they're fabricated on wafers using photo lithography. The wafers then cut up into many pieces each one, having a chip. Each of these pieces is the chip. And two chips make up a system. The key here is that Arm is quadrupling the number of cores instead of increasing thread counts. It's giving you cores. Cores are better than threads because threads are shared and cores are independent and much easier to virtualize. This is particularly important in situations where you want to be as efficient as possible sharing massive resources like the Cloud. Now, as you can see in the right hand side of the chart under the orange Arm is dramatically increasing the amount of capabilities compared to previous generations. And one of the other highlights to us is that last line that CCIX and CXL support again Arm maybe needs to name these better. These refer to Arms and memory sharing capabilities within and between processors. This allows CPU's GPU's NPS, et cetera to share resources very often efficiently especially compared to the way x86 works where everything is currently controlled by the x86 processor. CCIX and CXL support on the other hand will allow designers to program the system and share memory wherever they want within the system directly and not have to go through the overhead of a central processor, which owns the memory. So for example, if there's a CPU, GPU, NPU the CPU can say to the GPU, give me your results at a specified location and signal me when you're done. So when the GPU is finished calculating and sending the results, the GPU just signals the operation is complete. Versus having to ping the CPU constantly, which is overhead intensive. Now composability in that chart means the system it's a fixed. Rather you can programmatically change the characteristics of the system on the fly. For example, if the NPU is idle you can allocate more resources to other parts of the system. Now, Intel is doing this too in the future but we think Arm is way ahead. At least by two years this is also huge for Nvidia, which today relies on x86. A major problem for Nvidia has been coherent memory management because the utilization of its GPU is appallingly low and it can't be easily optimized. Last week, Nvidia announced it's intent to provide an AI capability for the data center without x86 I.e using Arm-based processors. So Nvidia another big Intel customer is also moving to Arm. And if it's successful acquiring Arm which is still a long shot this trend is only going to accelerate. But the bottom line is if Intel can't move fast enough to stem the momentum of Arm we believe Arm will capture 50% of the enterprise semiconductor spending by 2030. So how does Intel continue to lead? Well, it's not going to be easy. Remember we said, Intel, can't go it alone. And we posited that the company would have to initiate a joint venture structure. We propose a triumvirate of Intel, IBM with its power of 10 and memory aggregation and memory architecture And Samsung with its volume manufacturing expertise on the premise that it coveted in on US soil presence. Now upon further review we're not sure the Samsung is willing to give up and contribute its IP to this venture. It's put a lot of money and a lot of emphasis on infrastructure in South Korea. And furthermore, we're not convinced that Arvind Krishna who we believe ultimately made the call to Jettisons. Jettison IBM's micro electronics business wants to put his efforts back into manufacturing semi-conductors. So we have this conundrum. Intel is fighting AMD, which is already at seven nanometer. Intel has a fall behind in process manufacturing which is strategically important to the United States it's military and the nation's competitiveness. Intel's behind the curve on cost and architecture and is losing key customers in the most important market segments. And it's way behind on volume. The critical piece of the pie that nobody ever talks about. Intel must become more price and performance competitive with x86 and bring in new composable designs that maintain x86 competitive. And give the ability to allow customers and designers to add and customize GPU's, NPUs, accelerators et cetera. All while launching a successful Foundry business. So we think there's another possibility to this thought exercise. Apple is currently reliant on TSMC and is pushing them hard toward five nanometer, in fact sucking up a lot of that volume and TSMC is maybe not servicing some other customers as well as it's servicing Apple because it's a bit destructive, it is distracted and you have this chip shortage. So Apple because of its size gets the lion's share of the attention but Apple needs a trusted onshore supplier. Sure TSMC is adding manufacturing capacity in the US and Arizona. But back to our precarious scenario in the South China sea. Will the U S government and Apple sit back and hope for the best or will they hope for the best and plan for the worst? Let's face it. If China gains control of TSMC, it could block access to the latest and greatest process technology. Apple just announced that it's investing billions of dollars in semiconductor technology across the US. The US government is pressuring big tech. What about an Apple Intel joint venture? Apple brings the volume, it's Cloud, it's Cloud, sorry. It's money it's design leadership, all that to the table. And they could partner with Intel. It gives Intel the Foundry business and a guaranteed volume stream. And maybe the U S government gives Apple a little bit of breathing room and the whole big up big breakup, big tech narrative. And even though it's not necessarily specifically targeting Apple but maybe the US government needs to think twice before it attacks big tech and thinks about the long-term strategic ramifications. Wouldn't that be ironic? Apple dumps Intel in favor of Arm for the M1 and then incubates, and essentially saves Intel with a pipeline of Foundry business. Now back to IBM in this scenario, we've put a question mark on the slide because maybe IBM just gets in the way and why not? A nice clean partnership between Intel and Apple? Who knows? Maybe Gelsinger can even negotiate this without giving up any equity to Apple, but Apple could be a key ingredient to a cocktail of a new strategy under Pat Gelsinger leadership. Gobs of cash from the US and EU governments and volume from Apple. Wow, still a long shot, but one worth pursuing because as we've written, Intel is too strategic to fail. Okay, well, what do you think? You can DM me @dvellante or email me at david.vellante@siliconangle.com or comment on my LinkedIn post. Remember, these episodes are all available as podcasts so please subscribe wherever you listen. I publish weekly on wikibon.com and siliconangle.com. And don't forget to check out etr.plus for all the survey analysis. And I want to thank my colleague, David Floyer for his collaboration on this and other related episodes. This is Dave Vellante for theCUBE insights powered by ETR. Thanks for watching, be well, and we'll see you next time. (upbeat music)

>>don't talk mhm, >>Okay, thing is my presentation on coherent nonlinear dynamics and combinatorial optimization. This is going to be a talk to introduce an approach we're taking to the analysis of the performance of coherent using machines. So let me start with a brief introduction to easing optimization. The easing model represents a set of interacting magnetic moments or spins the total energy given by the expression shown at the bottom left of this slide. Here, the signal variables are meditate binary values. The Matrix element J. I. J. Represents the interaction, strength and signed between any pair of spins. I. J and A Chive represents a possible local magnetic field acting on each thing. The easing ground state problem is to find an assignment of binary spin values that achieves the lowest possible value of total energy. And an instance of the easing problem is specified by giving numerical values for the Matrix J in Vector H. Although the easy model originates in physics, we understand the ground state problem to correspond to what would be called quadratic binary optimization in the field of operations research and in fact, in terms of computational complexity theory, it could be established that the easing ground state problem is np complete. Qualitatively speaking, this makes the easing problem a representative sort of hard optimization problem, for which it is expected that the runtime required by any computational algorithm to find exact solutions should, as anatomically scale exponentially with the number of spends and for worst case instances at each end. Of course, there's no reason to believe that the problem instances that actually arrives in practical optimization scenarios are going to be worst case instances. And it's also not generally the case in practical optimization scenarios that we demand absolute optimum solutions. Usually we're more interested in just getting the best solution we can within an affordable cost, where costs may be measured in terms of time, service fees and or energy required for a computation. This focuses great interest on so called heuristic algorithms for the easing problem in other NP complete problems which generally get very good but not guaranteed optimum solutions and run much faster than algorithms that are designed to find absolute Optima. To get some feeling for present day numbers, we can consider the famous traveling salesman problem for which extensive compilations of benchmarking data may be found online. A recent study found that the best known TSP solver required median run times across the Library of Problem instances That scaled is a very steep route exponential for end up to approximately 4500. This gives some indication of the change in runtime scaling for generic as opposed the worst case problem instances. Some of the instances considered in this study were taken from a public library of T SPS derived from real world Veil aside design data. This feels I TSP Library includes instances within ranging from 131 to 744,710 instances from this library with end between 6880 13,584 were first solved just a few years ago in 2017 requiring days of run time and a 48 core to King hurts cluster, while instances with and greater than or equal to 14,233 remain unsolved exactly by any means. Approximate solutions, however, have been found by heuristic methods for all instances in the VLS i TSP library with, for example, a solution within 0.14% of a no lower bound, having been discovered, for instance, with an equal 19,289 requiring approximately two days of run time on a single core of 2.4 gigahertz. Now, if we simple mindedly extrapolate the root exponential scaling from the study up to an equal 4500, we might expect that an exact solver would require something more like a year of run time on the 48 core cluster used for the N equals 13,580 for instance, which shows how much a very small concession on the quality of the solution makes it possible to tackle much larger instances with much lower cost. At the extreme end, the largest TSP ever solved exactly has an equal 85,900. This is an instance derived from 19 eighties VLSI design, and it's required 136 CPU. Years of computation normalized to a single cord, 2.4 gigahertz. But the 24 larger so called world TSP benchmark instance within equals 1,904,711 has been solved approximately within ophthalmology. Gap bounded below 0.474%. Coming back to the general. Practical concerns have applied optimization. We may note that a recent meta study analyzed the performance of no fewer than 37 heuristic algorithms for Max cut and quadratic pioneer optimization problems and found the performance sort and found that different heuristics work best for different problem instances selected from a large scale heterogeneous test bed with some evidence but cryptic structure in terms of what types of problem instances were best solved by any given heuristic. Indeed, their their reasons to believe that these results from Mexico and quadratic binary optimization reflected general principle of performance complementarity among heuristic optimization algorithms in the practice of solving heart optimization problems there. The cerise is a critical pre processing issue of trying to guess which of a number of available good heuristic algorithms should be chosen to tackle a given problem. Instance, assuming that any one of them would incur high costs to run on a large problem, instances incidence, making an astute choice of heuristic is a crucial part of maximizing overall performance. Unfortunately, we still have very little conceptual insight about what makes a specific problem instance, good or bad for any given heuristic optimization algorithm. This has certainly been pinpointed by researchers in the field is a circumstance that must be addressed. So adding this all up, we see that a critical frontier for cutting edge academic research involves both the development of novel heuristic algorithms that deliver better performance, with lower cost on classes of problem instances that are underserved by existing approaches, as well as fundamental research to provide deep conceptual insight into what makes a given problem in, since easy or hard for such algorithms. In fact, these days, as we talk about the end of Moore's law and speculate about a so called second quantum revolution, it's natural to talk not only about novel algorithms for conventional CPUs but also about highly customized special purpose hardware architectures on which we may run entirely unconventional algorithms for combinatorial optimization such as easing problem. So against that backdrop, I'd like to use my remaining time to introduce our work on analysis of coherent using machine architectures and associate ID optimization algorithms. These machines, in general, are a novel class of information processing architectures for solving combinatorial optimization problems by embedding them in the dynamics of analog, physical or cyber physical systems, in contrast to both MAWR traditional engineering approaches that build using machines using conventional electron ICS and more radical proposals that would require large scale quantum entanglement. The emerging paradigm of coherent easing machines leverages coherent nonlinear dynamics in photonic or Opto electronic platforms to enable near term construction of large scale prototypes that leverage post Simoes information dynamics, the general structure of of current CM systems has shown in the figure on the right. The role of the easing spins is played by a train of optical pulses circulating around a fiber optical storage ring. A beam splitter inserted in the ring is used to periodically sample the amplitude of every optical pulse, and the measurement results are continually read into a refugee A, which uses them to compute perturbations to be applied to each pulse by a synchronized optical injections. These perturbations, air engineered to implement the spin, spin coupling and local magnetic field terms of the easing Hamiltonian, corresponding to a linear part of the CME Dynamics, a synchronously pumped parametric amplifier denoted here as PPL and Wave Guide adds a crucial nonlinear component to the CIA and Dynamics as well. In the basic CM algorithm, the pump power starts very low and has gradually increased at low pump powers. The amplitude of the easing spin pulses behaviors continuous, complex variables. Who Israel parts which can be positive or negative, play the role of play the role of soft or perhaps mean field spins once the pump, our crosses the threshold for parametric self oscillation. In the optical fiber ring, however, the attitudes of the easing spin pulses become effectively Qantas ized into binary values while the pump power is being ramped up. The F P J subsystem continuously applies its measurement based feedback. Implementation of the using Hamiltonian terms, the interplay of the linear rised using dynamics implemented by the F P G A and the threshold conversation dynamics provided by the sink pumped Parametric amplifier result in the final state of the optical optical pulse amplitude at the end of the pump ramp that could be read as a binary strain, giving a proposed solution of the easing ground state problem. This method of solving easing problem seems quite different from a conventional algorithm that runs entirely on a digital computer as a crucial aspect of the computation is performed physically by the analog, continuous, coherent, nonlinear dynamics of the optical degrees of freedom. In our efforts to analyze CIA and performance, we have therefore turned to the tools of dynamical systems theory, namely, a study of modifications, the evolution of critical points and apologies of hetero clinic orbits and basins of attraction. We conjecture that such analysis can provide fundamental insight into what makes certain optimization instances hard or easy for coherent using machines and hope that our approach can lead to both improvements of the course, the AM algorithm and a pre processing rubric for rapidly assessing the CME suitability of new instances. Okay, to provide a bit of intuition about how this all works, it may help to consider the threshold dynamics of just one or two optical parametric oscillators in the CME architecture just described. We can think of each of the pulse time slots circulating around the fiber ring, as are presenting an independent Opio. We can think of a single Opio degree of freedom as a single, resonant optical node that experiences linear dissipation, do toe out coupling loss and gain in a pump. Nonlinear crystal has shown in the diagram on the upper left of this slide as the pump power is increased from zero. As in the CME algorithm, the non linear game is initially to low toe overcome linear dissipation, and the Opio field remains in a near vacuum state at a critical threshold. Value gain. Equal participation in the Popeo undergoes a sort of lazing transition, and the study states of the OPIO above this threshold are essentially coherent states. There are actually two possible values of the Opio career in amplitude and any given above threshold pump power which are equal in magnitude but opposite in phase when the OPI across the special diet basically chooses one of the two possible phases randomly, resulting in the generation of a single bit of information. If we consider to uncoupled, Opio has shown in the upper right diagram pumped it exactly the same power at all times. Then, as the pump power has increased through threshold, each Opio will independently choose the phase and thus to random bits are generated for any number of uncoupled. Oppose the threshold power per opio is unchanged from the single Opio case. Now, however, consider a scenario in which the two appeals air, coupled to each other by a mutual injection of their out coupled fields has shown in the diagram on the lower right. One can imagine that depending on the sign of the coupling parameter Alfa, when one Opio is lazing, it will inject a perturbation into the other that may interfere either constructively or destructively, with the feel that it is trying to generate by its own lazing process. As a result, when came easily showed that for Alfa positive, there's an effective ferro magnetic coupling between the two Opio fields and their collective oscillation threshold is lowered from that of the independent Opio case. But on Lee for the two collective oscillation modes in which the two Opio phases are the same for Alfa Negative, the collective oscillation threshold is lowered on Lee for the configurations in which the Opio phases air opposite. So then, looking at how Alfa is related to the J. I. J matrix of the easing spin coupling Hamiltonian, it follows that we could use this simplistic to a p o. C. I am to solve the ground state problem of a fair magnetic or anti ferro magnetic ankles to easing model simply by increasing the pump power from zero and observing what phase relation occurs as the two appeals first start delays. Clearly, we can imagine generalizing this story toe larger, and however the story doesn't stay is clean and simple for all larger problem