(bright music) >> Hey welcome back, everybody. Jeff Frick here with theCUBE We are high atop San Francisco in the Sales Force Tower in the new Accenture offices, it's really beautiful and as part of that, they have their San Francisco Innovation Hubs. So it's five floors of maker's labs, and 3D printing, and all kinds of test facilities and best practices, innovation theater, and this studio which is really fun to be at. So we're talking about hybrid cloud and the development of cloud and multi-cloud and continuing on this path. Not only are customers on this path, but everyone is kind of on this path as things kind of evolve and transform. We are excited to have a couple of experts in the field we've got Larry Socher, he's the Global Managing Director of Intelligent Cloud Infrastructure Services growth and strategy at Accenture. Larry, great to see you again. >> Great to be here, Jeff. And Ajay Patel, he's the Senior Vice President and General Manager at Cloud Provider Software Business Unit at VMWare and a theCUBE alumni as well. >> Excited to be here, thank you for inviting me. >> So, first off, how do you like the digs up here? >> Beautiful place, and the fact we're part of the innovation team, thank you for that. >> So let's just dive into it. So a lot of crazy stuff happening in the marketplace. Lot of conversations about hybrid cloud, multi-cloud, different cloud, public cloud, movement of back and forth from cloud. Just want to get your perspective today. You guys have been in the middle of this for a while. Where are we in this kind of evolution? Everybody's still kind of feeling themselves out, is it, we're kind of past the first inning so now things are settling down? How do you kind of view the evolution of this market? >> Great question and I think Pat does a really nice job of defining the two definitions. What's hybrid versus multi? And simply put, we look at hybrid as when you have consistent infrastructure. It's the same infrastructure regardless of location. Multi is when you have disparate infrastructure, but are using them in a collective. So just from a from a level setting perspective, the taxonomy is starting to get standardized. Industry is starting to recognize hybrid is the reality. It's not a step in the long journey. It is an operating model that going to exist for a long time. So it's not about location. It's about how do you operate in a multi-cloud and a hybrid cloud world. And together at Accenture VMware have a unique opportunity. Also, the technology provider, Accenture, as a top leader in helping customers figure out where best to land their workload in this hybrid, multi-cloud world. Because workloads are driving decisions. >> Jeff: Right. >> We are going to be in this hybrid, multi-cloud world for many years to come. >> Do I need another layer of abstraction? 'Cause I probably have some stuff that's in hybrid and I probably have some stuff in multi, right? 'Cause those are probably not mutually exclusive, either. >> We talked a lot about this, Larry and I were chatting as well about this. And the reality is the reason you choose a specific cloud, is for those native differentiator capability. So abstraction should be just enough so you can make workloads portable. To be able to use the capability as natively as possible. And by fact that we now at VMware have a native VMware running on every major hyperscaler and on pram, gives you that flexibility you want of not having to abstract away the goodness of the cloud while having a common and consistent infrastructure while tapping into the innovations that the public cloud brings. So, it is the evolution of what we've been doing together from a private cloud perspective to extend that beyond the data center, to really make it an operating model that's independent of location. >> Right, so Larry, I'm curious your perspective when you work with customers, how do you help them frame this? I mean I always feel so sorry for corporate CIAOs. I mean they got security going on like crazy, they go GDPR now I think, right? The California regs that'll probably go national. They have so many things to be worried about. They go to keep up on the latest technology, what's happening in containers. I thought it was doc, now you tell me it's Kubernetes. It's really tough. So how do you help them kind of, put a wrapper around it? >> It's got to start with the application. I mean you look at cloud, you look at infrastructure more broadly I mean. It's there to serve the applications and it's the applications that really drive business value. So I think the starting point has to be application led. So we start off, we have our intelligent engineering guys, our platform guys, who really come in and look and do an application modernization strategy. So they'll do an assessment, you know, most of our clients given their scale and complexity usually have from 500 to 20,000 applications. You know, very large estates. And you got to start to figure out okay what's my current applications? A lot of times they'll use the six Rs methodology and they say hey okay what is it? I'm going to retire this, I no longer need it. It no longer has business value. Or I'm going to replace this with SaaS. I move it to sales force for example, or service now, etcetera . Then they're going to start to look at their workloads and say okay, hey, do I need to re-fact of reformat this. Or re-host it. And one of the things obviously, VMware has done a fantastic job is allowing you to re-host it using their software to find data center, you know, in the hyperscaler's environment. >> We call it just, you know, migrate and then modernize. >> Yeah, exactly. But the modernized can't be missed. I think that's where a lot of times we see clients kind of get in the trap, hey, i'm just going to migrate and then figure it out. You need to start to have a modernization strategy and then, 'cause that's ultimately going to dictate your multi and your hybrid cloud approach, is how those apps evolve and you know the dispositions of those apps to figure out do they get replaced. What data sets need to be adjacent to each other? >> Right, so Ajay, you know we were there when Pat was with Andy and talking about VMware on AWS. And then, you know, Sanjay is showing up at everybody else's conference. He's at Google Cloud talking about VMware on Google Cloud. I'm sure there was a Microsoft show I probably missed you guys were probably there, too. You know, it's kind of interesting, right, from the outside looking in, you guys are not a public cloud, per se, and yet you've come up with this great strategy to give customers the options to adopt VMware in a public cloud and then now we're seeing where even the public cloud providers are saying, "Here, stick this box in your data center". It's like this little piece of our cloud floating around in your data center. So talk about the evolution of the strategy, and kind of what you guys are thinking about 'cause you know you are clearly in a leadership position making a lot of interesting acquisitions. How are you guys see this evolving and how are you placing your bets? >> You know Pat has been always consistent about this and any strategy. Whether it's any cloud or any device. Any workload, if you will, or application. And as we started to think about it, one of the big things we focused on was meeting the customer where he was at in his journey. Depending on the customer, they may simply be trying to figure out working out to get on a data center. All the way, to how to drive an individual transformation effort. And a partner like Accenture, who has the breadth and depth and sometimes the vertical expertise and the insight. That's what customers are looking for. Help me figure out in my journey, first tell me where I'm at, where am I going, and how I make that happen. And what we've done in a clever way in many ways is, we've created the market. We've demonstrated that VMware is the only, consistent infrastructure that you can bet on and leverage the benefits of the private or public cloud. And I often say hybrid's a two-way street now. Which is they are bringing more and more hybrid cloud services on pram. And where is the on pram? It's now the edge. I was talking to the Accenture folks and they were saying the metro edge, right? So you're starting to see the workloads And I think you said almost 40 plus percent of future workloads are now going to be in the central cloud. >> Yeah, and actually there's an interesting stat out there. By 2022, seventy percent of data will be produced and processed outside the cloud. So I mean the edge is about to, as we are on the tipping point of IOT finally taking off beyond smart meters. We're going to see a huge amount of data proliferate out there. So the lines between between public and private have becoming so blurry. You can outpost, you look at, Antheos, Azure Stack for ages. And that's where I think VMware's strategy is coming to fruition. You know they've-- >> Sometimes it's great when you have a point of view and you stick with it against the conventional wisdom. And then all of a sudden everyone is following the herd and you are like, "This is great". >> By the way, Anjay hit on a point about the verticalization. Every one of our clients, different industries have very different paths there. And to the meaning that the customer where they're on their journey. I mean if you talk to a pharmaceutical, you know, GXP compliance, big private cloud, starting to dip their toes into public. You go to Mians and they've been very aggressive public. >> Or in manufacturing with Edge Cloud. >> Exactly. >> So it really varies by industry. >> And that's a very interesting area. Like if you look at all the OT environments of the manufacturing. We start to see a lot of end of life of environments. So what's that next generation of control systems going to run on? >> So that's interesting on the edge because and you've brought up networking a couple times while we've been talking as a potential gate, right, when one of them still in the gates, but we're seeing more and more. We were at a cool event, Churchill Club when they had psy links, micron, and arm talking about shifting more of the compute and store on these edge devices to accommodate, which you said, how much of that stuff can you do at the edge versus putting in? But what I think is interesting is, how are you going to manage that? There is a whole different level of management complexity when now you've got this different level of distributing computing. >> And security. >> And security. Times many, many thousands of these devices all over the place. >> You might have heard recent announcements from VMware around the Carbon Black acquisition. >> Yeah. >> That combined with our workspace one and the pulse IOT, we are now giving you the management framework whether it's for people, for things, or devices. And that consistent security on the client, tied with our network security with NSX all the way to the data center security. We're starting to look at what we call intrinsic security. How do we bake security into the platform and start solving these end to end? And have our partner, Accenture, help design these next generation application architectures, all distributed by design. Where do you put a fence? You could put a fence around your data center but your app is using service now and other SaaS services. So how do you set up an application boundary? And the security model around that? So it's really interesting times. >> You hear a lot about our partnership around software defined data center, around networking. With Villo and NSX. But we've actually been spending a lot of time with the IOT team and really looking and a lot of our vision aligns. Actually looking at they've been working with similar age in technology with Liota where, ultimately the edge computing for IOT is going to have to be containerized. Because you're going to need multiple modalware stacks, supporting different vertical applications. We were actually working with one mind where we started off doing video analytics for predictive maintenance on tires for tractors which are really expensive the shovels, et cetera. We started off pushing the data stream, the video stream, up into Azure but the network became a bottleneck. We couldn't get the modality. So we got a process there. They're now looking into autonomous vehicles which need eight megabits load latency band width sitting at the edge. Those two applications will need to co-exist and while we may have Azure Edge running in a container down doing the video analytics, if Caterpillar chooses Green Grass or Jasper, that's going to have to co-exist. So you're going to see the whole containerization that we are starting to see in the data center, is going to push out there. And the other side, Pulse, the management of the Edge, is going to be very difficult. >> I think the whole new frontier. >> Yeah absolutely. >> That's moving forward and with 5G IntelliCorp. They're trying to provide value added services. So what does that mean from an infrastructure perspective? >> Right, right. >> When do you stay on the 5G radio network versus jumping on a back line? When do you move data versus process on the edge? Those are all business decisions that need to be there into some framework. >> So you guys are going, we can go and go and go. But I want to follow up on your segway on containers. 