instances. And to find a more complicated example, we only need to go to n equals four for some choices of J J for n equals, for the story remains simple. Like the n equals two case. The figure on the upper left of this slide shows the energy of various critical points for a non frustrated and equals, for instance, in which the first bifurcated critical point that is the one that I forget to the lowest pump value a. Uh, this first bifurcated critical point flows as symptomatically into the lowest energy easing solution and the figure on the upper right. However, the first bifurcated critical point flows to a very good but sub optimal minimum at large pump power. The global minimum is actually given by a distinct critical critical point that first appears at a higher pump power and is not automatically connected to the origin. The basic C am algorithm is thus not able to find this global minimum. Such non ideal behaviors needs to become more confident. Larger end for the n equals 20 instance, showing the lower plots where the lower right plot is just a zoom into a region of the lower left lot. It can be seen that the global minimum corresponds to a critical point that first appears out of pump parameter, a around 0.16 at some distance from the idiomatic trajectory of the origin. That's curious to note that in both of these small and examples, however, the critical point corresponding to the global minimum appears relatively close to the idiomatic projector of the origin as compared to the most of the other local minima that appear. We're currently working to characterize the face portrait topology between the global minimum in the antibiotic trajectory of the origin, taking clues as to how the basic C am algorithm could be generalized to search for non idiomatic trajectories that jump to the global minimum during the pump ramp. Of course, n equals 20 is still too small to be of interest for practical optimization applications. But the advantage of beginning with the study of small instances is that we're able reliably to determine their global minima and to see how they relate to the 80 about trajectory of the origin in the basic C am algorithm. In the smaller and limit, we can also analyze fully quantum mechanical models of Syrian dynamics. But that's a topic for future talks. Um, existing large scale prototypes are pushing into the range of in equals 10 to the 4 10 to 5 to six. So our ultimate objective in theoretical analysis really has to be to try to say something about CIA and dynamics and regime of much larger in our initial approach to characterizing CIA and behavior in the large in regime relies on the use of random matrix theory, and this connects to prior research on spin classes, SK models and the tap equations etcetera. At present, we're focusing on statistical characterization of the CIA ingredient descent landscape, including the evolution of critical points in their Eigen value spectra. As the pump power is gradually increased. We're investigating, for example, whether there could be some way to exploit differences in the relative stability of the global minimum versus other local minima. We're also working to understand the deleterious or potentially beneficial effects of non ideologies, such as a symmetry in the implemented these and couplings. Looking one step ahead, we plan to move next in the direction of considering more realistic classes of problem instances such as quadratic, binary optimization with constraints. Eso In closing, I should acknowledge people who did the hard work on these things that I've shown eso. My group, including graduate students Ed winning, Daniel Wennberg, Tatsuya Nagamoto and Atsushi Yamamura, have been working in close collaboration with Syria Ganguly, Marty Fair and Amir Safarini Nini, all of us within the Department of Applied Physics at Stanford University. On also in collaboration with the Oshima Moto over at NTT 55 research labs, Onda should acknowledge funding support from the NSF by the Coherent Easing Machines Expedition in computing, also from NTT five research labs, Army Research Office and Exxon Mobil. Uh, that's it. Thanks very much. >>Mhm e >>t research and the Oshie for putting together this program and also the opportunity to speak here. My name is Al Gore ism or Andy and I'm from Caltech, and today I'm going to tell you about the work that we have been doing on networks off optical parametric oscillators and how we have been using them for icing machines and how we're pushing them toward Cornum photonics to acknowledge my team at Caltech, which is now eight graduate students and five researcher and postdocs as well as collaborators from all over the world, including entity research and also the funding from different places, including entity. So this talk is primarily about networks of resonate er's, and these networks are everywhere from nature. For instance, the brain, which is a network of oscillators all the way to optics and photonics and some of the biggest examples or metal materials, which is an array of small resonate er's. And we're recently the field of technological photonics, which is trying thio implement a lot of the technological behaviors of models in the condensed matter, physics in photonics and if you want to extend it even further, some of the implementations off quantum computing are technically networks of quantum oscillators. So we started thinking about these things in the context of icing machines, which is based on the icing problem, which is based on the icing model, which is the simple summation over the spins and spins can be their upward down and the couplings is given by the JJ. And the icing problem is, if you know J I J. What is the spin configuration that gives you the ground state? And this problem is shown to be an MP high problem. So it's computational e important because it's a representative of the MP problems on NPR. Problems are important because first, their heart and standard computers if you use a brute force algorithm and they're everywhere on the application side. That's why there is this demand for making a machine that can target these problems, and hopefully it can provide some meaningful computational benefit compared to the standard digital computers. So I've been building these icing machines based on this building block, which is a degenerate optical parametric. Oscillator on what it is is resonator with non linearity in it, and we pump these resonate er's and we generate the signal at half the frequency of the pump. One vote on a pump splits into two identical photons of signal, and they have some very interesting phase of frequency locking behaviors. And if you look at the phase locking behavior, you realize that you can actually have two possible phase states as the escalation result of these Opio which are off by pie, and that's one of the important characteristics of them. So I want to emphasize a little more on that and I have this mechanical analogy which are basically two simple pendulum. But there are parametric oscillators because I'm going to modulate the parameter of them in this video, which is the length of the string on by that modulation, which is that will make a pump. I'm gonna make a muscular. That'll make a signal which is half the frequency of the pump. And I have two of them to show you that they can acquire these face states so they're still facing frequency lock to the pump. But it can also lead in either the zero pie face states on. The idea is to use this binary phase to represent the binary icing spin. So each opio is going to represent spin, which can be either is your pie or up or down. And to implement the network of these resonate er's, we use the time off blood scheme, and the idea is that we put impulses in the cavity. These pulses air separated by the repetition period that you put in or t r. And you can think about these pulses in one resonator, xaz and temporarily separated synthetic resonate Er's if you want a couple of these resonator is to each other, and now you can introduce these delays, each of which is a multiple of TR. If you look at the shortest delay it couples resonator wanted to 2 to 3 and so on. If you look at the second delay, which is two times a rotation period, the couple's 123 and so on. And if you have and minus one delay lines, then you can have any potential couplings among these synthetic resonate er's. And if I can introduce these modulators in those delay lines so that I can strength, I can control the strength and the phase of these couplings at the right time. Then I can have a program will all toe all connected network in this time off like scheme, and the whole physical size of the system scales linearly with the number of pulses. So the idea of opium based icing machine is didn't having these o pos, each of them can be either zero pie and I can arbitrarily connect them to each other. And then I start with programming this machine to a given icing problem by just setting the couplings and setting the controllers in each of those delight lines. So now I have a network which represents an icing problem. Then the icing problem maps to finding the face state that satisfy maximum number of coupling constraints. And the way it happens is that the icing Hamiltonian maps to the linear loss of the network. And if I start adding gain by just putting pump into the network, then the OPI ohs are expected to oscillate in the lowest, lowest lost state. And, uh and we have been doing these in the past, uh, six or seven years and I'm just going to quickly show you the transition, especially what happened in the first implementation, which was using a free space optical system and then the guided wave implementation in 2016 and the measurement feedback idea which led to increasing the size and doing actual computation with these machines. So I just want to make this distinction here that, um, the first implementation was an all optical interaction. We also had an unequal 16 implementation. And then we transition to this measurement feedback idea, which I'll tell you quickly what it iss on. There's still a lot of ongoing work, especially on the entity side, to make larger machines using the measurement feedback. But I'm gonna mostly focused on the all optical networks and how we're using all optical networks to go beyond simulation of icing Hamiltonian both in the linear and non linear side and also how we're working on miniaturization of these Opio networks. So the first experiment, which was the four opium machine, it was a free space implementation and this is the actual picture off the machine and we implemented a small and it calls for Mexico problem on the machine. So one problem for one experiment and we ran the machine 1000 times, we looked at the state and we always saw it oscillate in one of these, um, ground states of the icing laboratoria. So then the measurement feedback idea was to replace those couplings and the controller with the simulator. So we basically simulated all those coherent interactions on on FB g. A. And we replicated the coherent pulse with respect to all those measurements. And then we injected it back into the cavity and on the near to you still remain. So it still is a non. They're dynamical system, but the linear side is all simulated. So there are lots of questions about if this system is preserving important information or not, or if it's gonna behave better. Computational wars. And that's still ah, lot of ongoing studies. But nevertheless, the reason that this implementation was very interesting is that you don't need the end minus one delight lines so you can just use one. Then you can implement a large machine, and then you can run several thousands of problems in the machine, and then you can compare the performance from the computational perspective Looks so I'm gonna split this idea of opium based icing machine into two parts. One is the linear part, which is if you take out the non linearity out of the resonator and just think about the connections. You can think about this as a simple matrix multiplication scheme. And that's basically what gives you the icing Hambletonian modeling. So the optical laws of this network corresponds to the icing Hamiltonian. And if I just want to show you the example of the n equals for experiment on all those face states and the history Graham that we saw, you can actually calculate the laws of each of those states because all those interferences in the beam splitters and the delay lines are going to give you a different losses. And then you will see that the ground states corresponds to the lowest laws of the actual optical network. If you add the non linearity, the simple way of thinking about what the non linearity does is that it provides to gain, and then you start bringing up the gain so that it hits the loss. Then you go through the game saturation or the threshold which is going to give you this phase bifurcation. So you go either to zero the pie face state. And the expectation is that Theis, the network oscillates in the lowest possible state, the lowest possible loss state. There are some challenges associated with this intensity Durban face transition, which I'm going to briefly talk about. I'm also going to tell you about other types of non aerodynamics that we're looking at on the non air side of these networks. So if you just think about the linear network, we're actually interested in looking at some technological behaviors in these networks. And the difference between looking at the technological behaviors and the icing uh, machine is that now, First of all, we're looking at the type of Hamilton Ian's that are a little different than the icing Hamilton. And one of the biggest difference is is that most of these technological Hamilton Ian's that require breaking the time reversal symmetry, meaning that you go from one spin to in the one side to another side and you get one phase. And if you go back where you get a different phase, and the other thing is that we're not just interested in finding the ground state, we're actually now interesting and looking at all sorts of states and looking at the dynamics and the behaviors of all these states in the network. So we started with the simplest implementation, of course, which is a one d chain of thes resonate, er's, which corresponds to a so called ssh model. In the technological work, we get the similar energy to los mapping and now we can actually look at the band structure on. This is an actual measurement that we get with this associate model and you see how it reasonably how How? Well, it actually follows the prediction and the theory. One of the interesting things about the time multiplexing implementation is that now you have the flexibility of changing the network as you are running the machine. And that's something unique about this time multiplex implementation so that we can actually look at the dynamics. And one example that we have looked at is we can actually go through the transition off going from top A logical to the to the standard nontrivial. I'm sorry to the trivial behavior of the network. You can then look at the edge states and you can also see the trivial and states and the technological at states actually showing up in this network. We have just recently implement on a two D, uh, network with Harper Hofstadter model and when you don't have the results here. But we're one of the other important characteristic of time multiplexing is that you can go to higher and higher dimensions and keeping that flexibility and dynamics, and we can also think about adding non linearity both in a classical and quantum regimes, which is going to give us a lot of exotic, no classical and quantum, non innate behaviors in these networks. Yeah, So I told you about the linear side. Mostly let me just switch gears and talk about the nonlinear side of the network. And the biggest thing that I talked about so far in the icing machine is this face transition that threshold. So the low threshold we have squeezed state in these. Oh, pios, if you increase the pump, we go through this intensity driven phase transition and then we got the face stays above threshold. And this is basically the mechanism off the computation in these O pos, which is through this phase transition below to above threshold. So one of the characteristics of this phase transition is that below threshold, you expect to see quantum states above threshold. You expect to see more classical states or coherent states, and that's basically corresponding to the intensity off the driving pump. So it's really hard to imagine that it can go above threshold. Or you can have this friends transition happen in the all in the quantum regime. And there are also some challenges associated with the intensity homogeneity off the network, which, for example, is if one opioid starts oscillating and then its intensity goes really high. Then it's going to ruin this collective decision making off the network because of the intensity driven face transition nature. So So the question is, can we look at other phase transitions? Can we utilize them for both computing? And also can we bring them to the quantum regime on? I'm going to specifically talk about the face transition in the spectral domain, which is the transition from the so called degenerate regime, which is what I mostly talked about to the non degenerate regime, which happens by just tuning the phase of the cavity. And what is interesting is that this phase transition corresponds to a distinct phase noise behavior. So in the degenerate regime, which we call it the order state, you're gonna have the phase being locked to the phase of the pump. As I talked about non degenerate regime. However, the phase is the phase is mostly dominated by the quantum diffusion. Off the off the phase, which is limited by the so called shallow towns limit, and you can see that transition from the general to non degenerate, which also has distinct symmetry differences. And this transition corresponds to a symmetry breaking in the non degenerate case. The signal can acquire any of those phases on the circle, so it has a you one symmetry. Okay, and if you go to the degenerate case, then that symmetry is broken and you only have zero pie face days I will look at. So now the question is can utilize this phase transition, which is a face driven phase transition, and can we use it for similar computational scheme? So that's one of the questions that were also thinking about. And it's not just this face transition is not just important for computing. It's also interesting from the sensing potentials and this face transition, you can easily bring it below threshold and just operated in the quantum regime. Either Gaussian or non Gaussian. If you make a network of Opio is now, we can see all sorts off more complicated and more interesting phase transitions in the spectral domain. One of them is the first order phase transition, which you get by just coupling to Opio, and that's a very abrupt face transition and compared to the to the single Opio phase transition. And if you do the couplings right, you can actually get a lot of non her mission dynamics and exceptional points, which are actually very interesting to explore both in the classical and quantum regime. And I should also mention that you can think about the cup links to be also nonlinear couplings. And that's another behavior that you can see, especially in the nonlinear in the non degenerate regime. So with that, I basically told you about these Opio networks, how we can think about the linear scheme and the linear behaviors and how we can think about the rich, nonlinear dynamics and non linear behaviors both in the classical and quantum regime. I want to switch gear and tell you a little bit about the miniaturization of these Opio networks. And of course, the motivation is if you look at the electron ICS and what we had 60 or 70 years ago with vacuum tube and how we transition from relatively small scale computers in the order of thousands of nonlinear elements to billions of non elements where we are now with the optics is probably very similar to 70 years ago, which is a table talk implementation. And the question is, how can we utilize nano photonics? I'm gonna just briefly show you the two directions on that which we're working on. One is based on lithium Diabate, and the other is based on even a smaller resonate er's could you? So the work on Nana Photonic lithium naive. It was started in collaboration with Harvard Marko Loncar, and also might affair at Stanford. And, uh, we could show that you can do the periodic polling in the phenomenon of it and get all sorts of very highly nonlinear processes happening in this net. Photonic periodically polls if, um Diabate. And now we're working on building. Opio was based on that kind of photonic the film Diabate. And these air some some examples of the devices that we have been building in the past few months, which I'm not gonna tell you more about. But the O. P. O. S. And the Opio Networks are in the works. And that's not the only way of making large networks. Um, but also I want to point out that The reason that these Nana photonic goblins are actually exciting is not just because you can make a large networks and it can make him compact in a in a small footprint. They also provide some opportunities in terms of the operation regime. On one of them is about making cat states and Opio, which is, can we have the quantum superposition of the zero pie states that I talked about and the Net a photonic within? I've It provides some opportunities to actually get closer to that regime because of the spatial temporal confinement that you can get in these wave guides. So we're doing some theory on that. We're confident that the type of non linearity two losses that it can get with these platforms are actually much higher than what you can get with other platform their existing platforms and to go even smaller. We have been asking the question off. What is the smallest possible Opio that you can make? Then you can think about really wavelength scale type, resonate er's and adding the chi to non linearity and see how and when you can get the Opio to operate. And recently, in collaboration with us see, we have been actually USC and Creole. We have demonstrated that you can use nano lasers and get some spin Hamilton and implementations on those networks. So if you can build the a P. O s, we know that there is a path for implementing Opio Networks on on such a nano scale. So we have looked at these calculations and we try to estimate the threshold of a pos. Let's say for me resonator and it turns out that it can actually be even lower than the type of bulk Pip Llano Pos that we have been building in the past 50 years or so. So we're working on the experiments and we're hoping that we can actually make even larger and larger scale Opio networks. So let me summarize the talk I told you about the opium networks and our work that has been going on on icing machines and the measurement feedback. And I told you about the ongoing work on the all optical implementations both on the linear side and also on the nonlinear behaviors. And I also told you a little bit about the efforts on miniaturization and going to the to the Nano scale. So with that, I would like Thio >>three from the University of Tokyo. Before I thought that would like to thank you showing all the stuff of entity for the invitation and the organization of this online meeting and also would like to say that it has been very exciting to see the growth of this new film lab. And I'm happy to share with you today of some of the recent works that have been done either by me or by character of Hong Kong. Honest Group indicates the title of my talk is a neuro more fic in silica simulator for the communities in machine. And here is the outline I would like to make the case that the simulation in digital Tektronix of the CME can be useful for the better understanding or improving its function principles by new job introducing some ideas from neural networks. This is what I will discuss in the first part and then it will show some proof of concept of the game and performance that can be obtained using dissimulation in the second part and the protection of the performance that can be achieved using a very large chaos simulator in the third part and finally talk about future plans. So first, let me start by comparing recently proposed izing machines using this table there is elected from recent natural tronics paper from the village Park hard people, and this comparison shows that there's always a trade off between energy efficiency, speed and scalability that depends on the physical implementation. So in red, here are the limitation of each of the servers hardware on, interestingly, the F p G, a based systems such as a producer, digital, another uh Toshiba beautification machine or a recently proposed restricted Bozeman machine, FPD A by a group in Berkeley. They offer a good compromise between speed and scalability. And this is why, despite the unique advantage that some of these older hardware have trust as the currency proposition in Fox, CBS or the energy efficiency off memory Sisters uh P. J. O are still an attractive platform for building large organizing machines in the near future. The reason for the good performance of Refugee A is not so much that they operate at the high frequency. No, there are particular in use, efficient, but rather that the physical wiring off its elements can be reconfigured in a way that limits the funding human bottleneck, larger, funny and phenols and the long propagation video information within the system. In this respect, the LPGA is They are interesting from the perspective off the physics off complex systems, but then the physics of the actions on the photos. So to put the performance of these various hardware and perspective, we can look at the competition of bringing the brain the brain complete, using billions of neurons using only 20 watts of power and operates. It's a very theoretically slow, if we can see and so this impressive characteristic, they motivate us to try to investigate. What kind of new inspired principles be useful for designing better izing machines? The idea of this research project in the future collaboration it's to temporary alleviates the limitations that are intrinsic to the realization of an optical cortex in machine shown in the top panel here. By designing a large care simulator in silicone in the bottom here that can be used for digesting the better organization principles of the CIA and this talk, I will talk about three neuro inspired principles that are the symmetry of connections, neural dynamics orphan chaotic because of symmetry, is interconnectivity the infrastructure? No. Next talks are not composed of the reputation of always the same types of non environments of the neurons, but there is a local structure that is repeated. So here's the schematic of the micro column in the cortex. And lastly, the Iraqi co organization of connectivity connectivity is organizing a tree structure in the brain. So here you see a representation of the Iraqi and organization of the monkey cerebral cortex. So how can these principles we used to improve the performance of the icing machines? And it's in sequence stimulation. So, first about the two of principles of the estimate Trian Rico structure. We know that the classical approximation of the car testing machine, which is the ground toe, the rate based on your networks. So in the case of the icing machines, uh, the okay, Scott approximation can be obtained using the trump active in your position, for example, so the times of both of the system they are, they can be described by the following ordinary differential equations on in which, in case of see, I am the X, I represent the in phase component of one GOP Oh, Theo f represents the monitor optical parts, the district optical Parametric amplification and some of the good I JoJo extra represent the coupling, which is done in the case of the measure of feedback coupling cm using oh, more than detection and refugee A and then injection off the cooking time and eso this dynamics in both cases of CNN in your networks, they can be written as the grand set of a potential function V, and this written here, and this potential functionally includes the rising Maccagnan. So this is why it's natural to use this type of, uh, dynamics to solve the icing problem in which the Omega I J or the eyes in coping and the H is the extension of the icing and attorney in India and expect so. Not that this potential function can only be defined if the Omega I j. R. A. Symmetric. So the well known problem of this approach is that this potential function V that we obtain is very non convicts at low temperature, and also one strategy is to gradually deformed this landscape, using so many in process. But there is no theorem. Unfortunately, that granted conventions to the global minimum of There's even Tony and using this approach. And so this is why we propose, uh, to introduce a macro structures of the system where one analog spin or one D O. P. O is replaced by a pair off one another spin and one error, according viable. And the addition of this chemical structure introduces a symmetry in the system, which in terms induces chaotic dynamics, a chaotic search rather than a learning process for searching for the ground state of the icing. Every 20 within this massacre structure the role of the er variable eyes to control the amplitude off the analog spins toe force. The amplitude of the expense toe become equal to certain target amplitude a uh and, uh, and this is done by modulating the strength off the icing complaints or see the the error variable E I multiply the icing complaint here in the dynamics off air d o p. O. On then the dynamics. The whole dynamics described by this coupled equations because the e I do not necessarily take away the same value for the different. I thesis introduces a symmetry in the system, which in turn creates security dynamics, which I'm sure here for solving certain current size off, um, escape problem, Uh, in which the X I are shown here and the i r from here and the value of the icing energy showing the bottom plots. You see this Celtics search that visit various local minima of the as Newtonian and eventually finds the global minimum? Um, it can be shown that this modulation off the target opportunity can be used to destabilize all the local minima off the icing evertonians so that we're gonna do not get stuck in any of them. On more over the other types of attractors I can eventually appear, such as limits I contractors, Okot contractors. They can also be destabilized using the motivation of the target and Batuta. And so we have proposed in the past two different moderation of the target amateur. The first one is a modulation that ensure the uh 100 reproduction rate of the system to become positive on this forbids the creation off any nontrivial tractors. And but in this work, I will talk about another moderation or arrested moderation which is given here. That works, uh, as well as this first uh, moderation, but is easy to be implemented on refugee. So this couple of the question that represent becoming the stimulation of the cortex in machine with some error correction they can be implemented especially efficiently on an F B. G. And here I show the time that it takes to simulate three system and also in red. You see, at the time that it takes to simulate the X I term the EI term, the dot product and the rising Hamiltonian for a system with 500 spins and Iraq Spain's equivalent to 500 g. O. P. S. So >>in >>f b d a. The nonlinear dynamics which, according to the digital optical Parametric amplification that the Opa off the CME can be computed in only 13 clock cycles at 300 yards. So which corresponds to about 0.1 microseconds. And this is Toby, uh, compared to what can be achieved in the measurements back O C. M. In which, if we want to get 500 timer chip Xia Pios with the one she got repetition rate through the obstacle nine narrative. Uh, then way would require 0.5 microseconds toe do this so the submission in F B J can be at least as fast as ah one g repression. Uh, replicate pulsed laser CIA Um, then the DOT product that appears in this differential equation can be completed in 43 clock cycles. That's to say, one microseconds at 15 years. So I pieced for pouring sizes that are larger than 500 speeds. The dot product becomes clearly the bottleneck, and this can be seen by looking at the the skating off the time the numbers of clock cycles a text to compute either the non in your optical parts or the dog products, respect to the problem size. And And if we had infinite amount of resources and PGA to simulate the dynamics, then the non illogical post can could be done in the old one. On the mattress Vector product could be done in the low carrot off, located off scales as a look at it off and and while the guide off end. Because computing the dot product involves assuming all the terms in the product, which is done by a nephew, GE by another tree, which heights scarce logarithmic any with the size of the system. But This is in the case if we had an infinite amount of resources on the LPGA food, but for dealing for larger problems off more than 100 spins. Usually we need to decompose the metrics into ah, smaller blocks with the block side that are not you here. And then the scaling becomes funny, non inner parts linear in the end, over you and for the products in the end of EU square eso typically for low NF pdf cheap PGA you the block size off this matrix is typically about 100. So clearly way want to make you as large as possible in order to maintain this scanning in a log event for the numbers of clock cycles needed to compute the product rather than this and square that occurs if we decompose the metrics into smaller blocks. But the difficulty in, uh, having this larger blocks eyes that having another tree very large Haider tree introduces a large finding and finance and long distance start a path within the refugee. So the solution to get higher performance for a simulator of the contest in machine eyes to get rid of this bottleneck for the dot product by increasing the size of this at the tree. And this can be done by organizing your critique the electrical components within the LPGA in order which is shown here in this, uh, right panel here in order to minimize the finding finance of the system and to minimize the long distance that a path in the in the fpt So I'm not going to the details of how this is implemented LPGA. But just to give you a idea off why the Iraqi Yahiko organization off the system becomes the extremely important toe get good performance for similar organizing machine. So instead of instead of getting into the details of the mpg implementation, I would like to give some few benchmark results off this simulator, uh, off the that that was used as a proof of concept for this idea which is can be found in this archive paper here and here. I should results for solving escape problems. Free connected person, randomly person minus one spring last problems and we sure, as we use as a metric the numbers of the mattress Victor products since it's the bottleneck of the computation, uh, to get the optimal solution of this escape problem with the Nina successful BT against the problem size here and and in red here, this propose FDJ implementation and in ah blue is the numbers of retrospective product that are necessary for the C. I am without error correction to solve this escape programs and in green here for noisy means in an evening which is, uh, behavior with similar to the Cartesian mission. Uh, and so clearly you see that the scaring off the numbers of matrix vector product necessary to solve this problem scales with a better exponents than this other approaches. So So So that's interesting feature of the system and next we can see what is the real time to solution to solve this SK instances eso in the last six years, the time institution in seconds to find a grand state of risk. Instances remain answers probability for different state of the art hardware. So in red is the F B g. A presentation proposing this paper and then the other curve represent Ah, brick a local search in in orange and silver lining in purple, for example. And so you see that the scaring off this purpose simulator is is rather good, and that for larger plant sizes we can get orders of magnitude faster than the state of the art approaches. Moreover, the relatively good scanning off the time to search in respect to problem size uh, they indicate that the FPD implementation would be faster than risk. Other recently proposed izing machine, such as the hope you know, natural complimented on memories distance that is very fast for small problem size in blue here, which is very fast for small problem size. But which scanning is not good on the same thing for the restricted Bosman machine. Implementing a PGA proposed by some group in Broken Recently Again, which is very fast for small parliament sizes but which canning is bad so that a dis worse than the proposed approach so that we can expect that for programs size is larger than 1000 spins. The proposed, of course, would be the faster one. Let me jump toe this other slide and another confirmation that the scheme scales well that you can find the maximum cut values off benchmark sets. The G sets better candidates that have been previously found by any other algorithms, so they are the best known could values to best of our knowledge. And, um or so which is shown in this paper table here in particular, the instances, uh, 14 and 15 of this G set can be We can find better converse than previously known, and we can find this can vary is 100 times faster than the state of the art algorithm and CP to do this which is a very common Kasich. It s not that getting this a good result on the G sets, they do not require ah, particular hard tuning of the parameters. So the tuning issuing here is very simple. It it just depends on the degree off connectivity within each graph. And so this good results on the set indicate that the proposed approach would be a good not only at solving escape problems in this problems, but all the types off graph sizing problems on Mexican province in communities. So given that the performance off the design depends on the height of this other tree, we can try to maximize the height of this other tree on a large F p g a onda and carefully routing the components within the P G A and and we can draw some projections of what type of performance we can achieve in the near future based on the, uh, implementation that we are currently working. So here you see projection for the time to solution way, then next property for solving this escape programs respect to the prime assize. And here, compared to different with such publicizing machines, particularly the digital. And, you know, 42 is shown in the green here, the green line without that's and, uh and we should two different, uh, hypothesis for this productions either that the time to solution scales as exponential off n or that the time of social skills as expression of square root off. So it seems, according to the data, that time solution scares more as an expression of square root of and also we can be sure on this and this production show that we probably can solve prime escape problem of science 2000 spins, uh, to find the rial ground state of this problem with 99 success ability in about 10 seconds, which is much faster than all the other proposed approaches. So one of the future plans for this current is in machine simulator. So the first thing is that we would like to make dissimulation closer to the rial, uh, GOP oh, optical system in particular for a first step to get closer to the system of a measurement back. See, I am. And to do this what is, uh, simulate Herbal on the p a is this quantum, uh, condoms Goshen model that is proposed described in this paper and proposed by people in the in the Entity group. And so the idea of this model is that instead of having the very simple or these and have shown previously, it includes paired all these that take into account on me the mean off the awesome leverage off the, uh, European face component, but also their violence