'Cause containers is such an important part of this story and an enabler to this story. And you guys made and aggressive move with Hep TO. We've had Craig McLuckie on when he was still at Google and Dan, great guys. But it's kind of funny right? 'Cause three years ago, everyone was going to DockerCon right? That was like, we're all about shows. That was the hot show. Now Docker's kind of faded and Kubernetes is really taking off. Why, for people that aren't familiar with Kubernetes, they probably hear it at cocktail parties if they live in the Bay area. Why is containers such an important enabler and what's so special about Kubernetes specifically? >> Do you want to go on the general or? >> Why don't your start off? >> I brought my products stuff for sure. >> If you look at the world its getting much more dynamic. Particularly as you start to get more digitally decoupled applications, you're starting, we've come from a world where a virtual machine might have been up for months or years to all the sudden you have containers that are much more dynamic, allowed to scale quickly, and then they need to be orchestrated. And that's essentially what Kubernetes does, is really start to orchestrate that. And as we get more distributed workloads, you need to coordinate them. You need to be able to scale up as you need for performance etcetera So Kubernetes is an incredible technology that allows you really to optimize the placement of that. So just like the virtual machine changed how we compute, containers now gives us a much more flexible, portable, you can run on any infrastructure at any location. Closer to the data etcetera to do that. >> I think the bold move we made is, we finally, after working with customers and partners like Accenture, we have a very comprehensive strategy. We announced Project Tanzu at our last VM World. And Project Tanzu really focused on three aspects of containers, How do you build applications, which is what Pivotal and the acquisition of Pivotal was driven around. How do we run these on a robust enterprise class run time? And what if you could take every vSphere ESX out there and make it a container platform. Now we have half a million customers. 70 million VM's. All the sudden, that run time we are container enabling with a Project Pacific. So vSphere 7 becomes a common place for running containers and VMs. So that debate of VMs or containers? Done, gone. One place or just spend up containers and resources. And then the more important part is how do I manage this? As you have said. Becoming more of a platform, not just an orchestration technology. But a platform for how do I manage applications. Where I deploy them where it makes more sense. I've decoupled my application needs from the resources and Kubernetes is becoming that platform that allows me to portably. I'm the Java Weblogic guy, right? So this is like distributed Weblogic Java on steroids, running across clouds. So pretty exciting for a middleware guy, this is the next generation middleware. >> And to what you just said, that's the enabling infrastructure that will allow it to roll into future things like edge devices. >> Absolutely. >> You can manage an Edge client. You can literally-- >> the edge, yeah. 'Cause now you've got that connection. >> It's in the fabric that you are going to be able to connect. And networking becomes a key part. >> And one of the key things, and this is going to be the hard part is optimization. So how do we optimize across particularly performance but even cost? >> And security, rewiring security and availability. >> So still I think my all time favorite business book is Clayton Christensen, "Innovator's Dilemma". One of the most important lessons in that book is what are you optimizing for? And by rule, you can't optimize for everything equally. You have to rank order. But what I find really interesting in this conversation and where we're going and the complexity of the size of the data, the complexity of what am I optimizing for now just begs for plight AI. This is not a people problem to solve. This is AI moving fast. >> Smart infrastructure going to adapt. >> Right, so as you look at that opportunity to now apply AI over the top of this thing, opens up tremendous opportunity. >> Absolutely, I mean standardized infrastructure allows you, sorry, allows you to get more metrics. It allows you to build models to optimize infrastructure over time. >> And humans just can't get their head around it. I mean because you do have to optimize across multiple dimensions as performance, as cost. But then that performance is compute, it's the network. In fact the network's always going to be the bottleneck. So you look at it, even with 5G which is an order magnitude more band width, the network will still lag. You go back to Moore's Law, right? It's a, even though it's extended to 24 months, price performance doubles, so the amount of data potentially can exponentially grow our networks don't keep pace. So that optimization is constantly going to have to be tuned as we get even with increases in network we're going to have to keep balancing that. >> Right, but it's also the business optimization beyond the infrastructure optimization. For instance, if you are running a big power generation field of a bunch of turbines, right, you may want to optimize for maintenance 'cause things are running in some steady state but maybe there's an oil crisis or this or that, suddenly the price rises and you are like, forget the maintenance right now, we've got a revenue opportunity that we want to tweak. >> You just talked about which is in a dynamic industry. How do I real time change the behavior? And more and more policy driven, where the infrastructure is smart enough to react, based on the policy change you made. That's the world we want to get to and we are far away from that right now. >> I mean ultimately I think the Kubernetes controller gets an AI overlay and then operators of the future are tuning the AI engines that optimize it. >> Right, right. And then we run into the whole thing which we talked about many times in this building with Dr. Rumman Chowdhury from Accenture. Then you got the whole ethics overlay on top of the business and the optimization and everything else. That's a whole different conversation for another day. So, before we wrap I just want to give you kind of last thoughts. As you know customers are in all different stages of their journey. Hopefully, most of them are at least off the first square I would imagine on the monopoly board. What does, you know, kind of just top level things that you would tell people that they really need just to keep always at the top as they're starting to make these considerations? Starting to make these investments? Starting to move workloads around that they should always have at the top of their mind? >> For me it's very simple. It's really about focus on the business outcome. Leverage the best resource for the right need. And design architectures that are flexible that give you choice, you're not locked in. And look for strategic partners, whether it's technology partners or services partners that allow you to guide. Because if complexity is too high, the number of choices are too high, you need someone who has the breadth and depth to give you that platform which you can operate on. So we want to be the ubiquitous platform from a software perspective. Accenture wants to be that single partner who can help them guide on the journey. So, I think that would be my ask is start thinking about who are your strategic partners? What is your architecture and the choices you're making that give you the flexibility to evolve. Because this is a dynamic market. Once you make decisions today, may not be the ones you need in six months even. >> And that dynanicism is accelerating. If you look at it, I mean, we've all seen change in the industry, of decades in the industry. But the rate of change now, the pace, things are moving so quickly. >> And we need to respond to competitive or business oriented industry. Or any regulations. You have to be prepared for that. >> Well gentleman, thanks for taking a few minutes and great conversation. Clearly you're in a very good space 'cause it's not getting any less complicated any time soon. >> Well, thank you again. And thank you. >> All right, thanks. >> Thanks. >> Larry and Ajay, I'm Jeff, you're watching theCUBE. We are top of San Francisco in the Sales Force Tower at the Accenture Innovation Hub. Thanks for watching. We'll see you next time.

>> from atop the Salesforce Tower in downtown San Francisco. It's the Q covering Accenture Innovation Date brought to you by ex center >> Hey, welcome back Your body jefe Rick here from the Cube were high atop San Francisco in the essential innovation hub. It's in the middle of the Salesforce Tower. It's a beautiful facility. They think you had it. The grand opening about six months ago. We're here for the grand opening. Very cool space. I got maker studios. They've got all kinds of crazy stuff going on. But we're here today to talk about Cloud in this continuing evolution about cloud in the enterprise and hybrid cloud and multi cloud in Public Cloud and Private Cloud. And we're really excited to have a couple of guys who really helping customers make this journey, cause it's really tough to do by yourself. CEOs are super busy. They worry about security and all kinds of other things. So centers, often a trusted partner. We got two of the leaders from center joining us today's Prasad Sankaran. He's the senior managing director of Intelligent Cloud infrastructure for Center Welcome and Larry Soccer, the global managing director. Intelligent cloud infrastructure offering from central gentlemen. Welcome. I love it. It intelligent cloud. What is an intelligent cloud all about? Got it in your title. It must mean something pretty significant. >> Yeah, I think First of all, thank you for having us, but you're absolutely Everything's around becoming more intelligent around using more automation. And the work that, you know we delivered to our clients and cloud, as you know, is the platform to which all of our clients are moving. So it's all about bringing the intelligence not only into infrastructure, but also into cloud generally. And it's all driven by software, >> right? It's just funny to think where we are in this journey. We talked a little bit before we turn the cameras on and there you made an interesting comment when I said, You know, when did this cloud for the Enterprise start? And you took it back to sass based applications, which, >> you know, you were sitting in the sales force builder. >> That's true. It isn't just the tallest building in here, and everyone all right, everyone's >> had a lot of focus on AWS is rise, etcetera. But the real start was really getting into sass. I mean, I remember We used to do a lot of Siebel deployments for CR M, and we started to pivot to sales, for some were moving from remedy into service. Now I mean, we went through on premise