s so that we can take into account more quantum effects off the g o p. O, such as the squeezing. And then we plan toe, make the simulator open access for the members to run their instances on the system. There will be a first version in September that will be just based on the simple common line access for the simulator and in which will have just a classic or approximation of the system. We don't know Sturm, binary weights and museum in term, but then will propose a second version that would extend the current arising machine to Iraq off F p g. A, in which we will add the more refined models truncated, ignoring the bottom Goshen model they just talked about on the support in which he valued waits for the rising problems and support the cement. So we will announce later when this is available and and far right is working >>hard comes from Universal down today in physics department, and I'd like to thank the organizers for their kind invitation to participate in this very interesting and promising workshop. Also like to say that I look forward to collaborations with with a file lab and Yoshi and collaborators on the topics of this world. So today I'll briefly talk about our attempt to understand the fundamental limits off another continues time computing, at least from the point off you off bullion satisfy ability, problem solving, using ordinary differential equations. But I think the issues that we raise, um, during this occasion actually apply to other other approaches on a log approaches as well and into other problems as well. I think everyone here knows what Dorien satisfy ability. Problems are, um, you have boolean variables. You have em clauses. Each of disjunction of collaterals literally is a variable, or it's, uh, negation. And the goal is to find an assignment to the variable, such that order clauses are true. This is a decision type problem from the MP class, which means you can checking polynomial time for satisfy ability off any assignment. And the three set is empty, complete with K three a larger, which means an efficient trees. That's over, uh, implies an efficient source for all the problems in the empty class, because all the problems in the empty class can be reduced in Polian on real time to reset. As a matter of fact, you can reduce the NP complete problems into each other. You can go from three set to set backing or two maximum dependent set, which is a set packing in graph theoretic notions or terms toe the icing graphs. A problem decision version. This is useful, and you're comparing different approaches, working on different kinds of problems when not all the closest can be satisfied. You're looking at the accusation version offset, uh called Max Set. And the goal here is to find assignment that satisfies the maximum number of clauses. And this is from the NPR class. In terms of applications. If we had inefficient sets over or np complete problems over, it was literally, positively influenced. Thousands off problems and applications in industry and and science. I'm not going to read this, but this this, of course, gives a strong motivation toe work on this kind of problems. Now our approach to set solving involves embedding the problem in a continuous space, and you use all the east to do that. So instead of working zeros and ones, we work with minus one across once, and we allow the corresponding variables toe change continuously between the two bounds. We formulate the problem with the help of a close metrics. If if a if a close, uh, does not contain a variable or its negation. The corresponding matrix element is zero. If it contains the variable in positive, for which one contains the variable in a gated for Mitt's negative one, and then we use this to formulate this products caused quote, close violation functions one for every clause, Uh, which really, continuously between zero and one. And they're zero if and only if the clause itself is true. Uh, then we form the define in order to define a dynamic such dynamics in this and dimensional hyper cube where the search happens and if they exist, solutions. They're sitting in some of the corners of this hyper cube. So we define this, uh, energy potential or landscape function shown here in a way that this is zero if and only if all the clauses all the kmc zero or the clauses off satisfied keeping these auxiliary variables a EMS always positive. And therefore, what you do here is a dynamics that is a essentially ingredient descend on this potential energy landscape. If you were to keep all the M's constant that it would get stuck in some local minimum. However, what we do here is we couple it with the dynamics we cooperated the clothes violation functions as shown here. And if he didn't have this am here just just the chaos. For example, you have essentially what case you have positive feedback. You have increasing variable. Uh, but in that case, you still get stuck would still behave will still find. So she is better than the constant version but still would get stuck only when you put here this a m which makes the dynamics in in this variable exponential like uh, only then it keeps searching until he finds a solution on deer is a reason for that. I'm not going toe talk about here, but essentially boils down toe performing a Grady and descend on a globally time barren landscape. And this is what works. Now I'm gonna talk about good or bad and maybe the ugly. Uh, this is, uh, this is What's good is that it's a hyperbolic dynamical system, which means that if you take any domain in the search space that doesn't have a solution in it or any socially than the number of trajectories in it decays exponentially quickly. And the decay rate is a characteristic in variant characteristic off the dynamics itself. Dynamical systems called the escape right the inverse off that is the time scale in which you find solutions by this by this dynamical system, and you can see here some song trajectories that are Kelty because it's it's no linear, but it's transient, chaotic. Give their sources, of course, because eventually knowledge to the solution. Now, in terms of performance here, what you show for a bunch off, um, constraint densities defined by M overran the ratio between closes toe variables for random, said Problems is random. Chris had problems, and they as its function off n And we look at money toward the wartime, the wall clock time and it behaves quite value behaves Azat party nominally until you actually he to reach the set on set transition where the hardest problems are found. But what's more interesting is if you monitor the continuous time t the performance in terms off the A narrow, continuous Time t because that seems to be a polynomial. And the way we show that is, we consider, uh, random case that random three set for a fixed constraint density Onda. We hear what you show here. Is that the right of the trash hold that it's really hard and, uh, the money through the fraction of problems that we have not been able to solve it. We select thousands of problems at that constraint ratio and resolve them without algorithm, and we monitor the fractional problems that have not yet been solved by continuous 90. And this, as you see these decays exponentially different. Educate rates for different system sizes, and in this spot shows that is dedicated behaves polynomial, or actually as a power law. So if you combine these two, you find that the time needed to solve all problems except maybe appear traction off them scales foreign or merely with the problem size. So you have paranormal, continuous time complexity. And this is also true for other types of very hard constraints and sexual problems such as exact cover, because you can always transform them into three set as we discussed before, Ramsey coloring and and on these problems, even algorithms like survey propagation will will fail. But this doesn't mean that P equals NP because what you have first of all, if you were toe implement these equations in a device whose behavior is described by these, uh, the keys. Then, of course, T the continue style variable becomes a physical work off. Time on that will be polynomial is scaling, but you have another other variables. Oxidative variables, which structured in an exponential manner. So if they represent currents or voltages in your realization and it would be an exponential cost Al Qaeda. But this is some kind of trade between time and energy, while I know how toe generate energy or I don't know how to generate time. But I know how to generate energy so it could use for it. But there's other issues as well, especially if you're trying toe do this son and digital machine but also happens. Problems happen appear. Other problems appear on in physical devices as well as we discuss later. So if you implement this in GPU, you can. Then you can get in order off to magnitude. Speed up. And you can also modify this to solve Max sad problems. Uh, quite efficiently. You are competitive with the best heuristic solvers. This is a weather problems. In 2016 Max set competition eso so this this is this is definitely this seems like a good approach, but there's off course interesting limitations, I would say interesting, because it kind of makes you think about what it means and how you can exploit this thes observations in understanding better on a low continues time complexity. If you monitored the discrete number the number of discrete steps. Don't buy the room, Dakota integrator. When you solve this on a digital machine, you're using some kind of integrator. Um and you're using the same approach. But now you measure the number off problems you haven't sold by given number of this kid, uh, steps taken by the integrator. You find out you have exponential, discrete time, complexity and, of course, thistles. A problem. And if you look closely, what happens even though the analog mathematical trajectory, that's the record here. If you monitor what happens in discrete time, uh, the integrator frustrates very little. So this is like, you know, third or for the disposition, but fluctuates like crazy. So it really is like the intervention frees us out. And this is because of the phenomenon of stiffness that are I'll talk a little bit a more about little bit layer eso. >>You know, it might look >>like an integration issue on digital machines that you could improve and could definitely improve. But actually issues bigger than that. It's It's deeper than that, because on a digital machine there is no time energy conversion. So the outside variables are efficiently representing a digital machine. So there's no exponential fluctuating current of wattage in your computer when you do this. Eso If it is not equal NP then the exponential time, complexity or exponential costs complexity has to hit you somewhere. And this is how um, but, you know, one would be tempted to think maybe this wouldn't be an issue in a analog device, and to some extent is true on our devices can be ordered to maintain faster, but they also suffer from their own problems because he not gonna be affect. That classes soldiers as well. So, indeed, if you look at other systems like Mirandizing machine measurement feedback, probably talk on the grass or selected networks. They're all hinge on some kind off our ability to control your variables in arbitrary, high precision and a certain networks you want toe read out across frequencies in case off CM's. You required identical and program because which is hard to keep, and they kind of fluctuate away from one another, shift away from one another. And if you control that, of course that you can control the performance. So actually one can ask if whether or not this is a universal bottleneck and it seems so aside, I will argue next. Um, we can recall a fundamental result by by showing harder in reaction Target from 1978. Who says that it's a purely computer science proof that if you are able toe, compute the addition multiplication division off riel variables with infinite precision, then you could solve any complete problems in polynomial time. It doesn't actually proposals all where he just chose mathematically that this would be the case. Now, of course, in Real warned, you have also precision. So the next question is, how does that affect the competition about problems? This is what you're after. Lots of precision means information also, or entropy production. Eso what you're really looking at the relationship between hardness and cost of computing off a problem. Uh, and according to Sean Hagar, there's this left branch which in principle could be polynomial time. But the question whether or not this is achievable that is not achievable, but something more cheerful. That's on the right hand side. There's always going to be some information loss, so mental degeneration that could keep you away from possibly from point normal time. So this is what we like to understand, and this information laws the source off. This is not just always I will argue, uh, in any physical system, but it's also off algorithm nature, so that is a questionable area or approach. But China gets results. Security theoretical. No, actual solar is proposed. So we can ask, you know, just theoretically get out off. Curiosity would in principle be such soldiers because it is not proposing a soldier with such properties. In principle, if if you want to look mathematically precisely what the solar does would have the right properties on, I argue. Yes, I don't have a mathematical proof, but I have some arguments that that would be the case. And this is the case for actually our city there solver that if you could calculate its trajectory in a loss this way, then it would be, uh, would solve epic complete problems in polynomial continuous time. Now, as a matter of fact, this a bit more difficult question, because time in all these can be re scared however you want. So what? Burns says that you actually have to measure the length of the trajectory, which is a new variant off the dynamical system or property dynamical system, not off its parameters ization. And we did that. So Suba Corral, my student did that first, improving on the stiffness off the problem off the integrations, using implicit solvers and some smart tricks such that you actually are closer to the actual trajectory and using the same approach. You know what fraction off problems you can solve? We did not give the length of the trajectory. You find that it is putting on nearly scaling the problem sites we have putting on your skin complexity. That means that our solar is both Polly length and, as it is, defined it also poorly time analog solver. But if you look at as a discreet algorithm, if you measure the discrete steps on a digital machine, it is an exponential solver. And the reason is because off all these stiffness, every integrator has tow truck it digitizing truncate the equations, and what it has to do is to keep the integration between the so called stability region for for that scheme, and you have to keep this product within a grimace of Jacoby in and the step size read in this region. If you use explicit methods. You want to stay within this region? Uh, but what happens that some off the Eigen values grow fast for Steve problems, and then you're you're forced to reduce that t so the product stays in this bonded domain, which means that now you have to you're forced to take smaller and smaller times, So you're you're freezing out the integration and what I will show you. That's the case. Now you can move to increase its soldiers, which is which is a tree. In this case, you have to make domain is actually on the outside. But what happens in this case is some of the Eigen values of the Jacobean, also, for six systems, start to move to zero. As they're moving to zero, they're going to enter this instability region, so your soul is going to try to keep it out, so it's going to increase the data T. But if you increase that to increase the truncation hours, so you get randomized, uh, in the large search space, so it's it's really not, uh, not going to work out. Now, one can sort off introduce a theory or language to discuss computational and are computational complexity, using the language from dynamical systems theory. But basically I I don't have time to go into this, but you have for heart problems. Security object the chaotic satellite Ouch! In the middle of the search space somewhere, and that dictates how the dynamics happens and variant properties off the dynamics. Of course, off that saddle is what the targets performance and many things, so a new, important measure that we find that it's also helpful in describing thesis. Another complexity is the so called called Makarov, or metric entropy and basically what this does in an intuitive A eyes, uh, to describe the rate at which the uncertainty containing the insignificant digits off a trajectory in the back, the flow towards the significant ones as you lose information because off arrows being, uh grown or are developed in tow. Larger errors in an exponential at an exponential rate because you have positively up north spawning. But this is an in variant property. It's the property of the set of all. This is not how you compute them, and it's really the interesting create off accuracy philosopher dynamical system. A zay said that you have in such a high dimensional that I'm consistent were positive and negatively upon of exponents. Aziz Many The total is the dimension of space and user dimension, the number off unstable manifold dimensions and as Saddam was stable, manifold direction. And there's an interesting and I think, important passion, equality, equality called the passion, equality that connect the information theoretic aspect the rate off information loss with the geometric rate of which trajectory separate minus kappa, which is the escape rate that I already talked about. Now one can actually prove a simple theorems like back off the envelope calculation. The idea here is that you know the rate at which the largest rated, which closely started trajectory separate from one another. So now you can say that, uh, that is fine, as long as my trajectory finds the solution before the projective separate too quickly. In that case, I can have the hope that if I start from some region off the face base, several close early started trajectories, they kind of go into the same solution orphaned and and that's that's That's this upper bound of this limit, and it is really showing that it has to be. It's an exponentially small number. What? It depends on the end dependence off the exponents right here, which combines information loss rate and the social time performance. So these, if this exponents here or that has a large independence or river linear independence, then you then you really have to start, uh, trajectories exponentially closer to one another in orderto end up in the same order. So this is sort off like the direction that you're going in tow, and this formulation is applicable toe all dynamical systems, uh, deterministic dynamical systems. And I think we can We can expand this further because, uh, there is, ah, way off getting the expression for the escaped rate in terms off n the number of variables from cycle expansions that I don't have time to talk about. What? It's kind of like a program that you can try toe pursuit, and this is it. So the conclusions I think of self explanatory I think there is a lot of future in in, uh, in an allo. Continue start computing. Um, they can be efficient by orders of magnitude and digital ones in solving empty heart problems because, first of all, many of the systems you like the phone line and bottleneck. There's parallelism involved, and and you can also have a large spectrum or continues time, time dynamical algorithms than discrete ones. And you know. But we also have to be mindful off. What are the possibility of what are the limits? And 11 open question is very important. Open question is, you know, what are these limits? Is there some kind off no go theory? And that tells you that you can never perform better than this limit or that limit? And I think that's that's the exciting part toe to derive thes thes this levian 10.

>>Hi everyone, This system A fellow from the University of Tokyo before I thought that would like to thank you she and all the stuff of entity for the invitation and the organization of this online meeting and also would like to say that it has been very exciting to see the growth of this new film lab. And I'm happy to share with you today or some of the recent works that have been done either by me or by character of Hong Kong Noise Group indicating the title of my talk is a neuro more fic in silica simulator for the commenters in machine. And here is the outline I would like to make the case that the simulation in digital Tektronix of the CME can be useful for the better understanding or improving its function principles by new job introducing some ideas from neural networks. This is what I will discuss in the first part and then I will show some proof of concept of the game in performance that can be obtained using dissimulation in the second part and the production of the performance that can be achieved using a very large chaos simulator in the third part and finally talk about future plans. So first, let me start by comparing recently proposed izing machines using this table there is adapted from a recent natural tronics paper from the Village Back hard People. And this comparison shows that there's always a trade off between energy efficiency, speed and scalability that depends on the physical implementation. So in red, here are the limitation of each of the servers hardware on, Interestingly, the F p G, a based systems such as a producer, digital, another uh Toshiba purification machine, or a recently proposed restricted Bozeman machine, FPD eight, by a group in Berkeley. They offer a good compromise between speed and scalability. And this is why, despite the unique advantage that some of these older hardware have trust as the currency proposition influx you beat or the energy efficiency off memory sisters uh P. J. O are still an attractive platform for building large theorizing machines in the near future. The reason for the good performance of Refugee A is not so much that they operate at the high frequency. No, there are particle in use, efficient, but rather that the physical wiring off its elements can be reconfigured in a way that limits the funding human bottleneck, larger, funny and phenols and the long propagation video information within the system in this respect, the f. D. A s. They are interesting from the perspective, off the physics off complex systems, but then the physics of the actions on the photos. So to put the performance of these various hardware and perspective, we can look at the competition of bringing the brain the brain complete, using billions of neurons using only 20 watts of power and operates. It's a very theoretically slow, if we can see. And so this impressive characteristic, they motivate us to try to investigate. What kind of new inspired principles be useful for designing better izing machines? The idea of this research project in the future collaboration it's to temporary alleviates the limitations that are intrinsic to the realization of an optical cortex in machine shown in the top panel here. By designing a large care simulator in silicone in the bottom here that can be used for suggesting the better organization principles of the CIA and this talk, I will talk about three neuro inspired principles that are the symmetry of connections, neural dynamics. Orphan, chaotic because of symmetry, is interconnectivity. The infrastructure. No neck talks are not composed of the reputation of always the same types of non environments of the neurons, but there is a local structure that is repeated. So here's a schematic of the micro column in the cortex. And lastly, the Iraqi co organization of connectivity connectivity is organizing a tree structure in the brain. So here you see a representation of the Iraqi and organization of the monkey cerebral cortex. So how can these principles we used to improve the performance of the icing machines? And it's in sequence stimulation. So, first about the two of principles of the estimate Trian Rico structure. We know that the classical approximation of the Cortes in machine, which is a growing toe the rate based on your networks. So in the case of the icing machines, uh, the okay, Scott approximation can be obtained using the trump active in your position, for example, so the times of both of the system they are, they can be described by the following ordinary differential equations on in which, in case of see, I am the X, I represent the in phase component of one GOP Oh, Theo F represents the monitor optical parts, the district optical parametric amplification and some of the good I JoJo extra represent the coupling, which is done in the case of the measure of feedback cooking cm using oh, more than detection and refugee A then injection off the cooking time and eso this dynamics in both cases of CME in your networks, they can be written as the grand set of a potential function V, and this written here, and this potential functionally includes the rising Maccagnan. So this is why it's natural to use this type of, uh, dynamics to solve the icing problem in which the Omega I J or the Eyes in coping and the H is the extension of the rising and attorney in India and expect so. >>Not that this potential function can only be defined if the Omega I j. R. A. Symmetric. So the well known problem of >>this approach is that this potential function V that we obtain is very non convicts at low temperature, and also one strategy is to gradually deformed this landscape, using so many in process. But there is no theorem. Unfortunately, that granted convergence to the global minimum of there's even 20 and using this approach. And so this is >>why we propose toe introduce a macro structure the system or where one analog spin or one D o. P. O is replaced by a pair off one and knock spin and one error on cutting. Viable. And the addition of this chemical structure introduces a symmetry in the system, which in terms induces chaotic dynamics, a chaotic search rather than a >>learning process for searching for the ground state of the icing. Every 20 >>within this massacre structure the role of the ER variable eyes to control the amplitude off the analog spins to force the amplitude of the expense toe, become equal to certain target amplitude. A Andi. This is known by moderating the strength off the icing complaints or see the the error variable e I multiply the icing complain here in the dynamics off UH, D o p o on Then the dynamics. The whole dynamics described by this coupled equations because the e I do not necessarily take away the same value for the different, I think introduces a >>symmetry in the system, which in turn creates chaotic dynamics, which I'm showing here for solving certain current size off, um, escape problem, Uh, in which the exiled from here in the i r. From here and the value of the icing energy is shown in the bottom plots. And you see this Celtics search that visit various local minima of the as Newtonian and eventually finds the local minima Um, >>it can be shown that this modulation off the target opportunity can be used to destabilize all the local minima off the icing hamiltonian so that we're gonna do not get stuck in any of them. On more over the other types of attractors, I can eventually appear, such as the limits of contractors or quality contractors. They can also be destabilized using a moderation of the target amplitude. And so we have proposed in the past two different motivation of the target constitute the first one is a moderation that ensure the 100 >>reproduction rate of the system to become positive on this forbids the creation of any non tree retractors. And but in this work I will talk about another modulation or Uresti moderation, which is given here that works, uh, as well as this first, uh, moderation, but is easy to be implemented on refugee. >>So this couple of the question that represent the current the stimulation of the cortex in machine with some error correction, they can be implemented especially efficiently on an F B G. And here I show the time that it takes to simulate three system and eso in red. You see, at the time that it takes to simulate the X, I term the EI term, the dot product and the rising everything. Yet for a system with 500 spins analog Spain's equivalent to 500 g. O. P. S. So in f b d a. The nonlinear dynamics which, according to the digital optical Parametric amplification that the Opa off the CME can be computed in only 13 clock cycles at 300 yards. So which corresponds to about 0.1 microseconds. And this is Toby, uh, compared to what can be achieved in the measurements tobacco cm in which, if we want to get 500 timer chip Xia Pios with the one she got repetition rate through the obstacle nine narrative. Uh, then way would require 0.5 microseconds toe do this so the submission in F B J can be at least as fast as, ah one gear repression to replicate the post phaser CIA. Um, then the DOT product that appears in this differential equation can be completed in 43 clock cycles. That's to say, one microseconds at 15 years. So I pieced for pouring sizes that are larger than 500 speeds. The dot product becomes clearly the bottleneck, and this can be seen by looking at the the skating off the time the numbers of clock cycles a text to compute either the non in your optical parts, all the dog products, respect to the problem size. And and if we had a new infinite amount of resources and PGA to simulate the dynamics, then the non in optical post can could be done in the old one. On the mattress Vector product could be done in the low carrot off, located off scales as a low carrot off end and while the kite off end. Because computing the dot product involves the summing, all the terms in the products, which is done by a nephew, Jay by another tree, which heights scares a logarithmic any with the size of the system. But this is in the case if we had an infinite amount of resources on the LPGA food but for dealing for larger problems off more than 100 spins, usually we need to decompose the metrics into ah smaller blocks with the block side that are not you here. And then the scaling becomes funny non inner parts linear in the and over you and for the products in the end of you square eso typically for low NF pdf cheap P a. You know you the block size off this matrix is typically about 100. So clearly way want to make you as large as possible in order to maintain this scanning in a log event for the numbers of clock cycles needed to compute the product rather than this and square that occurs if we decompose the metrics into smaller blocks. But the difficulty in, uh, having this larger blocks eyes that having another tree very large Haider tree introduces a large finding and finance and long distance started path within the refugee. So the solution to get higher performance for a simulator of the contest in machine eyes to get rid of this bottleneck for the dot product. By increasing the size of this at the tree and this can be done by organizing Yeah, click the extra co components within the F p G A in order which is shown here in this right panel here in order to minimize the finding finance of the system and to minimize the long distance that the path in the in the fpt So I'm not going to the details of how this is implemented the PGA. But just to give you a new idea off why the Iraqi Yahiko organization off the system becomes extremely important toe get good performance for simulator organizing mission. So instead of instead of getting into the details of the mpg implementation, I would like to give some few benchmark results off this simulator, uh, off the that that was used as a proof of concept for this idea which is can be found in this archive paper here and here. I should result for solving escape problems, free connected person, randomly person minus one, spin last problems and we sure, as we use as a metric the numbers >>of the mattress Victor products since it's the bottleneck of the computation, uh, to get the optimal solution of this escape problem with Nina successful BT against the problem size here and and in red here there's propose F B J implementation and in ah blue is the numbers of retrospective product that are necessary for the C. I am without error correction to solve this escape programs and in green here for noisy means in an evening which is, uh, behavior. It's similar to the car testing machine >>and security. You see that the scaling off the numbers of metrics victor product necessary to solve this problem scales with a better exponents than this other approaches. So so So that's interesting feature of the system and next we can see what is the real time to solution. To solve this, SK instances eso in the last six years, the time institution in seconds >>to find a grand state of risk. Instances remain answers is possibility for different state of the art hardware. So in red is the F B G. A presentation proposing this paper and then the other curve represent ah, brick, a local search in in orange and center dining in purple, for example, and So you see that the scaring off this purpose simulator is is rather good and that for larger politicizes, we can get orders of magnitude faster than the state of the other approaches. >>Moreover, the relatively good scanning off the time to search in respect to problem size uh, they indicate that the FBT implementation would be faster than risk Other recently proposed izing machine, such as the Hope you know network implemented on Memory Sisters. That is very fast for small problem size in blue here, which is very fast for small problem size. But which scanning is not good on the same thing for the >>restricted Bosman machine implemented a PGA proposed by some group in Brooklyn recently again, which is very fast for small promise sizes. But which canning is bad So that, uh, this worse than the purpose approach so that we can expect that for promise sizes larger than, let's say, 1000 spins. The purpose, of course, would be the faster one. >>Let me jump toe this other slide and another confirmation that the scheme scales well that you can find the maximum cut values off benchmark sets. The G sets better cut values that have been previously found by any other >>algorithms. So they are the best known could values to best of our knowledge. And, um, or so which is shown in this paper table here in particular, the instances, Uh, 14 and 15 of this G set can be We can find better converse than previously >>known, and we can find this can vary is 100 times >>faster than the state of the art algorithm and cp to do this which is a recount. Kasich, it s not that getting this a good result on the G sets, they do not require ah, particular hard tuning of the parameters. So the tuning issuing here is very simple. It it just depends on the degree off connectivity within each graph. And so this good results on the set indicate that the proposed approach would be a good not only at solving escape problems in this problems, but all the types off graph sizing problems on Mexican province in communities. >>So given that the performance off the design depends on the height of this other tree, we can try to maximize the height of this other tree on a large F p g A onda and carefully routing the trickle components within the P G A. And and we can draw some projections of what type of performance we can achieve in >>the near future based on the, uh, implementation that we are currently working. So here you see projection for the time to solution way, then next property for solving this escape problems respect to the prime assize. And here, compared to different with such publicizing machines, particularly the digital and, you know, free to is shown in the green here, the green >>line without that's and, uh and we should two different, uh, prosthesis for this productions either that the time to solution scales as exponential off n or that >>the time of social skills as expression of square root off. So it seems according to the data, that time solution scares more as an expression of square root of and also we can be sure >>on this and this production showed that we probably can solve Prime Escape Program of Science 2000 spins to find the rial ground state of this problem with 99 success ability in about 10 seconds, which is much faster than all the other proposed approaches. So one of the future plans for this current is in machine simulator. So the first thing is that we would like to make dissimulation closer to the rial, uh, GOP or optical system in particular for a first step to get closer to the system of a measurement back. See, I am. And to do this, what is, uh, simulate Herbal on the p a is this quantum, uh, condoms Goshen model that is proposed described in this paper and proposed by people in the in the Entity group. And so the idea of this model is that instead of having the very simple or these and have shown previously, it includes paired all these that take into account out on me the mean off the awesome leverage off the, uh, European face component, but also their violence s so that we can take into account more quantum effects off the g o p. O, such as the squeezing. And then we plan toe, make the simulator open access for the members to run their instances on the system. There will be a first version in September that will >>be just based on the simple common line access for the simulator and in which will have just a classical approximation of the system. We don't know Sturm, binary weights and Museum in >>term, but then will propose a second version that would extend the current arising machine to Iraq off eight f p g. A. In which we will add the more refined models truncated bigger in the bottom question model that just talked about on the supports in which he valued waits for the rising problems and support the cement. So we will announce >>later when this is available, and Farah is working hard to get the first version available sometime in September. Thank you all, and we'll be happy to answer any questions that you have.

>>Hello. Nice to meet you. My name is yes. Um Escuela. I'm a professor in a university in Japan. So today I want to introduce my research. That title is a really time emotional detection using e g with riel time knowing the reduction. First of all, I want to introduce myself. My major is system identification and signal processing for large removed and by American signal process for owner off them. A common technique. It's most magical. Modern by you creation using this opportunity identification method. So today topic it's e easy modern by the Barriers Council with heavy notes. We call this technique the concept moody. Now what is a concept? I mean, the concept is Japanese world because studies are first in Japan. So consider is similar to emotion and sensibility, but quite different. The commercial nous sensibility is innate ability. The concert is acquired after birth, so concept is similar to how to be So we focus on this can see using the brain signals. As for the brain Sina, there is ah, many way to know the brain. For example, the optical leading X c T m i m e g e g optical topography um, function and my by using these devices, we have three areas off research, for example, like neural engineering area for obligation, including new market neuroscience area for understanding the mechanism a medically oil area for treatment. So but it's very important to use, depending on the purpose. So what did they can be obtained? Uh, in the case of e g, we can see the activity of neurons that scalp the case of in years so we can attain the river off oxygen bar Pratt The case off natural and safe Alagem we can see the activity of new uh, that contact is neck case off position. Martian topography. We can get activity off reception by the contact list. If we use that, I we can measure the amount of blood by the contractors. These devices are showing these figures. So our motivation is to get the concept question using their model by system identification where it's not removed on. The second motivation is to theorize that's simple and small cancer X election using the each information when we use the ever my the large scale and the expensive on binding. So it is unuseful. So we focus on the EEG because the e g iss Moscow inexpensive a non binding on to use. So we focus on the energy. So e g is actually a potential from the major from the scalp that detective data is translated to the pregnancy domain. And if you can see domain that their point to 44. We call the data death of it 4 to 6. We called a cedar with on 17. 14 were called the Alfa Hour and 14 to 26. We called a better work in a conventional method we want if we want use the cats a deep sleep, we use that death of it in a case of light sleep we used a secretive and so but this is just only the sensible method. So we cannot use that for all the film Actuary accuracies under the 20%. So we need to define the situation original. So recall this technique council modeling. So these are the block diagram Kansi the concept What? So this field this part eyes for the noise, this part for the mathematical model. So we calculate this transfer function like this. This is a discrete time water, and, uh, this time, uh, is continuous time model. So then we really right this part Thio Discrete time water. So we cull Create, uh, this part us like this This'll first part on the second part is calculated by the party application so we can get this the argumentative model. So then that we were right this part by using that the transfer function transport formation. So we right this argument ID model like this. So the off about the inverse and better off the inverse is the point as this equation. So each the coefficient is corrugated by this equation on. But then we calculate a way too busy with beaver by using this because of a least squares algorithm. So we call this identification method the self joining identification method. Um, that this is an example of stories modeling. The first of all, we decide we gather the data like a story. It's moving. So we move the small beans, try to trade at 41 hour. So last 10 minutes we used as stories and we measure that culture soul for sliced levin Onda. We associate the egg and we measure the 8000 data. Uh, in 17 years we? Yeah, that's a 17 years. So in the case, off the simple, easy universes that there are many simply devices in the world like this so many of them the There we calculate the signal nodes. Lazio, The signal means the medical easy system on the each device made it sn Lazio. And we investigate 58 kinds off devices on almost off All devices are noise devices. So I'm also asked about to various parts more device that best. So my answer is anything. Our skill is, you know, processing on def. With love. Data can be obtained from the device. No, but what device? He may use the same result commission. Our novelty is level Signal processing on our system is structured by 17 years Data for one situation. So the my answer is what? Anything. So we applied this system to Arial product. We call this product concern Analyzer. In a concept analyzer, you can see the concept that right the our time a concept dinner influence Solis sickness concentration on like so that we combine that this can't say analyzer And the camera system We made the euro system your account so pretty show it this is in Eureka. Well, this is, uh, e g system and we can get can say by using the iPhone on the, uh, we combine the camera system by the iPhone camera and if the cancer is higher than the 6% 60% so automatically recorded like this. Mhm. So every time we wear the e g devices, we can see the no awareness, the constant way. That's so finally we combine the each off cancer. So like that this movie, so we can see the thes one days. Can't say the movie s Oh, this is a miracle. On the next example, it's neuro marketing using a constant analyzer. So this is a but we don't know what is the number one point. So then we analyze the deeds CME by using concert analyzer so we can get the rial time concept then that we can see the one by one situation like this. So this is the interest level and we can see the high interest like this. So the recorded a moment automatically on the next one is really application. The productive design. Ah, >>Japanese professor has come up with a new technology she claims can read minds, she says. The brainwave analysis system will help businesses better understand their customers, needs workers at a major restaurant chain or testing a menu item that is being developed. This device measures brain waves from the frontal lobes of people who try the product. An application analyzes five feelings how much they like something and their interest, concentration, stress and sleepiness. >>The >>new menu item is a cheese souffle topped with kiwi, orange and other fruit. The APP checks the reaction of a person who sees the souffle for the first time. Please open your eyes. When she sees the souffle, the like and interest feelings surge on the ground. This proves the desert is visually appealing. Now please try it. After the first bite, the like level goes up to 60. That shows she likes how the dessert tastes. After another bite, the like level reaches 80. She really enjoys the taste of the souffle. It scores high in terms of both looks and taste, but there's an unexpected problem. When she tries to scoop up the fruit, the stress level soars to 90. I didn't know where to put the spoon. I felt it was a little difficult to eat. It turned out it was difficult to scoop up the fruit with a small spoon. So people at the restaurant chain are thinking of serving this a flavor with a fork instead. Green well. How could be the difference with the device? We can measure emotional changes in minute detail in real time. This is a printing and design firm in Tokyo. >>It >>designs direct mail and credit card application forms. The company is using the brainwave analyzing system to improve the layout of its products. The idea is to make them easier to read during this test, The subject wears an eye tracking device to record where she's looking. In addition to the brainwave analyzing device, her eye movements are shown by the red dots on the screen. Stress levels are indicated on the graph on the left. Please fill out the form. This is a credit card application form. Right after she turns her eyes to this section, her stress levels shoots up. It was difficult to read as each line contained 60 characters, so they decided to divide the section in two, cutting the length of the lines by half 15 a Hong Kong. This system is very useful for us. We can offer differentiated service to our clients by providing science based solutions. The brain wave analyzed. >>Okay, uh, now the we construct a concert detection like this. Like this. Like concentration, interest sickness stories contain, like comfortable, uncomfortable. I'm present the rats emotion, deadly addictive case lighting, comfort, satisfaction and the achievement. So finally we conquer more presentation. So in this presentation, we introduce the our such we construct the council question Onda we demonstrate that c street signal processing and we apply the proposed method to Arial product. Uh, we named the constant riser. So this is the first in the world, that's all. Thank you so much.