collaboration, email todo 360 5 So So we've actually been at it for quite a while in the particularly the SAS world. And it's only more recently that we started to see that kind of push to the, you know, the public pass, and it's starting to cloud native development. But But this journey started, you know, it was that 78 years ago that we really start to see some scale around it >> and tell me if you agree. I think really, what? The sales forces of the world and the service now is of the world off. 3 65 kind of broke down some of those initial barriers which were all really about security and security. Security secure. It's always too here where now security is actually probably an attribute >> and loud can brink Absolutely. In fact, I'm in those barriers took years to bring down. I still saw clients where they were forcing salesforce tor service. Now to put you know instances on Prime, and I think I think they finally woke up toe. You know, these guys invested ton in their security organizations. You know, there's a little of that needle in the haystack. You know, if you breach a data set, you know what you're getting after. But when you happen to sail sports, it's a lot harder. And so you know. So I think that security problems, I've certainly got away. We still have some compliance, regulatory things, data sovereignty. But I think security and not not that it's all by any means that you know, it's always giving an ongoing problem. But I think they're getting more comfortable with their data being up in the public domain, right? Not public. >> I think it also help them with their progress towards getting cloud native. So, you know, you pick certain applications which were obviously hosted by sales force and other companies, and you did some level of custom development around it. And now I think that's paved the way for more complex applications and different workloads now going into, you know, the public cloud and the private cloud. But that's the next part of the journey, >> right? So Let's back up 1/2 a step cause then, as you said, a bunch of stuff then went into Public Cloud, right? Everyone's putting in AWS and Google. Um, IBM has got a public how there was a lot more. They're not quite so many as there used to be. Um, but then we ran into a whole new home, Those of issues, right, Which is kind of opened up this hybrid cloud. This multi cloud world, which is you just can't put everything into a public clouds there certain attributes that you need to think about and yet from the application point of view, before you decide where you deploy that. So I'm just curious. If you can share now, would you guys do with clients? How should they think about applications? How, after they think about what to deploy where I >> think I'll start in the, You know, Larry has a lot of expertise in this area. I think you know, we have to obviously start from an application centric perspectives. You got to take a look at you know where your applications have to live water. What are some of the data implications on the applications or do you have by way of regulatory and compliance issues? Or do you have to do as faras performance because certain applications have to be in a high performance environment? Certain other applications don't think a lot of these factors will then drive where these applications need to recite. And then what we're seeing in today's world is really accomplish. Complex, um, situation where you have a lot of legacy, but you also have private as well as public cloud. So you approach it from an application perspective. >> Yeah. I mean, if you really take a look at Army, you look at it centers clients, and we were totally focused on up into the market Global 2000 savory. You know, clients typically have application portfolios ranging from 520,000 applications. And really, I mean, if you think about the purpose of cloud or even infrastructure for that, they're there to serve the applications. No one cares if your cloud infrastructure is not performing the absolute. So we start off with an application monetization approach and ultimately looking, you know, you know, with our tech advisory guys coming in, there are intelligent engineering service is to do the cloud native and at mod work our platforms. Guys, who do you know everything from sales forward through ASAP. They should drive a strategy on how those applications going to evolve with its 520,000 and determined hey, and usually using some like the six orders methodology. And I'm I am I going to retire this Am I going to retain it? And I'm gonna replace it with sass. Am I gonna re factor in format? And it's ultimately that strategy that's really gonna dictate a multi in and, you know, hybrid cloud story. So it's based on the applications data, gravity issues where they gonna reside on their requirements around regulatory, the requirements for performance, et cetera. That will then dictate the cloud strategies. I'm you know, not a big fan of going in there and just doing a multi hybrid cloud strategy without a really good up front application portfolio approach, right? How we're gonna modernize that >> it hadn't had a you segment. That's a lot of applications. And you know, how do you know the old thing? How do you know that one by that time, how do you help them pray or size? Where they should be focusing on. Yes, >> it. Typically, what we do is work with our clients to do a full application portfolio analysis, and then we're able to then segment the applications based on, you know, important to the business and some of the factors that both of us mentioned. And once we have that, then we come up with an approach where certain sets of applications have moved to sass certain other applications you moved past. So you know, you're basically doing the re factoring and the modernization, and then certain others, you know, you can just, you know, lift and shift. So it's really a combination off both modernization as well as migration. It's a combination off that, but to do that, you have initially look at the entire set of applications and come up with that approach. >> I'm just curious where within that application assessment, where is cost savings? Where is, uh, this is just old and where is opportunities to innovate faster? Because we know a lot of lot of talk really. Days has cost savings, but what the real advantages is execution speed if you can get it. >> If >> you could go back three or four years and we had there was a lot of CEO discussions around cost savings. I'm not really have seen our clients shift. It costs never goes away, obviously right. But there's a lot greater emphasis now on business agility. You know, howto innovate faster, get, get new capabilities, market faster to change my customer experience. So it's really I t is really trying to step up and, you know, enabled the business toe to compete in the marketplace. So we're seeing a huge shift in emphasis or focus at least starting with, you know, how do I get better business agility outta leverage to cloud and cloud native development to get there upper service levels? Actually, we started seeing increase on Hey, you know, these applications need to work. It's actress, So obviously cost still remains a factor, but we seem much more, you know, much more emphasis on agility, you know, enabling the business on giving the right service levels of right experience to the user. Little customers. Big pivot there, >> Okay. And let's get the definitions out because you know a lot of lot of conversation about public clouds. Easy private clouds, easy but hybrid cloud and multi cloud and confusion about what those are. How do you guys define them? How do you help your customers think about the definition? Yes, >> I think it's a really good point. So what we're starting to see is there were a lot of different definitions out there. But I think as I talk to my clients and our partners, I think we're all starting to come toe. You know, the same kind of definition on multi cloud. It's really about using more than one cloud. But hybrid, I think, is a very important concept because hybrid is really all about the placement off the workload or where your application is going to run on. And then again, it goes to all of these points that we talked about data, gravity and performance and other things. Other factors. But it's really all about where do you place the specific workload >> if you look at that, so if you think about public, I mean obviously gives us the innovation of the public providers. You look at how fast Amazon comes out with new versions of Lambda etcetera, so that's the innovations. There obviously agility. You could spend up environments very quickly which is, you know, one of the big benefits of it. The consumption economic models. So that is the number of drivers that are pushing in the direction of public. You know, on the private side, they're still it's quite a few benefits that don't get talked about as much. Um, so you know, if you look at it performance, you know, if you think the public world, you know, although they're scaling up larger T shirts, et cetera, they're still trying to do that for a large array of applications on the private side, you can really Taylor somethingto very high performance characteristics. Whether it's you know, 30 to 64 terabyte Hana, you can get a much more focused precision environment for business critical workloads like that article, article rack. You know, the Duke clusters everything about fraud analysis. So that's a big part of it. Related to that is the data gravity that Prasad just mentioned. You know, if I've got a 64 terrified Hana database, you know, sitting in my private cloud, it may not be that convenient to go and put get that data shared up in red shift or in Google's tensorflow. So So there's some data gravity out. Networks just aren't there. The Laden sea of moving that stuff around is a big issue. And then a lot of people of investments in their data centers. I mean, the other piece, that's interesting. His legacy, you know, You know, as we start to look at the world a lot, there's a ton of Could still living in, You know, whether it's you, Nick system, that IBM mainframes. There's a lot of business value there, and sometimes the business cases aren't aren't necessarily there toe to replace them. Right. And in world of digital, the decoupling where I can start to use micro service is we're seeing a lot of trends. We worked with one hotel to take the reservation system. You know, Rapid and Micro Service is, um, we then didn't you know, open shift couch base, front end. And now when you go against, you know, when you go and browsing properties, you're looking at rates you actually going into distributed database cash on, you know, in using the latest cloud native technologies that could be dropped every two weeks or every three or four days for my mobile application and It's only when it goes, you know, when the transaction goes back, to reserve the room that it goes back there. So we're seeing a lot of power with digital decoupling, but we still need to take advantage of, you know, we've got these legacy applications. So So the data centers air really were trying to evolve them. And really, just, you know, how do we learn everything from the world of public and struck to bring those saints similar type efficiencies to the to the world of private? And really, what we're saying is this emerging approach where I can start to take advantage of the innovation cycles that land is that you know, the red shifts the azure functions of the public world. But then maybe keep some of my more business critical regulated workloads. You know, that's the other side of the private side, right? I've got G X p compliance. If I've got hip data that I need to worry about GDP are you know, the whole set of regular two requirements Over time, we do anticipate the public guys will get much better and more compliant. In fact, they made great headway already, but they're Still not a number of clients are still, you know, not 100% comfortable from rail client's perspective. >> Gotta meet Teresa Carlson. She'll change him. Who runs that AWS Public Sector is doing amazing things, obviously with big government contracts. But but you raise real inching point later. You almost described what I would say is really a hybrid application in this thing. This hotel example that you use because it's is, you know, kind of break in the application and leveraging micro service is to do things around the core that allowed to take advantage of some this agility and hyper fast development, yet still maintain that core stuff that either doesn't need to move Works fine. Be too expensive. Drea Factor. It's a real different weight. Even think about workloads and applications into breaking those things into bits. >> And we see that pattern all over the place. I'm gonna give you the hotel Example Where but finance, you know, look at financial service. Is retail banking so open banking a lot. All those rito applications are on the mainframe. I'm insurance claims and and you look at it, the business value, replicating a lot of like the regulatory stuff, the locality stuff. It doesn't make sense to write it. There's no rule inherent business values of I can wrap it, expose it and you know the micro service's architecture now. D'oh cloud native front end. That's gonna give me a 360 view a customer, Change the customer experience. You know, I've got a much you know, I can still get that agility. The the innovation cycles by public. Bye bye. Wrapping my legacy environment >> in person, you rated jump in and I'll give you something to react to, Which is which is the single glass right now? How do I How did I manage all this stuff now? Not only do I have distributed infrastructure now, I've got distributed applications and the thing that you just described and everyone wants to be that single pane of glass Everybody wants to be the app that's upon everybody. Screen. How are you seeing people deal with the management complexity of these kind of distributed infrastructures? If you will Yeah, >> I think that that's that's an area that's, ah, actually very topical these days because, you know, you're starting to see more and more workers. Goto private cloud and so you've got a hybrid infrastructure you're starting to see move movement from just using the EMS to, you know, the cantinas and Cuban Edie's. And, you know, we talked about Serval s and so on. So all of our clients are looking for a way, and you have different types of users as well. Yeah, developers. You have data scientists. You have, you know, operators and so on. So they're all looking for that control plane that allows them access and a view toe everything that is out there that is being used in the enterprise. And that's where I think you know, a company like Accenture were able to use the best of breed toe provide that visibility to our clients. >> Yeah. I mean, you hit the nail on the head. It's becoming, you know, with all the promise of cloud and all the power. And these new architectures is becoming much more dynamic, ephemeral, with containers and kubernetes with service computing that that one application for the hotel, they're actually started, and they've got some actually, now running a native us of their containers and looking at serverless. So you gonna even a single application can span that and one of things we've seen is is first. You know, a lot of our clients used to look at, you know, application management, you know, different from their their infrastructure. And the lines are now getting very blurry. You need to have very tight alignment. You take that single application. You know, if any my public side goes down or my mid tier with my you know, you know, open shipped on VM where it goes down on my back and mainframe goes down. Or the networks that connected to go down the devices that talked it. It's a very well, despite the power, very complex environment. So what we've been doing is first we've been looking at, you know, how do we get better synergy across what we you know, application service is teams that do the application manager an optimization cloud infrastructure, you know, how do we get better alignment that are embedded security, You know, how do you know what are managed to Security Service's and bringing those together? And then what we did was we looked at, you know, we got very aggressive of cloud for a strategy and, you know, how do we manage the world of public. But when looking at the public providers of hyper scale er's and how they hit incredible degrees of automation, we really looked at, said and said, Hey, look, you gotta operate differently in this new world. What can we learn from how the public guys they're doing that? We came up with this concept We call it running different. You know, how do you operate differently in this new multi speed? You know, you know, hot, very hybrid world across public, private demon, legacy environment and started looking say OK, what is it that they do? You know, first they standardize, and that's one of the big challenges you know, going to almost all of our clients in this a sprawl. And you know, whether it's application sprawl, its infrastructure, sprawl and >> my business is so unique. The Larry no business out there has the same process that we have. So we started make you know how to be >> standardized like center hybrid cloud solution apart with HP. Envy em where we, you know, how do we that was an example. So we can get thio because you can't automate unless you standardise. So that was the first thing you know, standardizing service catalog. Standardizing that, um, you know, the next thing is the operating model. They obviously operate differently. So we've been putting a lot of time and energy and what I call a cloud and agile operating model. And also a big part of that is truly you hear a lot about Dev ops right now, but truly putting the security and and operations into Deb set cops of bringing, you know, the development in the operations much tied together. So spending a lot of time looking at that and transforming operations re skilling the people you know, the operators of the future aren't eyes on glass there. Developers, they're writing the data ingestion, the analytic algorithms, you know, to do predictive operations. They're writing the automation script to take work, you know, test work out. Right. And over time, they'll be tuned in the air. Aye, aye. Engines to really optimize the environment and then finally has presided. Looted thio. Is that the platforms that control planes? That doing that? So, you know, we What we've been doing is we've had a significant investments in the eccentric cloud platform, our infrastructure automation platforms and then the application teams with it with our my wizard framework, and we've been starting to bring that together. You know, it's an integrated control plane that can plug into our clients environments to really manage seamlessly, you know, and provide, you know, automation Analytics. Aye, aye. Across APS, cloud infrastructure and even security. Right. And that, you know, that really is a iob is right. I mean, that's delivering on, you know, as the industry starts toe define and really coalesce around, eh? I ops, that's what we use. >> So just so I'm clear that so it's really your layer your software layer kind of management layer that that integrates all these different systems and provides kind of a unified view. Control, I reporting et cetera. Right >> Exactly. Then can plug in and integrate, you know, third party tools. I had to do some strategic function. >> I'm just I'm just >> curious is one of the themes that we here out in the press right now is this is this kind of pull back of public cloud app. Some of them are coming back. Or maybe it was, you know, kind of a rush. Maybe a little bit too aggressively. What are some of the reasons why people are pulling stuff back out of public clouds, that just with the wrong it was just the wrong application? The costs were not what we anticipated to be. We find it, you know, what are some of the reasons that you see after coming back in house? Yeah, >> I think it's >> a variety of factors. I mean, it's certainly cost, I think is one. So as there are multiple private options and you know, we don't talk about this, but the hyper skills themselves are coming out with their own different private options, like Aunt Ours and out pulls and other stack and on. And Ali Baba has obsessed I and so on. So you see a proliferation of that and you see many more options around private cloud. So I think the cost is certainly a factor. The second is I think data gravity is, I think, a very important point because as you're starting to see how different applications have to work together, then that becomes a very important point. The third is just about compliance, and, you know, the regulatory environment. As we look across the globe, you know, even outside the U. S. We look at Europe and other parts of Asia as clients and moving more to the cloud. You know, that becomes an important factor. So as you start to balance these things, I think you have to take a very application centric view. You see some of those some some maps moving back, and and I think that's the part of the hybrid world is that you know, you can have a nap running on the private cloud and then tomorrow you can move this. Since it's been containerized to run on public and it's, you know, it's all managed that look >> e. I mean, cost is a big factor if you actually look at it. Most of our clients, you know, they typically you were big cap ex businesses, and all of a sudden they're using this consumption consumption model. And they weren't really They didn't have a function to go and look at the thousands or millions of lines of it, right? You know, as your statement, exactly think they misjudged, you know, some of the scale on B e e. I mean, that's one of the reasons we started. It's got to be an application lead modernization that really that will dictate that. And I think in many cases, people didn't may not have thought through which application. What data? There The data, gravity data. Gravity's a conversation I'm having just by with every client right now. You know, I've got a 64 terabyte hana, and that's the core. My crown jewels. That data, you know, how do I get that to tensorflow? How'd I get that >> right? But if Andy was >> here, though, Andy would say, we'll send down the snow. The snow came from which virgin snow plows Snowball snowball. Well, they're snowballs. But we've seen the >> hold of a truck killer >> that comes out and he'd say, Take that and stick it in the cloud. Because if you've got that data in a single source right now, you can apply multitude of applications across that thing. So they you know they're pushing. Get that date end in this single source course than to move it, change it, you know you run it. All these micro lines of billing statement take >> the hotel. I mean, their data stolen the mainframe. So if they may want need to expose it? Yeah, they have a database cash, and they move it out. You know, the particulars of data sets get larger, it becomes, you know, the data. Gravity becomes a big issue. Because no matter how much you know, while Moore's law might be might have elongated from 18 to 24 months, the network will always be the bottle, Mac. So ultimately, we're seeing, you know, a CZ. We proliferate more and more data, all data sets get bigger and better than network becomes more of a bottleneck. Conned. That's a lot of times you gotta look at your applications. They have. I've got some legacy database I need to get. Thio. I need this to be approximately somewhere where I don't have, you know, high bandwith o r. Right Or, you know, highlight and see type or so egress costs a pretty big deals. My date is up in the cloud, and I'm gonna get charged for pulling it off. You know that That's been a big issue. >> You know, it's funny, I think, and I think a lot of the issue, obviously complexity building. It's a totally different building model, but I think to a lot of people will put stuff in a public cloud and then operated as if they bought it. And they're running in the data center in this kind of this. Turn it on, turn it off when you need it. Everyone turns. Everyone loves to talk about the example turning it on when you need it. But nobody ever talks about turning it off when you don't. But but the