>>from around the globe. It's the Cube covering HP. Discover Virtual experience Brought to you by HP. >>Hello, everyone. Welcome to the Cube's coverage we're covering HP Discover Virtual experience. 2020. I'm John Furrier, host of the Cube. Great online experience. Check it out. A lot of content go poke around a lot of Cube interviews. A lot of content from HP. It's their virtual conference. HP Discover virtual experience. We have Cube alumni Justin Hotard, who's now s VP and general manager of HP Japan. Justin, great to see you virtually here for the virtual experience. How you doing >>Doing well, John. Great to see you again. A swell and really glad to be here. >>You know, just reminiscing about our previous interview a couple times. You know Jeff Frick is interviewed. I've interviewed HP Discover a couple years ago. Um, service provider Edge now is booming. Everyone's working at home. Everyone is seeing the global pandemic play out on a global stage and impacting our lives. But anyone in the in the I T. Business or technology business is seeing the massive gaps and the areas that need to be worked on. This is something that we're gonna dig into it, I think is really interesting conversation as someone who's in Japan. Honestly, Big telco presence, but also part of the global stage. So I want to get into that. But before we do, tell us about your new role at HP. What are you working on and what are you doing? >>Yes. So, John, currently, I'm the president of HP Japan. I'm responsible is the managing director of Japan and also the managing managing director. Our business in China as well. So keeping myself busy these days. >>A pack your own a lot of zoom calls, conference calls, could imagine the work. You're doing pretty big disruptions. I want to get your thoughts as an industry participant and who's seen these ways before. What is some of the disruptions that you're seeing right now? I see there will document in terms of VM or video, um, VPNs under proficient. Where are you seeing the big disruption? Because those are the obvious low hanging fruit. But it's certainly being an impact. The disruptions or creating opportunities, but major challenges right now. What's your thoughts? >>You >>know, I think I think specific and, uh John and we're seeing in Japan, and a big pillar is, you know, this is really a big inflection point in terms of how people work, and as you as you know, you think about Japan. The culture and the economy has been very reliant on face to face in relation, relationship driven. It's also there's been some traditional paper based activity in that space, as well as things like the Hong Kong stamp away. You sign documents to get you're not just for government approval, but even in private transactions. So all of that is actually under a great way to change. And so the obvious part is, we talk about virtualization and VD I It's really forcing people to rethink, um, you know, work flows and it's not, you know, it's not just one thing. Generally, it's across many, many parts. Education, manufacturing, obviously, obviously traditional enterprise. You touched on Zoom and other virtualization and beady eye, but it's it's I think it's coming across all industries right now. Based on this change, >>what's going on in Japan? Specifically, I know that some GDP numbers were coming in pre covert. I'll see when Covic it's given some of the things you were just talking about how they do business. The culture there must be impacted by the covert 19. What do you what you're seeing there, and how do they move forward? What is some of the changes that need to happen? What do you see? >>Yeah, I mean, I think you touched on. I think the economy that was already under pressure. Um, then you have Cove. It hit. Um, you know, Japan has a huge has had a huge tourism business booming based on the growth in Asia and obviously particularly in China, all of that gets hit. And, uh huh. And then, obviously, you know, the traditional way of doing business has been challenged over the past few months, but it's actually creating quite a bit of opportunity. And some of it is some of it is similar to what you see in other parts of the world. But, you know, we've seen many of the Japanese companies and medical devices and pharmaceuticals jump into innovation and everything from masks toe, um, you know, investment in, you know, in virology and other and, you know, in other areas and testing and all the things that you see, but beyond that we're also seeing is a lot, a lot more discussion around innovation. One place that we're seeing it immediately is education. There's a huge initiative around connecting uh, schools, primary schools, great schools and bringing technology into those schools is a way to accelerate the learning experience. I think obviously in this in this new world in the short term help manage on and ensure continuity of learning through through social distancing and some of the challenges that and everybody has, you know, in in primary education. >>It's interesting, you know, those traditional things like you mentioned just signatures converting at the digitally signatures of the stamping thing you mentioned. Also, the face to face with education, every vertical up is going to be disrupted and an opportunity. So that's what you guys see. That transformation is part of that. What are some of the patterns you see emerging so that your customers and prospects can capture it? What is some of the highlights? What's the big picture? >>Yeah, I think I think at a high level we talk a lot about digital transformation and remote work. These, by the way, were discussed before Covic hit, so I think it's It's just an acceleration. The other one is really around edge, and I ot, um Japan. Obviously great tradition of manufacturing this actually is gonna probably create new investment around manufacturing. Is Japan looks to build its manufacturing base is part of what we expect from the government stimulus programs out there. Um, but they're investing in. And I don't think the factory that will be built tomorrow is gonna is going to start off with a traditional labour view. In fact, it's going to start very, very organized against robotics AI using using i O. T. Using sensors to drive greater levels of automation. A lot of that exists today, but I think this this event just creates more opportunities for acceleration, particularly Greenfield. So we're having conversations with customers around all those areas right now. >>You know, one of the biggest observations I would say in the past 10 years, looking at the wave we've been on and looking at the massive wave coming in now is culture is always a part of the blocker of adoption, and you're kind of getting at some of this with the world you're in now, >>where >>the culture has to shift pretty radically fast. Whether it's the remote workforce, the remote workplace, workloads with robotics and AI everything work related workplace workloads, workflow was with the work. We're forced. I mean, always changing, right? So this is a critical cultural thing. Your thoughts on this because this has to move faster. What are you seeing as catalysts? Any kind of technology? Enablement. What's the What's the What's the data tell you? >>Yeah, yeah, I think I think a couple of things were, you know, we're seeing I think, one that we're seeing that given that we've obviously seen in the rest of the world for a number of years now is a is a shift, that consumption. And we've seen that grow from customers, right? So they're looking at How do we accelerate this experience, how they stand it up? How did they get it? Running and consumption as a service, you know, as a service, models are becoming even more attractive, and so we're seeing new interest in that as a way to build things, to scale things, to create flexibility for future growth. And it's not, you know, it's not just public cloud, it's it's public cloud and on premise applications. It's integration into the virtualization stack, obviously, with, um, you know, with players like VM Ware and Nutanix and Red Hat, it's ah, you know, with open shift containers. It's bringing all of that, you know, bringing all of that scale and flexibility and the other good place. Honestly, we're still seeing it is even in some of our traditional businesses, and we had a very large consumption model in a traditional transaction processing business and for that customer was about creating the flexibility for growth. Um, and so I think we're you know, I think we really are on the brink of a very different I t model in, you know, certainly in Japan to enable a lot of this innovation and to provide more more flexibility and more automation for, you know, for companies there in the businesses. >>And I just want to just validate that by seeing the day that we're looking at in the interviews we've had and even our internal conversation with our editorial Cuban research teams is, is it's happening now in the change you can't ignore it. You could ignore in the past were not ready for it. People process technology. Three pillars of transformation with Cove ID and we've seven, which is having this debate with our team this past month where it's not so much an acceleration in the future. The future got pulled to today, and people are now seeing it and saying, Wow, I need to move because the consequences of not changing are obvious. It's not like a hypothetical. You're starting to see specific use cases where the folks that under invested or didn't make the right bets might be on the wrong side of history coming out of covitz. So to your point about growth is a really key point. This >>is what >>everyone is thinking about right now. So I got to ask you, what solutions do you guys have ready to help customers? Because right now, solutions Walk are really all that matters. It walks that fine line between making it and not making it's having the right solutions is key. >>Yeah, and actually, you know, I think one of things you mentioned a great example of what you're talking about in transformation right in the airline industry. You know, we're seeing that we're going to see this in in Japan, right? This is a place where based if a service was considered a premium experience where you go to kiosks and automation. But now I think we're going to see now we're seeing already interested complete and an automation right bag check bag drop. And that stuff's been talked about for many years. But now it's an acceleration of the experience, and the difference is going to be no longer is it going to be a premium to talk to someone? It's actually about speed. So that's a place where, you know, obviously that's a heavily impacted industry. But as we see it come back in Japan and probably throughout Asia, I think we're gonna see a very different model. And to your question on, uh, you know, to your question on technologies, when I see us doing is really kind of three pieces I think you've got You've got solutions like VD. I were literally out of the box and we built a partners so that customers that are small, medium or large that wants something standard that they could just take into it quickly. We have a platform for also things like SD wan to our business, and we're seeing significant growth there, obviously, you know, mobile access, wireless access, Another place where we're seeing demand, just building on our core business and really seeing healthy growth. I mentioned education is one vertical, but we're seeing it in, obviously in places like manufacturing and on. I'm expecting this even more broken enterprise there as this customer, Aziz, many of our customers come back to the office and bring employees back in. And you can't. You can't have a traditional, you know, just density of desks, right? You've really got to think about how people have mobility and have flexibility to make being distancing and and even even kind of the in and out of office, right? How do I mean by that? That work experience in the productivity, whether I'm in the office for a couple days and how so? I think those are places where we see the technology. Then we talk about consumption service. So the flexibility consume it as a service which in all of those solutions we have offers around and then ultimately even a pop it out or hp fs our financial services, giving customers flexibility and payment options, which for many people that are cash strapped solves a real challenge, right? We talk a lot about the technology but fundamental business challenge of saying yes, I want to invest today. I need to get my work, my workforce up in productive with beady eye. But so they can start generating revenue and cash flow, but one of the cash flow to invest in that productivity. And so this becomes a place where, you know, we're just seeing a lot of traction with our customers. We can help them actually get that up and running, not not created huge cash flow outlay upfront and making get productive and get back on their feet. And definitely in the mid market and the smaller businesses, we're seeing a lot of a lot of activity there. >>That's a huge point, because right now, more than ever, that need is there because of the financial hardships that we're seeing that's evident and well reported. Having that financial flexibilities primary, that's a key thing. So that's great. So good to hear that. The second thing I want to ask you on the business side that's important is not just a financing because you want to have that consumption buy as you go from a cloud technology like standpoint as a service. But now you've got the financial support check. Next step is ecosystem. What are you guys doing on the ecosystem side? If I'm trying to rebuild my business or have a growth strategy check technology check. I'm gonna get some business help on the finance side. Third is partners. What's the status there? >>Yeah, yeah, I think there's I think there's a couple things. One is there's obviously the global relationships we have, you know, close relationship with VM Ware. You know that Nutanix relationship red hat, others that were standing up solutions that some of things I mentioned like me. I literally packaged out of the box experience with a complete turnkey solution, right? So so our partners don't even have to. You don't have to optimize that they can. They can just deploy and enable their their customers. I think the other place in Japan, it's you know what? We didn't touch on it earlier, but one of the really important things and is most of our customers depend on their vendors, depend on their partners, actually do a lot of their I t work. It's a little bit unique in Japan versus the rest of the world. And so this is a place to We're spending a lot of time with our partners with our entire partner ecosystem to make sure they're ready. And I was just actually in a conversation yesterday with a partner talking about the investments they're bringing their they're putting in to really bring that that core innovation around, um around beady eye and around around SD win for as an example and working with them to make sure that they've got all the tools they need from us so that what they can deliver into their into their ecosystem is very turnkey and easy. And I think I think that's really, really, really important. So it's not just the, you know, the global technology relationships that we talked about certainly in Japan, it's also about it's about stitching together. That entire ecosystem that, you know that allows the the end customer toe have ah have a turnkey experience and everybody that's involved in that delivery, you know, to have to have a seamless experience to get these customers up and running. >>And it's great to you guys had that foundational services, but also now with some great acquisitions. You got the cloud native experience across environments and then the reality of the edge Actually, work force in workplaces are changing. VD I etcetera. But you've got edge exploding. You guys also made a great has been years of investments and edge. So with telco and WiFi, all kind of coming together kind of sets up for a nice kind of front end piece with the APP development piece going on. You're seeing that in Japan as well. >>Yeah, I think all of our major telcos there have you have announced five G projects projects and launch is we've got a new you know, we've got a new entrant in the telco space Pakatan launched just a couple months ago. Therefore G solution. But I think all of that is very favorable to driving greater levels of connectivity. And I think you know, it's a lot of times we talk about five G. We talk about kind of the next mobile hands when we think about the next mobile device or handset. But it's also a lot of the private lt and connectivity, and I think we'll see that actually, the intersection of five G and WiFi. In some cases, we're having conversations about, you know, are there opportunities in five G and as the back whole and actually using WiFi in a smaller medium sized office home? And so there's a number of things like that that I think will be compelling and great opportunities for growth, because Japan's an incredibly A. So you know, John is incredibly well connected society and a lot of connectivity, but but I think this is also creating new demand. I mean, people weren't working at home all the time and way. Obviously, you see that in other countries where maybe media streaming and video conferencing we're working on the plans where people got their original Internet service. I think in Japan that's even more so because this tradition, if I go to the office at work and I know when I'm home, I'm relaxing. I mean, this is fundamentally under a huge shift right now, and so I think it's gonna be a you know, a really significant wave of growth and five g n and wife by as this this new. Imagine this new, this new remote work experience this new mobile work experience happens a >>lot of architecture to really work a little bit. Not radical, but certainly transforming. And its benefits. Exciting time, tough environment. Right now, let people working hard have to come out of it. But it's super exciting from a tech perspective. What it can enable. Really appreciate. Of course, we're here in the HP Discover virtual experience bringing you the best content. So I have to ask you, what sessions? Um, do you think people should turn into for the virtual experience? >>Well, you know, it's of course, the one that I think everyone has to make. And I never liked the missus is the keynote is that obviously Antonio always gives us not only, you know, some of the great technologies and launches, but but also really a vision of where we see the industry going to. I think Tom ones foundational. But we've got some great sessions on consumption and as a service that are actually set up for some of our customers and partners in Japan and across Asia. And I think those will be really good discussions, you know, with, uh, you know, with folks like our CTO commercial coffee and our our global general manager for green like white. So I'd encourage folks to turn into, you know, to really learn about as a service because I think a lot of times we talk about the cloud and we think about public Cloud only. Um and I think for certainly for many of my customers and partners in Japan, um, I think with everything we just talked about, the cloud is gonna be an inevitable reality. But the cloud is an architecture, and that's where some of these new technologies and services that we're bringing out will be will be really, really valuable, whether it's in storage or it's in compute virtualization, enabling collaboration or some things that we're doing right now, John. But be a video video conference, but but also also even just in automating the data center and bringing, you know, being a new levels of productivity back into some of the traditional data center. A swee as we need to do that in order to enable the new edge and some of these new applications around AI and machine learning that are necessary, Teoh to support the growth of the economy. But you know net net. I think this is going to be. These are all things they're going to support growth and recovery. So I think it's a great opportunity and discover for our customers and partners to learn what they could do to help accelerate that and and and accelerate the recovery. >>Certainly, Cloud has shown the way it's operating model. It's not just public, it's on premise. It's an edge is so it's not just multi cloud either. It's multi environment. This is where the market's going. So you guys are on the right track. Justin really appreciate the time. But I want to ask the final question. I want you to complete this sentence for me as we end this out on our virtual experience, Our competitive advantage HP HP is competitive advantage to our clients is that we are blank. >>Our competitive advantage is that we are the best partner, deeply understanding their needs and bringing them the right innovation and value that they need to deliver their business outcomes and in this case, obviously recover and get back to growth. >>There's a whole chart. Managing director of President of HP Japan great to see you. Congratulations on your new role over there on Asia Pacific. Um, and thanks for checking in on the virtual experience. Thanks for coming in. And good to see you again. >>Great. Great to see you, John. Thanks again for having time for me. And best of luck for a successful discover virtual experience. >>Awesome. Okay, I'm John Furry here in the Cube studios, getting the remote injuries for this virtual experience for HP Discover. Thanks for watching. >>Yeah. Yeah, yeah, yeah.

>> Announcer: Live from Las Vegas, it's theCUBE covering Magento Imagine 2019, brought to you by Adobe. >> Welcome to the theCUBE, I'm Lisa Martin at The Wynn, in Las Vegas for Magento Imagine 2019. This is a three day event. You can hear a lot of exciting folks networking behind me, talking tech, talking e-commerce innovation and we're pleased to welcome fresh off the keynote stage a couple of guests from HP. We've got Gillian Campbell, the Head of Omni-channel Strategy and Operations. Gillian, thank you for joining us. >> Thank you for asking us. >> Our pleasure and Herriot Stobo, Director of Omni-channel Innovation and Solutions, also from HP. Welcome. >> Thank you very much. >> So Gillian fresh off the keynote stage, enjoyed your presentation this morning. >> Gillian: Thank you. >> Everybody I think in the world knows HP. Those of us consumers going, you know what actually, that reminds me, I need a new printer. >> We can help you. >> Thank you, excellent. Whether I'm shopping online or in a store. So you gave this really interesting keynote this morning talking about what HP is doing, starting at Apache. You really transform this shopping experience. Talk to us a little bit about HP, as I think you've mentioned it as a $50 billion start up and from a digital experience perspective, what you needed to enable. >> Yeah, so as I said, HP have been around for 80 years and in 2015, we became our own entity, HP Inc., and really started looking at how do we enable digital to be pervasive through everything that we do. Our internal processes are reached to customers and identified a great opportunity to really take leading edge and our digital commerce capabilities and we already had some early proof points and APG so we launched a global initiative and we're now on that journey to enable that best in class experience through the digital platforms. >> So Herriot talk to us about, you're based in Singapore. >> Yes. >> What were some of the market dynamics that really made it obvious that this is where we want to start building out this omni-channel strategy starting in Apache? Is it, you know whether, Gillian you mentioned it before. We started retail spaces, some being expensive. Is it more mobile experience and expectations on consumer's part? >> I think we've got a mix of different starting points across Asia. We've got some mega cities like Hong Kong and Singapore rising, Tokyo. And then we've got you know emerging markets across South-East Asia. We don't necessarily have any single market place that controls the entire market as we might see in other regions and so we've had a lot of runway to go and experiment and try new things. We also have an ecosystem of branded retail in Asia. Not in all markets, predominantly in India but also in some markets in South-East Asia that allow us to really blend the experience across both offline and online and to give customers choice at the end of the day. Let them decide how they want to shop and interact with our brand. So we have been running Magento 1 since we first launched our online store businesses in Indonesia and Thailand about six years ago and then we moved into China, replatformed, lexi-platform onto Magento 1 and then that was really the foundation of what we decided to go and build upon to become a global program. so we already had some proof points under our belt with Magento so. >> And what were some of those early wins that really started to make this really obvious that this omni-channel experience, the ability to give customers choice? Whether they want to start the process online, finish it in store, vice verse, or at least have the opportunity to have a choice? What were some of those early wins and business outcomes that you started to see? >> I think even just from because we're all, customers are people. Whether you're a corporate customer, a small business, or a consumer, we're all people and we all know that we shop that way. So essentially the storyline on that back to HP was we have to enable experiences that we would want to experience as well and it was quite a shift for a tech company who were really all about the products to be thinking about, well, how do we really enable that end to end experience? And as Herriot said, the runway was open. We already had some proof points. I was new in the job so I was like all listening to, you know, what the team were telling me. We have a great opportunity here and took that formered as a new concept for the company. We got funding approval and you know the rest is the history and the journey that we're on. So I think it was just taking a different perspective and a different approach and working with a team who already had the, built some of that credibility and others proof points with the earlier deployments and I think we kind of took a risk at the time when we started the engagement with Magento. They weren't in that leadership quadrant and we took a risk to say, let's partner with an energizing company and do something a little bit different and we're still here working towards it so I think that for me was the breakthrough, was just having the tenacity to say, we're gonna drive this path forward. It may not be how we would have done things in the past, but we're a different company now. and we had much more thinner air cover to be able to do that. >> Little bit more agility and flexibility. >> Yeah, absolutely. So you guys, you talked about, Gillian about all the buyers. We are the consumers and we have this expectation, growing expectation that I want to be able to get any and transact anything that I want to buy, whether I'm a procuring person for a company and I'm traveling but I need to approve expenses or I'm a salesperson maybe sitting next to a medium-small business customer. I need to have the option at least to have this store front. What are the things that you guys launched in Apache, leverage be the power of Magento Commerce was click to collect. So tell me a little bit about from maybe an e-commerce cultural perspective, what is it that makes people want to have the ability to start online and actually complete the transaction in a physical location? >> Essentially I was in the Advisory Board yesterday and one of the other customers of Magento said, "Until we can invent a way to touch and feel online, "there's always gonna be a need to have, "outlets where you can go touch and feel." and I think with the click and collect, some of our products are, you know, high-end PCs and gaming devices and printers that is hard to get a good appreciation of what it looks and feels like online. So if you're gonna be spending you know, a significant money you may want to go in and be able to see the colors, feel the finish. You know some of our newer products with the leather portfolios is not something you can truly appreciate without touching it. So I think we have to enable again those customers who do want to experience, feel the weight, you know feel the finish, see the color scheme 'cause its usually important, again not for all customers. Some customers are quite happy to spend thousands of dollars on an online purchase without seeing it and then making sure they have a good facility to be able to, well if they wanted to, to return if they got the normal the product. >> As we look though at like we talked about, this consumerization of everything where we have this expectation and the numbers, I think you even mentioned it maybe in your keynote, Gillian, the numbers of, or somebody did this morning, like upwards of half of all transactions are starting on mobile so we got to start there. What are some of the things that you guys have seen in region in terms of mobile conversions? >> So there's still a massive gap between desktop and mobile conversions, first of all. I mean we're not anywhere near parity between the two. But obviously we're seeing a huge volume of traffic coming in as well and it's shifting that way, so you would expect it to drop as result. I think with Magento what we've seen over the, you know, past few deployments that we've been running and that were over 8% improven. But the desktop conversions are far higher. I mean in terms of improvement and actual conversion so we've still got a long way to go. There and that's a naturative process, that's a journey that probably never ends in terms of ongoing optimization and experimentation. So yeah a lot happening there. I think just on the click and collect topic as well that you were asking about people wanting to start their journey online and then come into bricks and mortar. We're seeing a huge uptake on it just by experimenting, by piloting. Over 26% of our consumer notebooks in India that we've put onto this program were being collected in store and this is in environments which are inherently chaotic on the streets. You don't want to go out there but actually I'm passing that way anyway so it's just easier for me to pick it up on the way home and probably quicker 'cause I can collect in two hours. So it's just giving people customer choice, no additional incentive and it seems to take. So now we're expanding out regionally. >> So you said there's, this morning, Gillian, in your keynote eight markets covered, mostly Apache, but also in Latin America. >> We just started in Latin America, again, the development process is not just as simple as we're switching on. So we've been doing a lot of work for this past six months with Latin America. The team there, they're super excited to get launched. There's some differences there, we've talked about the regional variation around fulfillment models that we have to adapt towards but the intent is to get Latin America deployed, leveraging some of the layer lengths from what we've done in Asia specific and then starting to move around into more the near region and then ultimately back into the US and Canada. >> So as you look forward and of course you've mentioned we're on this journey right, what are some of the key learnings that you're going to apply? You mentioned this morning, something that was very intriguing and that was, respect the integrity of the Magento platform. Talk about that in context of some of the other learnings that you'd recommend for colleagues and similar or other industries to be able to achieve what you have on a global scale. >> I think from the outset, there was this kind of like baggage of deployments of capabilities not just in commerce but deployment of capabilities across HP that we had not respected the integrity of the platform. We had adjusted the code and developed on the code to make it HP specific and with the new HP Inc. company one of the guided principles was no, when we buy the leverage software applications respect it for what it is and adjust business processes and adjust integration rather than adjust the core so that we can get the advantage of the longer term opportunity without creating such like. So it was really just a foundational, you know, let's not go in here with a mindset that we know better than the core. The core is there for a reason and then build around that and ensure the integration and I think you know with Herriot's leadership, we've been able to you know, just keep that firm is why we can be successful and be successful longer term as well. So that all the, and one of the things we talked about yesterday also is the excellent capabilities that are coming with Adobe and the integration that we talked about the recommendation of Adobe Sensei and integrate that with Magento Core. If you don't keep to the respect the integrity, those upgrades and capabilities become really hard to take benefit of so we're really excited about, you know, again, sticking with the core and enabling and growing with the core with Magento and Adobe. >> I would just build on it, I mean I think its never gonna be easy running a global commerce platform. Single instance, multiple countries, you know, 27 markets to get started with. Who knows where we're gonna end. Its always gonna be a challenge so we have to keep it as simple as possible. These upgrades are fast and furious and that's great and we all gets lots of benefit but if we start going down our own path, we've lost it. We've lost the benefit. >> And that's one of the things too that Jason Wolfsteen said this morning was that the word Magento was gonna be enabling businesses to achieve without getting in their way and it kind of sounds Herriot, like you're saying the same thing. That we've gotta be able to respect the technologies that we're building so we don't get in our own way and we keep it simple as we wanna expand globally. Ultimately at the end of the day, you're creating these personalized experiences with consumers and that personalization is so important because it's more and more not only are we transacting or wanting to on mobile but we want our brands like HP to know us. We want you to know our brand value, you know our average order value so that we can become part of the experience but also ideally get rewarded for being loyal. >> Yeah. >> Yeah, I mean, I mean just coming to mobile again but you know, 2.3 delivers the native PWA capabilities which we're super excited to get started with. You know we've got so many used cases for this straight away, right out the box but you know we've got to do it gradually, do it the right way. I think we're also aware that we're not gonna be able to run with PWA in all markets straight away 'cause not all markets are ready for it quite frankly. User behavior- >> Is that a cultural thing? >> It's purely cultural. Maybe technical and just technical ecosystems as well. Places like China in particular, where, you know, customers use app stores but they use app stores from every single phone manufacturer right there. That's where the customer is. We can't just move away from that so we need to keep some of those legacy approaches for a little while and then yeah test in other regions and then take the learnings when we're ready to adopt it. >> Exciting so here we are at, this is the first Magento Imagine since the Adobe acquisition. Gillian, let's wrap things up with you. What are your, you mentioned you were part of the Customer Advisory Board yesterday, just some of your perspectives on this years' event now that Magento is powering the Adobe commerce cloud. >> I actually attended the Adobe Summit a few weeks ago here also in Vegas and started to see the thread of commerce coming into that conference and then seeing the Adobe, the experience, coming into Magento and I just think it's a perfect combination of opportunities especially for a company like HP where we were linked in to connect, you know, marketing and sales and support across the customer journey and the capabilities with Adobe and some of the marketing stack, and then the commerce stack, and there was support bringing that together is a super exciting opportunity for us. You know the partnership that we have with both Adobe and Magento again as one as I really, they were just starting what the next journey was gonna look like. >> We feel that about so many things, we're just starting, but Gillian, Herriot, it's been a pleasure to have you on theCUBE for Magento Imagine 2019. Thank you both for your time. >> Thank you, thank you. >> Our pleasure. I'm Lisa Martin and you're watching theCUBE live from The Wynn Las Vegas at Magento Imagine 2019. Thanks for watching. (light music)

>> Live from Las Vegas, it's theCUBE! Covering Dell Technologies World 2019. Brought to you by Dell Technologies and its ecosystem partners. >> Hello everyone. Welcome back to theCUBE's live coverage here in Las Vegas for Dell Technologies World. I'm John Furrier with Dave Vellante. Dave, we've got Pat Gelsinger back on theCUBE. He stopped by yesterday, did a flyby after his keynote to kick off our intro section. He's back for the sit-down. >> (laughs) Welcome back. >> I can't get enough of you, Pat. >> CEO of VMware, Pat Gelsinger. >> Yeah, I love to photobomb you guys, so it was great. >> Anytime. I know you're super busy, business is going great. And you know, what a three years its been. I remember the keynote you gave at VMworld a few years ago. This was really on a time where, I would call it the seminal moment for you because you saw a vision, and we've talked privately and on theCUBE about, and you gave this speech of this is going to be the preferred future, and it was very visionary-oriented, but it ended up happening. That became the beginning of a run for VMware. And since then, you've been kind of chipping away and filling in all the tech pieces, the business model, and deals, with Amazon and now Azure and others. How are you feeling about it? What's the highlights? What's your perspective of where we are now? What's the notable accomplishments? >> Well you know, it's been just great. And you think about the run that we've been on where we, five years ago, we described a hybrid future. And you know, most people said, what are you, stupid? And you know, student body right to the public cloud. And now everybody is starting to understand the difficulty of replatforming, right? And says wow, this is really hard. I can spend millions and millions of dollars, in fact, one customer's estimate was that they were going to spend almost $1 billion replatforming all their applications to the cloud. And when they got them cloud-native, what do they have? The same apps. So imagine going to your board and saying I'm going to spend $1 billion just so I can be on the cloud, but give you no new business value. You've got to be kidding! And that's why this hybrid future, and as I like to joke, Andy, five years ago, Andy Jassy said if you're running your own data center, you're stupid. And Pat said if you're using Amazon, you're stupid. And now we're doing bro hugs on stage with each other. (laughter) >> And by the way, hybrid, you picked that trend that was right. Multi-cloud, though, came out of more a reality, less of an operating vision, 'cause hybrid cloud, you know, you saw the dots, connected those dots, but I think multi-cloud was much more of just a reality. When people started to realize that as I started doing stuff on premises, wow, I got native workloads on the cloud, and there are benefits for being in the cloud first for certain workloads. But then the multi-cloud thing comes up. >> And I think everybody has started to realize, and I really, as I would say, I think every CIO needs a three-cloud strategy. Making their private data centers into a proper operating private cloud. And some of this week's announcements, I'm sure we'll get back to those a little bit, to me are just a huge dimension. You know, VMware Cloud on Dell EMC, you know, a huge accelerant of making your private data center op like a private cloud, right, at scale. Second, you need a primary public cloud partner. And I think most people should pick a primary. Not one, a primary, and then a secondary cloud, right, you know, as their partners. And then you have your range of SAS offerings. And I think that needs to be the core, right, of every IT, CIO's strategy for the future. And our objective is to create an environment between what we're doing with VMware Cloud Foundation, and now VMware Cloud on Dimension. What we're doing with Amazon, our preferred partner for the public cloud offering. What we announced this week with Azure, right? Our 4000 other cloud partners, including, you know, very successful relationship with IBM. And saying, okay, that's your infrastructure. And the bulk of your workloads should run on a VMware environment that we can operate across that, with the same tools, the same interfaces, the same security, the same management tools, and then use the other cloud services as they bring you business value. You're a fan of Tensorflow? Go for it, baby. Right? You know, and use it in your app. You love function as a service with Lambda, go for it. But the bulk of your workload should lay in here and use these where they have business value. >> And to follow up on the three legs of the cloud stool, the CIO's legs, number three is for what? Is it for risk mitigation, exit strategies, or more specific best-of-breed, horses-for-courses type of workloads. >> Yes, yes, and yes. To some degree, really it's saying, nobody wants to say, I'm only in one. Right? Nobody wants to lock in for it. Also you know, clearly, hey, you know, these are technologies that break. You get more resilience that way, right? You want to be able to manage your cost environments. There's clearly this view of okay, you know, if I can do one, two, and three, I can do N. 