kind of clothes on our conversation I won't talk about a I and applied a I. CoSine is a lot of talk in the market place, but a time machine learning. But as you guys know pride better than anybody, it's the application of a I and specific applications, which really on unlocks the value. And as we're sitting here talking about this complexity, I can't help but think that, you know, applied a I in a management layer like your run differently, set up to actually know when to turn things on, when to turn things off when you moved in but not moved, it's gonna have to be machines running that right cause the data sets and the complexity of these systems is going to be just overwhelming. Yeah, yeah, >> absolutely completely agree with you in fact. Ah, essential. We actually referred to the Seoul area as Applied intelligence. Ah, and that's our guy, right? And, uh, it is absolutely to add more and more automation Move everything Maur toe where it's being run by the machine rather than, you know, having people really working on these things >> yet, e I mean, if you think you hit the nail on the head, we're gonna a eyes e. I mean, given how things getting complex, more ephemeral, you think about kubernetes et cetera. We're gonna have to leverage a humans or not to be able to get, you know, manage this. The environment is important, right? What's interesting way we've used quite effectively for quite some time. But it's good at some stuff, not good at others. So we find it's very good at, like, ticket triage, like ticket triage, chicken routing, et cetera. You know, any time we take over account, we tune our AI ai engines. We have ticket advisers, etcetera. That's what probably got the most, you know, most bang for the buck. We tried in the network space. Less success to start even with, you know, commercial products that were out there. I think where a I ultimately bails us out of this is if you look at the problem. You know, a lot of times we talked about optimizing around cost, but then performance. I mean, and it's they they're somewhat, you know, you gotta weigh him off each other. So you've got a very multi dimensional problem on howto I optimize my workloads, particularly. I gotta kubernetes cluster and something on Amazon, you know, sums running on my private cloud, etcetera. So we're gonna get some very complex environment. And the only way you're gonna be ableto optimize across multi dimensions that cost performance service levels, you know, and then multiple options don't do it public private, You know, what's my network costs etcetera. Isn't a I engine tuning that ai ai engines? So ultimately, I mean, you heard me earlier on the operators. I think you know, they write the analytic albums, they do the automation scripts, but they're the ultimate ones who then tune the aye aye engines that will manage our environment, right. And I think it kubernetes will be interesting because it becomes a link to the control plane optimize workload placement between >> when the best thing to you. Then you have dynamic optimization can. You might be up to my tanks at us right now, but you might be optimizing for output the next day. So exists really a you know, kind of Ah, never ending >> when you got you got to see them >> together with it. And multi dimension optimization is very difficult. So I mean, you know, humans can't get their head around. Machines can, but they need to be trained. >> Well, Prasad, Larry, Lots of great opportunities for for centuries bring that expertise to the table. So thanks for taking a few minutes to walk through some of these things. Our pleasure. Thank you. Raise Prasad is Larry. I'm Jeff. You're watching the Cube. We are high above San Francisco in the Salesforce Tower. Theis Center. Innovation have in San Francisco. Thanks for watching. We'll see you next time

(upbeat music) >> Hey, welcome back everybody. Jeff Frick here with theCUBE. We are high atop San Franciscso, in the Salesforce Tower in the brand new Accenture, the Innovation Hub. It opened up, I don't know, six months ago or so. We were here for the opening. It's a really spectacular space with a really cool Cinderella stair, so if you come, make sure you check that out. We're talking about cloud and the evolution of cloud, and hybrid cloud, and clearly, two players that are right in the middle of this, helping customers get through this journey, and do these migrations are Accenture and HPE. So we're excited to have our next guest, Melissa Besse. She is the Managing Director, Intelligent Cloud and Infrastructure Strategic Partnerships, at Accenture. Melissa, welcome. >> Thanks Jeff. >> And joining us from HP is David Stone. He is the VP of Ecosystem Sales. David great to see you. >> Great, thanks for having me. >> So, let's just jump into it. The cloud discussion has taken over for the last 10 years, but it's really continuing to evolve. It was kind of this new entrance, with AWS coming on the scene, one of the great lines that Jeff Bezos talks about, is they had no competition for seven years. Nobody recognized that the bookseller, out on the left hand edge, was coming in to take their infrastructure business. But as things have moved to public cloud, now there's hybrid cloud, now all applications, or work loads, are right for public clouds, so now, all the Enterprises are trying to figure this out, they want to make their moves but it's complicated. So, first of all, let's talk about some of the vocabulary, hybrid cloud versus Multi-Cloud. What do those terms mean to you and your customers? Let's start with you, Melissa. >> Sure. So when you think of Multi-Cloud, right, we're seeing a big convergence of, I would say, a Multi-Cloud operating model, that really has to integrate across all the clouds. So, you have your public cloud providers, you have your SaaS, like Salesforce, work day, you have your PAS, right. And so when you think of Multi-Cloud, any customer is going to have a plethora, of all of these types of clouds. And really being able to manage across those, becomes critical. When you think of Hybrid-Cloud, Hybrid-Cloud is really thinking about the placement of Ous. We usually look at it from a data perspective, right. Are you going to in the public, or in the private space? And you kind of look at it from that perspective. And it really enables that data movement across both, of those clouds. >> So what do you see, David, in your customers? >> I see a lot of the customers, that we see today, are confused, right? The people who have gone to the Public Cloud, had scratched their heads and said, "Geez, what do I do?", "It's not as cheap as I thought it was going to be." So, the ones who are early adopters, are confused. The ones who haven't moved, yet, are really scratching their head as well, right. Because if you don't the right strategy, you'll end up getting boxed in. You'll pay a ton of money to get your data in, and you'll pay a ton of money to get your data out. And so, all of our customers, you know, want the right hybrid strategy. And, I think that's where the market, and I know Accenture and HPE, clearly see the market really becoming a hybrid world. >> It's interesting, you said it's based on the data, and you just talked about moving data in and out. Where we more often here it talked about workload, this kind of horses for courses, you know. It's a workload specific, should be deployed in this particular, kind of infrastructure configuration. But you both mention data, and there's a lot of conversation, kind of pre-cloud, about data gravity and how expensive it is to move the data, and the age old thing, do you move the compute to the data, or move the data to the compute? There's a lot of advantages, if you have that data in the cloud, but you're highlighting a couple of the real negatives, in terms of potential cost implications, and we didn't even get into regulations, and some of the other things that drive workloads to stay, in the data center. So, how should people start thinking about these variables, when they're trying to figure out what to do next? >> Accenture's position definitely, like when we started off on our Hybrid Cloud journey, was to capture the workload, right. And, once you have that workload, you could really balance the public benefits of speed, innovation, and consumption, with the private benefits of, actual regulation, data gravity, and performance, right. And so, our whole approach and big bet, has been to- Basically, we had really good leading public capabilities, cause we got into the market early. But we knew our customers were not going to be able to, migrate their entire estate over to public. And so in doing that, we said okay, if we create a hybrid capability, that is highly automated, that is consumed like public, and that is standard, we'd be able to offer our customers a way to pick really, the right workload, in the right place, at the right price. And that was really what our whole goal was. >> Go ahead. >> Yeah, and so just to add on to what Melissa said, I think we also think about, at least, you know, keeping the data in a place that you want, but then being cloud adjacent, so getting in the right data centers, and we often use a cloud saying, to bring the cloud to the data. So, if you have the right hybrid strategy, you put the data where it makes the most sense. Where you want to maintain the security and privacy, but then have access to the APIs, and whatever else you might need to get the full advantages, of the public cloud. >> Yeah, and we here a lot of the data center providers like, Equinix and stuff, talking about features, like direct connect and, you know, to have this proximity between the public cloud, and the stuff that's in your private cloud, so that you do have, you know, low latency, and you can, when you do have to move things, or you do need to access that data, it's not so far away. I'm curious about the impact of companies like, Salesforce in the Salesforce tower, here in San Francisco, at the center offices, and office 365, and Work Day, on how can the adoption of the SaaS applications, have changed the conversation about cloud, and what's important and not important, it used to be security, I don't trust anything outside my data center, and know I might argue that public clouds are more secure, in some ways that private cloud, you don't have disgruntled employees per se, running around the data centers unplugging things. So, how it the adoption of things like Office 365, clearly Microsoft's leveraged that in a big way, to grow their own cloud presence, change the conversation about what's good about cloud, what's not good about cloud, why should we move in this direction. David, you have a thought? >> No, look, I think it's a great question, and I think if you think about the, as Melissa said, the used cases, right. And, how Microsoft has successfully pivoted, their business to it as a service model, right. And so what I think it's done, it's opened up innovation, and a lot of the Salesforces of the world, have adapted their business models. And that's truly to your point, a SaaS based offer, and so when you can do a Work Day, or Salesforce.com implementation, sure, it's been built, it's tested and everything else. I think what then becomes the bigger question, and the bigger challenge is, most companies are sitting on a thousand applications, that have been built over time. And what do you do with those, right? And so, in many cases you need to be connected, to those SaaS space providers, but you need the right hybrid strategy, again, to be able to figure out, how to connect those SaaS space services, to whatever you're going to do, with those thousand workloads. And those thousand workloads, running on different things, you need the right strategy, to figure out where to put the actual workloads. And, as people are trying to go, I know one of the questions that comes up is, do you migrate? Or do you modernize? >> David: And so, as people put that strategy together, I think how you tie to those SaaS space services, clearly ties into your hybrid strategy. >> I would agree, and so, as David mentioned, right. That's where the cloud adjacency, you're seeing a lot of blur, between public and private, I mean, Google's providing Bare-metal as a service. So it is actually dedicated, hybrid cloud capabilities, right. So you're seeing a lot of everyone, and as David talked about, all of the surrounding applications around your SAP, around your oracle. When we created our Exensor Hyper Cloud, we were going after the Enterprise workload. But there's a lot of legacy and other ones, that need that data, and or, the Salesforce data. Whatever the data is, right. And really be able to utilize it when they need to, in a real low latency. >> So, I was wondering I we could unpack, the Accenture Hybrid Cloud. >> Melissa: Sure. >> What is that? Is that your guys own cloud? Is this, you know, kind of the solution set? I've heard that mentioned a couple times. So what is the Accenture Hybrid Cloud? >> So Accenture Hybrid Cloud, was a big bet that we made, as we saw the convergence of MultiCloud. We really said, we know, everything is not going to go public. And in some cases, it's all coming back. And so, customers really needed a way, to look at all of their workloads, right. Because part of the issue with, the getting the cost and benefits out of public is, the workload goes but you really aren't able, to get out of the data center. We term it the "Wild Animal Park", because there's a lot of applications that, right, are you going to modernize, are you going to let them to end of life. So there's a lot of things you have to consider, to truly exit the data center strategy. And so, Accenture Hybrid Cloud is actually, a big bet we made, it is a highly automated, standard private cloud capability, that really augments all of the leading capability, we had in the cloud area. It is, it's differentiated, we made a big bet with HPE, it's differentiated on it's hardware. One of the reasons, when we were going after the Enterprise, was they need large compute, and large storage requirements. And what we're able to do is, when we created this, use some of our automation differentiation. We have actually a client, that we had in the existing I-O-N environment, and we were actually able to achieve, some significant benefits, just from the automation. We got 50 percent in the provisioning of applications. We got 40 percent in the provisioning of the V.M. And we were able to take a lot of what I'll call, the manual tasks, and down to, it was like 62 percent reduction in the effort. As well as, 33 percent savings overall, in getting things production ready. So, this capability is highly automated. It will actually repeat the provisioning, at the application level, because we're going after the Enterprise workloads. And it will create these, it's an ASA that came from government, so it's highly secured, and it really was able to preserve, I think what our customer needed. And being able to span that public/private, capability they need out there in the hybrid world. >> Yeah, I was going to say, I don't know that there's enough talk, about the complexity of the management in these worlds. Nobody ever wants to talk about writing, the CIS Admin piece of the software, right? It's all about the core functionality. Let's shift gears a little bit and talk about HPC, a lot of conversation about high performance computing, a lot going on with A.I. and machine learning now. Which, you know, most of those benefits are going to be, realized in a specific application, right? It's machine learning or artificial intelligence, applied to a specific application. So, again, you guys make big iron, and have been making big iron for a long time, what is this kind of hybrid cloud open up, in terms of, for HPE to have the big heavy metal, and still have kind of the agility and flexibility, of a cloud type of infrastructure. >> Yeah, no, I think it's a great question. I think if you think about HPE's strategy has been, in this area of high performance compute. That we bought the company S.G.I. And as you have seen the announcements, we're hopefully going to close on the Cray acquisition as well. And so we in the world of the data continuing to expand, and at huge volumes. The need to have incredible horsepower to drive that, that's associated with it, now all of this really requires, where's your data being created, and where's it actually being consumed? And so, you need to have the right edge, to cloud strategy in everything. And so, in many cases, you need enough compute at the edge, to be able to compute and do stuff in real time. But in many cases you need to feed all that data, back into another cloud or some sort of mother. HPE, you know, type of high performance compute environment, that can actually run the more, advanced A.I. machine learning type of applications, to really get the insights and tune the algorithms. And then, push some of those APIs and applications, back to the edge. So, it's an area of huge investment, it's an area where because of the latency, you know, things like the autonomous driving, and things like that. You can't put all that stuff into the public cloud. But you need the public cloud, or you need cloud type capability, if you will, to be able to compute and make the right decisions, at the right time. So, it's about having the right compute technology, at the right place, at the right time, at the right cost, and the right perform. >> A lot of rights, good opportunity for Accenture. So, I mean it's funny as we talk about hybrid cloud, and that kind of new, verbs around cloud-like things. Is where we're going to see the same thing, kind of the edge versus the data center comparison, in terms of where the data is, where the processing is, because it's going to be this really dynamic situation, and how much can you push out of the edge, cause, you know, there's no air conditioning a lot of times, and the power might not be that great, and maybe connectivity is a little bit limited. So, you know, Edge offers a whole bunch of, different challenges that you can control for, in a data center but it is going to be this crazy, kind of hybrid world there too, in terms of where the allocation of those resources are. You guys get into the deeper end of that model, Melissa? >> Yeah, so we're definitely working with HPE, to create some of, I'll call it our edge managed services, again, going back to what we were saying about the data, right, we saw the centralization of data with the cloud, with the initial entrance into the cloud, now we're seeing the decentralization of that data, back out to the Edge. With that, right, in these hybrid cloud models, you're really going to need- They require a lot of high performance compute, especially for certain industries, right? If you take a look at gas, oil, and exploration, if you look at media processing, right, all of these need to be able to do that. One of the things, and depending on where it's located, if it's on the Edge, how you're going to feedback the data as we talked about. And so, we're looking at, how do you take this foundation, right, this, I'll call it Exensor Hybrid architecture, right. Take that, and play that intermediate role. I'm going to call it intermediary, right, because you really need a really good, you know, global data map, you need a good supply chain, right. Really to make sure that the data, no matter where it's coming from, is going to be available for that application, at the right time. With, right, the ability to do it at speed. And so, all of these things are factors, as you look at our whole Exensor Hybrid Cloud strategy, right. And being able to manage that, Edge to core and then back up to Cloud, etcetera. >> Right, now I wonder if you could share some stories, cause the value proposition around Cloud, is significantly shifted for those who are paying attention, right. But it's not about cost, it's not about cost savings, I mean there's a lot of that in there and that's good, but really the opportunity is about speed. Speed and innovation. And enabling more innovation across your Enterprise, with more people having more access to more data, to build more apps, and really, to react. Are people getting that? Or, are they still, the customer still kind of encumbered, by this kind of transition phase, they're still trying to sort it out, or do they get it? That really this opportunity is about speed, speed, speed. >> No, go ahead. I mean we use a phrase first off, it's, "fear no cloud", right. To your point, you know, how do you figure out the right strategy. But, I think within that you get, what's the right application? And how do you, you know, fit it in to the overall strategy, of what you're trying to do. >> Yeah. >> And I think the other thing that we're seeing is, you know, customers are trying to figure that out. We have a whole, right, when you start with that application map, you know, there could be 500 to 1000 workloads, right, and applications, and how are you going to, some you're going to retain, some you're going to retire, some you're going to (stutters) refactor for the cloud, or for your private cloud capability. Whatever it is, you're going to be looking at doing, I think, you know, we're seeing early adopters, like even the hyperscalers, themselves, right. They recognize the speed, so you know, we're working with Google for instance. They wanted to get into the Bare-metal, as a service capability, right. Them actually building it, getting it out to market would take so much longer. We already had this whole Exensor Hybrid Cloud architecture, that was cloud adjacent, so we had sub-millisecond latency, right. And so, they're the ones, right, everyone's figuring out that utilizing all of these, I'll call it platforms and prebook capabilities. Many of our partners have them as well, is really allowing them that innovation, get products to market sooner, be able to respond to their customers. Because it is, as we talked about in this multicloud world, lots of things that you have to manage, if you can get pieces from multiple, you know, from a partner, right, that can provide more of the services that you need, it really enables the management of those clouds sources. >> Right, so we're going to wrap it up, but I just want to give you the last word in terms of, what's the most consistent blind spot, that you see when you're first engaging with a customer, who's relatively early on this journey, that they miss, that you see over, and over, and over, and you're like, you know, these are some of the thing you really got to think about, that they haven't thought about. >> Yeah so, for me, I think it's- the cloud isn't about a destination, it's about an experience. And so, how do you get- you talked about the operations, but how do you provide that overall experience? I like to use this simple analogy, that if you and I needed a car, for five or 10, or 15 minutes, you go get an Uber. Cause it's easy, it's quick. If you need a car for a couple days, you do a rental car. You need a car for a year, you might do a lease. You need a car for three, four year, you probably by it, right? And so, if you use that analogy and think, Hmmm, I need a workload application for five/six years, putting something at a persistent workload, that you know about on a public cloud, may be the right answer, but it might be a lot more cost prohibited. But, if you need something, that you can stand up in five minutes, and shut it right back down, the public cloud is absolutely, the right way to go, as long as you can deal with the security requirements, and stuff. And so, if you think about, what are the actual requirements, is it cost, is it performance, you've talked about speed and everything else. It's really trying to figure out how you get an experience, and the only experience that can really hit you, what you need to do today, is having the right hybrid strategy. And every company, I know Accenture was out, way in front of the market on public cloud, and now they've come to the realization, so has many other places. The world is going to be hybrid, it's going to be multicloud. And as long as you can have an experience, and a partner, that can manage, you know, help you define the right path, you'll be on the right journey. >> Jeff: Melissa. >> I think blind spot we run into is, it does start off as a cost savings activity. And there really, it really