'Cause most people are also going to end up picking, oh, I'm in Hong Kong. Okay, I need a Hong Kong cloud, because my data can only go there. You know, I'm in Malaysia, oh, they require all data to be there. 'Cause a practicality, if you're a big enterprise company, it's not just going to be three. You're going to need to be four, five, and six as well, for regional. And then you're going to acquire somebody, they're using a different partner. It really says, build an operational environment that works that way. Give myself business flexibility. I have application flexibility, and if I've done that, I really can move to the other environments that my business requires. >> I think one of the reasons why you guys have been so successful, if I go back five or six years, I remember you laying out the market, the market segmentation, you're obviously close to customers. You're a very clear thinker. You've obviously looked at the market for multi-cloud. How do you describe that, how do you look at the TAM, how big is it? >> Well you know, if you think about cloud today, right, we're closing in on $100 billion of the public cloud. You add SAS to it, you know, you got almost another $100 billion at that level. And you know, the overall data center market is probably on the order of, you know, $1 trillion-ish. >> Give or take. (laughs) >> Yeah, on that order. And then you know, you throw the operations costs inside of it, you're probably looking at something that's, you know, on the order of $2 trillion as well. So this is a big market, right? You know, part of the excitement that people are seeing in this cloud environment, is that they can just go faster. And as I described in the keynote today, we want to enable every one of our customers to stop looking down and look up, right? Spend less time looking down at the infrastructure. We're going to operationalize it, we're going to automate it for you, we're going to take care of it so that every one of your engineers can become software engineers building app and business value. >> I want to ask you on that point, because one of the things, I was talkin' last night, the analyst said at the briefing or the reception was, having a debate with one of the strategists in Dell, and I'm like, look it, outcomes are great at the top of the stack. Looking up, you want outcomes. But during the OSI stack days, no one cared about outcomes. It was either token ring or Ethernet. Speed won, so certain things have to be speed-driven, world-class, and keep getting better. And so that's what we're seeing as an infrastructure requirement. Horizontal scalability, operational scale. So that's a speeds and feeds game. So the outcome there is faster (laughs), and simpler. Up the stack, data becomes a big part of that. That, more, is where we see outcome. Do you see it that way, Pat? Because you know, again, infrastructure is often, that's how they said it on stage. We want to have whole new-paved, new infrastructure for this generation, essentially a refresh of infrastructure. Okay. Well, what does it look like? It's got to be fast, got to be flexible, software-defined. Your thoughts? >> So you know, clearly, I mean, what we're trying to do is we build this common infrastructure layer. And build an environment that allows you to be fast, but also allows you to be in control and cost-effective. Because if you would say, oh, I just want to be fast, ah, that doesn't work, right? We still have limited budgets, and you know, people, someday there's a CFO day of reckoning. But you also have to realize, part of the hybrid cloud laws that I described this morning, you know, one of those is the laws of physics, right? Hey, my factory automation for robotics needs to be 40 milliseconds, period. And if I round-trip to the cloud at 150 milliseconds, guess what? (laughs) >> Latency. >> Right. You know, my image recognition for being able to detect my autonomous vehicle is less than 50 milliseconds. I can't round-trip to the cloud. It has to be fast, right, but we also need to be able to push more of this data, more of the inference of my machine learning and AI closer to the edge. That's why, you know, you heard Michael talk about, and Jeff talk about this explosion of data. Most of that data will be at the edge. Why? Because every camera, you know, every sensor will be developing it, and I'm not going to round-trip it to the cloud because of economics. I can't afford to take all that data to the cloud. It's not just the latency. >> Latency matters. >> Yeah. And so for that, so I can't take it to the cloud, I got to be able to compute locally. I got to be able to apply the inference of my AI models locally, but you know, I also then need to scale aspects of cloud as well. My third law, of course, was regulation, where you know, guess what? I was just with a major customer in Latin America, and they said they are repatriating 100% of their data and applications out of the public cloud, 'cause the new president, right, is assisting on data only in his country for all of their nationalized resources and assets. >> So that's driving the change. This brings up the multi-cloud kind of thing earlier. You guys got to play in all the ponds out there, in the industry. But let's talk about on-stage here at Dell Technologies World. You were on-stage with Michael Dell and Satya Nadella, and I was lookin' up there. I'm like, man, the generational knowledge of the three people on-stage, the history. >> (laughs) I think that just means I'm getting old. (laughs) >> Well I mean, you've seen it all. I mean, from Intel, to EMC, to VMware. Dave and I, Dave's a historian of tech, as he'll self-claims, but I'm up there, I was pretty blown away. You guys are leading the industry. What kind of moment was that for you, because now you've got Microsoft doing a deal with VMware. Who would've thought that would happen? >> Well, maybe two different aspects to it. You know, one is, I've known Satya for over 25 years. You know, he was sort of going through the Microsoft ranks, Windows NT, SQL, et cetera. (laughter) You know, at the same time I was. So we got to know each other. Almost 25 years since our first interactions. When Michael Dell first came to Intel to meet Andy Grove to get microprocessors so he could start his business, I was there. So I mean, these relationships are decades old. So in that view, it's sort of like, hey Satya, how's the wife, you know. (laughter) Hey Michael, how's Susan doing? Really, it-- >> But you haven't even gone anywhere, you're still in the industry. (laughs) >> Yeah. But then to be able, the announcement was really pretty special in the sense that I call it 20 years in the making. You know, not a year or two, 20 years in the making, 'cause VMware and Microsoft has essentially been at odds with each other for two decades. You know, at that level. And to be able to be on-stage and saying, that's right, we're cooperating on cloud, we're cooperating on client, and we're cooperating on futures, okay, that's a pretty big statement as well. And I think customers respond very positively to that. And you know, I'm-- >> It's been a bold move, and you also made a bold move with the cloud, too, Pat. I got to say, that was another good call. Partnering with Andy Jassy. Again, once, both idiots, I guess, calling each other clever, you know. (laughs) Hey, public cloud, at odds, partner. Boom. >> And I really think this idea, moving headwinds to tailwinds. And you know, the Amazon partnership with Andy, and as we say, it's our preferred cloud partner, VMware Cloud, our native US hub, VMware-offered service. You know, super committed to it. We're closing in on 2000 customers on that now. >> Clarify the Amazon relation. I saw some press articles that kind of missed, skewed a little bit. They kind of made it sound like the Azure deal was similar to the Amazon deal. So just explain the difference between the VMware deal with AWS and Andy Jassy, that relationship, and the other cloud ones. Take a minute to explain that. >> Yeah, thank you. And what we're doing with Amazon is VMware is offering a cloud service that I operate for customers, that runs on Amazon. And that is a VMware-delivered service. They're our preferred partner. We're not bashful about that, that if we have the choice, that's the one to go to. It's going to be best. But what we've done now with Azure is we've made the VMware Cloud Foundation, the same underlying components, available with CloudSimple and Virtustream, they're partners, to have a VMware Cloud Foundation offering delivered by Microsoft as a first-party service. So VMware Cloud, VMware is delivering it. In the Azure for VMware services, that's being delivered and supported by Microsoft. >> And that's the same deal you did with IBM. >> It's very, the same-- >> Google and other ones. >> Yeah, the same as we've done with our 4000 other cloud partners, right? And obviously, Virtustream and CloudSimple are part of that 4000, and they're making the VMware Cloud Foundation available to Azure customers now. >> And what's the benefits to VMware's customers for those deals? >> Well, imagine that you're somebody in, Walmart was quoted in the press release, as an example. Walmart's a big VMware customer. Walmart is also a big Azure customer. So their ability to say, oh, I can have a hybrid environment makes a lot of sense for that kind of customer. So we really do see it as saying, you know-- >> Customer-driven, basically. >> Absolutely. And people said, which are you going to sell to us? Well in most cases, customers have already decided who their major cloud partners our. We're saying that VMware offering, even though we're first and best with Amazon, we're saying as they make their cloud choices, we'll have a valid VMware Cloud Foundation offering available. >> And best, I want to understand best. Best is, in part, anyway, because of the engineering you guys have done. When we interviewed Andy Jassy in November at re:Invent, he said you can't have a lot of these types of partnerships. And it's very deep integration. Is that why it's best? And what makes it best? >> Yeah, I call it first and best for two reasons. One is because we are engineering, we are co-engineering, the bits first get done on VMware Cloud, and then we make 'em available to the other partners. That's where we're doing the core engineering, the innovation. Andy has hundreds of engineers working on this. I have hundreds of engineers working on it. So it's first and best from an engineering sense. And, given it's my service and my offering, we're selling it aggressively in the marketplace, positioning it as part of the broader set of solutions and leveraging that, like you saw this week with the Dell EMC offering, VMware Cloud on Dell EMC. It's leveraging all that first and best work to now bring it on-premise as well. So it really is both the engineering as as a go-to-market. >> I'm going to ask some CEO questions. (laughs) So Tom Sweet has said they're happy to have the Class V transaction behind them. I'm sure you're glad, too. Thank you. That was very generous of you. >> (laughs) >> You've been incredibly good at acquisitions. I mean, obviously Nicira, Heptio, CloudHealth, AirWatch, I mean, on and on. >> VeloCloud. >> VeloCloud. I mean, most acquisitions, frankly, don't live up to their objectives. I think that's not the case for VMware. So now you're, good news is you draw off a lot of cash, so you're building up that pot again. How do you see, going forward, use of that cash? R and D, M and A, maybe you could make some comments there to the extent you can? >> Yeah, and you know, we said the primary ways we use cash, stock buybacks and M and A. And that continues. We did the special one-time dividend, which helped Dell go public. Everybody's happy. The market's responded super positively on both the Dell side. They're up, what, 40% since they go public. VMware up almost 50% this year. Just tremendous. >> Tremendous, $80 billion value now, awesome. >> Yeah, just tremendous. And, right then, we said going forward, it's business as usual for us. We're going to continue to do stock buybacks. We're going to continue to do M and A's. As you've said, we're good at this acquisitions stuff. And part of that is, I call it, imagine you're a hot startup company. And you say, do I want to be part of VMware? And we try to answer these questions. Do we have vision alignment? >> (laughs) >> Second is, can we accelerate your vision? Because most startups, you know, I mean, you talk about unicorns and so on like that. But what really motivates them is their vision. And if they believe their vision is going to be accelerated as part of VMware, so they're on this and we're going to turn 'em to that, aw man, they get excited. Do we have a cultural fit? I mean, with every CEO of our acquisitions, and HR does, we really, are they going to fit our team? Because you know, cultural issues, you can't butt your heads day and night. Life's too short. >> Certainly VMware, you guys are (laughs) that culture's very hardcore. Work hard, play hard. (laughter) >> Yeah, and you know, it has to be this deep drive for technical innovation, right? The technical due diligence that we do with our startups. Right? It's sort of like, you know, this is like a PhD exam for these, I mean, they really got to know their stuff. >> Yeah, so people don't fit in the culture at VMware, and there-- >> And we've said no to a number of potential acquisitions over cultural issues as well, if they're just not going to fit. And hey, we're not going to be perfect, but the fact that we can bring these companies in, accelerate their vision, give 'em a culture that they're excited about. You know, we have maybe 90-ish% success rate. The industry average is below 50% >> Yeah, fantastic track record. I mean-- >> And that just gives us the ability to do organic and inorganic innovation, which to me is like, a potent recipe. >> And you got the radio conference coming up. What will your talk, theCUBE will be there. Pat, you've created great shareholder value. You turned those headwinds into tailwinds, and we were watchin' the whole time. It's been great to watch. And what's next? You have your VMware tattoo still on from VMworld? (laughter) Like you have a jail tattoo? >> No, I'll tell you >> Cute tattoo. >> a little inside, I'll tell you a little inside story. My wife, you know, after the VMworld keynote with the tattoo on, we were leavin' on vacation two weeks later. And all she said to me after the keynote was what's that tattoo thing, it better be gone by the time we leave for vacation. (laughter) It's like, there was no, honey, that was a great keynote today, it's like, that better be gone! (laughs) >> Nothin's better than watchin' that video and that CUBE sticker we had on your hand. Pat, great to see you, as always. Great commentary, great analysis. Congratulations on all the success with VMware. Again, the transformation's just getting started. We're seeing a lot more good things for you guys as well. >> Yeah, and you know, this has been a great week in some ways. I sort of joked this morning on-stage that, it almost felt like VMworld. We talked about VMware technologies and that Dell partnership accelerating so well. >> It's not AMCWorld, it's DellWorld now, it's a whole new vibe. >> (laughs) And you know, with that, you know, I just really believe in the superpowers that I talk about, we're just getting started. So we're going to be doing this a long time together. >> What's on your plate in front of you now? You got VMworld coming up in a few months. Priorities, objectives, what's on your plate? >> Well, I have to leave some of the secrets for what we're cookin' up for VMworld this year. But some of these steps clearly, in the developer container space, super important for us to really make some progress there. Obviously, we'll have some incremental cloud announcements as well. >> ContainerWare rhymes with VMware. (laughs) >> Yes, that's very good! We have an advertisement on that coming out, so a new ad. But it really is, I think, that topic area's one that, how can we really solve that for customers that really can deploy at scale containerized environments for an enterprise workload. So, excited about that area. And you know, maybe just a few deliverables from what we announced this week. >> Alright, take your CEO of VMware hat off, put your CUBE analyst hat on. What's the most important story here at Dell Technologies World, if you were a commentator? You can't say VMware 'cause that's biased, but you got to be objective. You can say VMware if an objective. What's the most important storyline here as a backdrop for Dell Technology Worlds, what's the real net net to customers? >> Well you know, I think, and I'll say, as exciting as the Microsoft announcements were, I think the most important thing was VMware Cloud on Dell EMC, on-prem. Because to me, you know, the fact, I go to CIOs, and I've done this probably five times since the keynote finished on Monday. And I say, how many of you have fully updated your hardware, your firmware, your operating systems, your networking stack, your compute stack, your management on the latest releases, all of them patched, upgraded appropriately for your environment? >> And they say, their eyes roll. (laughs) >> And the answer is none. Not some, none. I have customers that are askin' me to extend support for vSphere 4.5. It's like, what, that's been EOL'ed for a year and a half, what are you talking about, right?! But the reality is that most people go to the cloud, public cloud, not because it's more cost-effective or because it's better, it's because it's easier. So what we've really said is we can make easy in the private cloud and truly deliver that hybrid cloud experience. And I think the customers really experience the TCO benefits, the acceleration, the reductions in their operational environments, the personnel associated with it, the security benefits of being always patched, upgraded the most release. You know, now you're talkin' about attacking that other $1 trillion of operational costs that they're bearing in the personnel and so on. To me, that is like, so powerful if we really get that engine going. >> And the simplicity that comes out of that, is just-- >> You know, and again, the demo that we showed. That was the VMware Cloud on AWS being able to demonstrate, now, a complete picture into the on-premise environment. That's powerful. >> Pat Gelsinger, CEO of VMware. I know he's got to go. Thanks for your generous time, I know you're really busy. Again, Pat Gelsinger. >> Love you guys, thank you. >> Thanks, Pat. >> Love you too. Pat Gelsinger, CEO of VMware, creating a lot of shareholder values, got a lot of tailwinds at their back. VMworld's coming up, theCUBE, of course, will be there with two sets. As usual, theCUBE cannons, two sets here, firing cannonballs of content here at Dell Technology World. I'm Jeff Furrier with Dave Vellante, stay with us for more after this short break. (electronic music)