is so much more about, how are you going to manage that enterprise workload? How are you going to worry about the data? Are you going to have access to it? Are you going to be able to make it fluid, right? The whole essence of cloud, right, what it disrupted was the thought, that something had to stay in one place, right. And, where the real time decisions were being made. Where things needed to happen. Now, through all the different clouds, as well as, that you had to own it yourself, right. I mean, everyone always thought, okay, I'll take all the, you know, I.T. department, and very protective of everything that it wanted to keep. Now, it's about saying, all right, how do I utilize, the best of each of these multiclouds, to stand up, what I'll call, what their core capability is as a customer, right. Are they doing the next chip design? Are they, you know, doing financial market models, right? That requires a high performance capability, right. So, when you start to think about all of this stuff, right, that's the true power, is having a strategy that looks at those outcomes. What am I trying to achieve in getting my products, and services to market, and touching the customers I need. Versus, oh, I'm going to move this out to an infrastructure, because that's what cloud, it'll save me money, right. That's typically the downfall we see, because they're not looking at it from the workload, or the application. >> Same old story, right? Focus on your core differentiator, and outsource the heavy lifting on the stuff, (laughs) that's not your core. Alright, well Melissa, David, thanks for taking a minute, and I really enjoyed the conversation. >> Thanks, Jeff. >> She's Melissa, He's David, and I'm Jeff Frick, you're watching theCUBE. We are high above the San Francisco skyline, in the Salesforce tower at the Accenture Innovation Hub. Thanks for watching, we'll see you next time. (tech music)

(upbeat music) >> Hey, welcome back everybody. Jeff Frick here with theCUBE. We are high atop San Franciscso, in the Salesforce Tower in the brand new Accenture, the Innovation Hub. It opened up, I don't know, six months ago or so. We were here for the opening. It's a really spectacular space with a really cool Cinderella stair, so if you come, make sure you check that out. We're talking about cloud and the evolution of cloud, and hybrid cloud, and clearly, two players that are right in the middle of this, helping customers get through this journey, and do these migrations are Accenture and HPE. So we're excited to have our next guest, Melissa Besse. She is the Managing Director, Intelligent Cloud and Infrastructure Strategic Partnerships, at Accenture. Melissa, welcome. >> Thanks Jeff. >> And joining us from HP is David Stone. He is the VP of Ecosystem Sales. David great to see you. >> Great, thanks for having me. >> So, let's just jump into it. The cloud discussion has taken over for the last 10 years, but it's really continuing to evolve. It was kind of this new entrance, with AWS coming on the scene, one of the great lines that Jeff Bezos talks about, is they had no competition for seven years. Nobody recognized that the bookseller, out on the left hand edge, was coming in to take their infrastructure business. But as things have moved to public cloud, now there's hybrid cloud, now all applications, or work loads, are right for public clouds, so now, all the Enterprises are trying to figure this out, they want to make their moves but it's complicated. So, first of all, let's talk about some of the vocabulary, hybrid cloud versus Multi-Cloud. What do those terms mean to you and your customers? Let's start with you, Melissa. >> Sure. So when you think of Multi-Cloud, right, we're seeing a big convergence of, I would say, a Multi-Cloud operating model, that really has to integrate across all the clouds. So, you have your public cloud providers, you have your SaaS, like Salesforce, work day, you have your PAS, right. And so when you think of Multi-Cloud, any customer is going to have a plethora, of all of these types of clouds. And really being able to manage across those, becomes critical. When you think of Hybrid-Cloud, Hybrid-Cloud is really thinking about the placement of Ous. We usually look at it from a data perspective, right. Are you going to in the public, or in the private space? And you kind of look at it from that perspective. And it really enables that data movement across both, of those clouds. >> So what do you see, David, in your customers? >> I see a lot of the customers, that we see today, are confused, right? The people who have gone to the Public Cloud, had scratched their heads and said, "Geez, what do I do?", "It's not as cheap as I thought it was going to be." So, the ones who are early adopters, are confused. The ones who haven't moved, yet, are really scratching their head as well, right. Because if you don't the right strategy, you'll end up getting boxed in. You'll pay a ton of money to get your data in, and you'll pay a ton of money to get your data out. And so, all of our customers, you know, want the right hybrid strategy. And, I think that's where the market, and I know Accenture and HPE, clearly see the market really becoming a hybrid world. >> It's interesting, you said it's based on the data, and you just talked about moving data in and out. Where we more often here it talked about workload, this kind of horses for courses, you know. It's a workload specific, should be deployed in this particular, kind of infrastructure configuration. But you both mention data, and there's a lot of conversation, kind of pre-cloud, about data gravity and how expensive it is to move the data, and the age old thing, do you move the compute to the data, or move the data to the compute? There's a lot of advantages, if you have that data in the cloud, but you're highlighting a couple of the real negatives, in terms of potential cost implications, and we didn't even get into regulations, and some of the other things that drive workloads to stay, in the data center. So, how should people start thinking about these variables, when they're trying to figure out what to do next? >> Accenture's position definitely, like when we started off on our Hybrid Cloud journey, was to capture the workload, right. And, once you have that workload, you could really balance the public benefits of speed, innovation, and consumption, with the private benefits of, actual regulation, data gravity, and performance, right. And so, our whole approach and big bet, has been to- Basically, we had really good leading public capabilities, cause we got into the market early. But we knew our customers were not going to be able to, migrate their entire estate over to public. And so in doing that, we said okay, if we create a hybrid capability, that is highly automated, that is consumed like public, and that is standard, we'd be able to offer our customers a way to pick really, the right workload, in the right place, at the right price. And that was really what our whole goal was. >> Go ahead. >> Yeah, and so just to add on to what Melissa said, I think we also think about, at least, you know, keeping the data in a place that you want, but then being cloud adjacent, so getting in the right data centers, and we often use a cloud saying, to bring the cloud to the data. So, if you have the right hybrid strategy, you put the data where it makes the most sense. Where you want to maintain the security and privacy, but then have access to the APIs, and whatever else you might need to get the full advantages, of the public cloud. >> Yeah, and we here a lot of the data center providers like, Equinix and stuff, talking about features, like direct connect and, you know, to have this proximity between the public cloud, and the stuff that's in your private cloud, so that you do have, you know, low latency, and you can, when you do have to move things, or you do need to access that data, it's not so far away. I'm curious about the impact of companies like, Salesforce in the Salesforce tower, here in San Francisco, at the center offices, and office 365, and Work Day, on how can the adoption of the SaaS applications, have changed the conversation about cloud, and what's important and not important, it used to be security, I don't trust anything outside my data center, and know I might argue that public clouds are more secure, in some ways that private cloud, you don't have disgruntled employees per se, running around the data centers unplugging things. So, how it the adoption of things like Office 365, clearly Microsoft's leveraged that in a big way, to grow their own cloud presence, change the conversation about what's good about cloud, what's not good about cloud, why should we move in this direction. David, you have a thought? >> No, look, I think it's a great question, and I think if you think about the, as Melissa said, the used cases, right. And, how Microsoft has successfully pivoted, their business to it as a service model, right. And so what I think it's done, it's opened up innovation, and a lot of the Salesforces of the world, have adapted their business models. And that's truly to your point, a SaaS based offer, and so when you can do a Work Day, or Salesforce.com implementation, sure, it's been built, it's tested and everything else. I think what then becomes the bigger question, and the bigger challenge is, most companies are sitting on a thousand applications, that have been built over time. And what do you do with those, right? And so, in many cases you need to be connected, to those SaaS space providers, but you need the right hybrid strategy, again, to be able to figure out, how to connect those SaaS space services, to whatever you're going to do, with those thousand workloads. And those thousand workloads, running on different things, you need the right strategy, to figure out where to put the actual workloads. And, as people are trying to go, I know one of the questions that comes up is, do you migrate? Or do you modernize? >> David: And so, as people put that strategy together, I think how you tie to those SaaS space services, clearly ties into your hybrid strategy. >> I would agree, and so, as David mentioned, right. That's where the cloud adjacency, you're seeing a lot of blur, between public and private, I mean, Google's providing Bare-metal as a service. So it is actually dedicated, hybrid cloud capabilities, right. So you're seeing a lot of everyone, and as David talked about, all of the surrounding applications around your SAP, around your oracle. When we created our Exensor Hyper Cloud, we were going after the Enterprise workload. But there's a lot of legacy and other ones, that need that data, and or, the Salesforce data. Whatever the data is, right. And really be able to utilize it when they need to, in a real low latency. >> So, I was wondering I we could unpack, the Accenture Hybrid Cloud. >> Melissa: Sure. >> What is that? Is that your guys own cloud? Is this, you know, kind of the solution set? I've heard that mentioned a couple times. So what is the Accenture Hybrid Cloud? >> So Accenture Hybrid Cloud, was a big bet that we made, as we saw the convergence of MultiCloud. We really said, we know, everything is not going to go public. And in some cases, it's all coming back. And so, customers really needed a way, to look at all of their workloads, right. Because part of the issue with, the getting the cost and benefits out of public is, the workload goes but you really aren't able, to get out of the data center. We term it the "Wild Animal Park", because there's a lot of applications that, right, are you going to modernize, are you going to let them to end of life. So there's a lot of things you have to consider, to truly exit the data center strategy. And so, Accenture Hybrid Cloud is actually, a big bet we made, it is a highly automated, standard private cloud capability, that really augments all of the leading capability, we had in the cloud area. It is, it's differentiated, we made a big bet with HPE, it's differentiated on it's hardware. One of the reasons, when we were going after the Enterprise, was they need large compute, and large storage requirements. And what we're able to do is, when we created this, use some of our automation differentiation. We have actually a client, that we had in the existing I-O-N environment, and we were actually able to achieve, some significant benefits, just from the automation. We got 50 percent in the provisioning of applications. We got 40 percent in the provisioning of the V.M. And we were able to take a lot of what I'll call, the manual tasks, and down to, it was like 62 percent reduction in the effort. As well as, 33 percent savings overall, in getting things production ready. So, this capability is highly automated. It will actually repeat the provisioning, at the application level, because we're going after the Enterprise workloads. And it will create these, it's an ASA that came from government, so it's highly secured, and it really was able to preserve, I think what our customer needed. And being able to span that public/private, capability they need out there in the hybrid world. >> Yeah, I was going to say, I don't know that there's enough talk, about the complexity of the management in these worlds. Nobody ever wants to talk about writing, the CIS Admin piece of the software, right? It's all about the core functionality. Let's shift gears a little bit and talk about HPC, a lot of conversation about high performance computing, a lot going on with A.I. and machine learning now. Which, you know, most of those benefits are going to be, realized in a specific application, right? It's machine learning or artificial intelligence, applied to a specific application. So, again, you guys make big iron, and have been making big iron for a long time, what is this kind of hybrid cloud open up, in terms of, for HPE to have the big heavy metal, and still have kind of the agility and flexibility, of a cloud type of infrastructure. >> Yeah, no, I think it's a great question. I think if you think about HPE's strategy has been, in this area of high performance compute. That we bought the company S.G.I. And as you have seen the announcements, we're hopefully going to close on the Cray acquisition as well. And so we in the world of the data continuing to expand, and at huge volumes. The need to have incredible horsepower to drive that, that's associated with it, now all of this really requires, where's your data being created, and where's it actually being consumed? And so, you need to have the right edge, to cloud strategy in everything. And so, in many cases, you need enough compute at the edge, to be able to compute and do stuff in real time. But in many cases you need to feed all that data, back into another cloud or some sort of mother. HPE, you know, type of high performance compute environment, that can actually run the more, advanced A.I. machine learning type of applications, to really get the insights and tune the algorithms. And then, push some of those APIs and applications, back to the edge. So, it's an area of huge investment, it's an area where because of the latency, you know, things like the autonomous driving, and things like that. You can't put all that stuff into the public cloud. But you need the public cloud, or you need cloud type capability, if you will, to be able to compute and make the right decisions, at the right time. So, it's about having the right compute technology, at the right place, at the right time, at the right cost, and the right perform. >> A lot of rights, good opportunity for Accenture. So, I mean it's funny as we talk about hybrid cloud, and that kind of new, verbs around cloud-like things. Is where we're going to see the same thing, kind of the edge versus the data center comparison, in terms of where the data is, where the processing is, because it's going to be this really dynamic situation, and how much can you push out of the edge, cause, you know, there's no air conditioning a lot of times, and the power might not be that great, and maybe connectivity is a little bit limited. So, you know, Edge offers a whole bunch of, different challenges that you can control for, in a data center but it is going to be this crazy, kind of hybrid world there too, in terms of where the allocation of those resources are. You guys get into the deeper end of that model, Melissa? >> Yeah, so we're definitely working with HPE, to create some of, I'll call it our edge managed services, again, going back to what we were saying about the data, right, we saw the centralization of data with the cloud, with the initial entrance into the cloud, now we're seeing the decentralization of that data, back out to the Edge. With that, right, in these hybrid cloud models, you're really going to need- They require a lot of high performance compute, especially for certain industries, right? If you take a look at gas, oil, and exploration, if you look at media processing, right, all of these need to be able to do that. One of the things, and depending on where it's located, if it's on the Edge, how you're going to feedback the data as we talked about. And so, we're looking at, how do you take this foundation, right, this, I'll call it Exensor Hybrid architecture, right. Take that, and play that intermediate role. I'm going to call it intermediary, right, because you really need a really good, you know, global data map, you need a good supply chain, right. Really to make sure that the data, no matter where it's coming from, is going to be available for that application, at the right time. With, right, the ability to do it at speed. And so, all of these things are factors, as you look at our whole Exensor Hybrid Cloud strategy, right. And being able to manage that, Edge to core and then back up to Cloud, etcetera. >> Right, now I wonder if you could share some stories, cause the value proposition around Cloud, is significantly shifted for those who are paying attention, right. But it's not about cost, it's not about cost savings, I mean there's a lot of that in there and that's good, but really the opportunity is about speed. Speed and innovation. And enabling more innovation across your Enterprise, with more people having more access to more data, to build more apps, and really, to react. Are people getting that? Or, are they still, the customer still kind of encumbered, by this kind of transition phase, they're still trying to sort it out, or do they get it? That really this opportunity is about speed, speed, speed. >> No, go ahead. I mean we use a phrase first off, it's, "fear no cloud", right. To your point, you know, how do you figure out the right strategy. But, I think within that you get, what's the right application? And how do you, you know, fit it in to the overall strategy, of what you're trying to do. >> Yeah. >> And I think the other thing that we're seeing is, you know, customers are trying to figure that out. We have a whole, right, when you start with that application map, you know, there could be 500 to 1000 workloads, right, and applications, and how are you going to, some you're going to retain, some you're going to retire, some you're going to (stutters) refactor for the cloud, or for your private cloud capability. Whatever it is, you're going to be looking at doing, I think, you know, we're seeing early adopters, like even the hyperscalers, themselves, right. They recognize the speed, so you know, we're working with Google for instance. They wanted to get into the Bare-metal, as a service capability, right. Them actually building it, getting it out to market would take so much longer. We already had this whole Exensor Hybrid Cloud architecture, that was cloud adjacent, so we had sub-millisecond latency, right. And so, they're the ones, right, everyone's figuring out that utilizing all of these, I'll call it platforms and prebook capabilities. Many of our partners have them as well, is really allowing them that innovation, get products to market sooner, be able to respond to their customers. Because it is, as we talked about in this multicloud world, lots of things that you have to manage, if you can get pieces from multiple, you know, from a partner, right, that can provide more of the services that you need, it really enables the management of those clouds sources. >> Right, so we're going to wrap it up, but I just want to give you the last word in terms of, what's the most consistent blind spot, that you see when you're first engaging with a customer, who's relatively early on this journey, that they miss, that you see over, and over, and over, and you're like, you know, these are some of the thing you really got to think about, that they haven't thought about. >> Yeah so, for me, I think it's- the cloud isn't about a destination, it's about an experience. And so, how do you get- you talked about the operations, but how do you provide that overall experience? I like to use this simple analogy, that if you and I needed a car, for five or 10, or 15 minutes, you go get an Uber. Cause it's easy, it's quick. If you need a car for a couple days, you do a rental car. You need a car for a year, you might do a lease. You need a car for three, four year, you probably by it, right? And so, if you use that analogy and think, Hmmm, I need a workload application for five/six years, putting something at a persistent workload, that you know about on a public cloud, may be the right answer, but it might be a lot more cost prohibited. But, if you need something, that you can stand up in five minutes, and shut it right back down, the public cloud is absolutely, the right way to go, as long as you can deal with the security requirements, and stuff. And so, if you think about, what are the actual requirements, is it cost, is it performance, you've talked about speed and everything else. It's really trying to figure out how you get an experience, and the only experience that can really hit you, what you need to do today, is having the right hybrid strategy. And every company, I know Accenture was out, way in front of the market on public cloud, and now they've come to the realization, so has many other places. The world is going to be hybrid, it's going to be multicloud. And as long as you can have an experience, and a partner, that can manage, you know, help you define the right path, you'll be on the right journey. >> Jeff: Melissa. >> I think blind spot we run into is, it does start off as a cost savings activity. And there really, it really is so much more about, how are you going to manage that enterprise workload? How are you going to worry about the data? Are you going to have access to it? Are you going to be able to make it fluid, right? The whole essence of cloud, right, what it disrupted was the thought, that something had to stay in one place, right. And, where the real time decisions were being made. Where things needed to happen. Now, through all the different clouds, as well as, that you had to own it yourself, right. I mean, everyone always thought, okay, I'll take all the, you know, I.T. department, and very protective of everything that it wanted to keep. Now, it's about saying, all right, how do I utilize, the best of each of these multiclouds, to stand up, what I'll call, what their core capability is as a customer, right. Are they doing the next chip design? Are they, you know, doing financial market models, right? That requires a high performance capability, right. So, when you start to think about all of this stuff, right, that's the true power, is having a strategy that looks at those outcomes. What am I trying to achieve in getting my products, and services to market, and touching the customers I need. Versus, oh, I'm going to move this out to an infrastructure, because that's what cloud, it'll save me money, right. That's typically the downfall we see, because they're not looking at it from the workload, or the application. >> Same old story, right? Focus on your core differentiator, and outsource the heavy lifting on the stuff, (laughs) that's not your core. Alright, well Melissa, David, thanks for taking a minute, and I really enjoyed the conversation. >> Thanks, Jeff. >> She's Melissa, He's David, and I'm Jeff Frick, you're watching theCUBE. We are high above the San Francisco skyline, in the Salesforce tower at the Accenture Innovation Hub. Thanks for watching, we'll see you next time. (tech music)