(bright music) >> Welcome back everyone to theCUBE's coverage of AWS re:Invent 2021 in person. So a live event, physical in-person, also virtual hybrid. So a lot of great action online, check out the website. All the videos are there on theCUBE, as well as what's going on all of the actions on site and theCUBE's here. I'm John Furrier, your host with Dave Vellante, my cohost. Finally, we've got David Brown, VP of Elastic Compute Cloud. EC2, the bread and butter. Our favorite part of Amazon. David, great to have you back on theCUBE in person. >> John, it's great to be back. It's the first time I'd been on theCUBE in person as well. A lot of virtual events with you guys, but it's amazing to be back at re:Invent. >> We're so excited for you. I know, Matt Garman and I've talked in the past. We've talked in the past. EC2 is just an amazing product. It's always been the core block of AWS. More and more action happening and developers are now getting more action and there's well, we wrote a big piece about it. What's going on? The Silicon's really paying off. You've got to also general purpose Intel and AMD, and you've got the custom silicon, all working together. What's the new update? Give us a scoop. >> Well, John, it's actually 15 years of EC2 this year and I've been lucky to be on that team for 14 years and so incredible to see the growth. It's been an amazing journey. The thing that's really driven us, two things. One is supporting new workloads. And so what are the workloads that customers have available out there trying to do on the cloud that we don't support and launch new instance types. And that's the first thing. The second one is price performance. How do we give customers more performance at a continuously decreasing price year-over-year? And that's just driven innovation across EC2 over the years with things like Graviton. All of our inferential chips are custom silicon, but also instance types with the latest Intel Ice Lake CPU's, latest Milan. We just announced the AMD Milan instance. It's just constantly innovation across the ever-increasing list of instances. So super exciting. >> So instances become the new thing. Provision an instance, spin up an instance. Instance becomes, and you can get instances, flavors, almost like flavors, right? >> David: Yeah. >> Take us through the difference between an instance and then the EC2 itself. >> That's correct, yeah. So we actually have, by end of the year, right now we have over 475 different instances available to you whether it's GPU accelerators, high-performance computing instances, memory optimized, just enormous number. We'll actually hit 500 by the end of the year, but that is it. I mean, customers are looking for different types of machines and those are the instances. >> So the Custom Silicon, it's one of the most interesting developments. We've written about it. AWS secret weapon is one of them. I wonder if you could take us back to the decision points and the journey. The Annapurna acquisition, you started working with them as a partner, then you said, all right, let's just buy the company. >> David: Yeah. >> And then now, you're seeing the acceleration, your time to tapeout is way, way compressed. Maybe what was the catalyst and maybe we can get into where it's going. >> Yeah, absolutely. Super interesting story 'cause it actually starts all the way back in 2008. In 2008, EC2 had actually been around for just a little under two years. And if you remember back then, everybody was like, will virtualize and hypervisors, specialization would never really get you the same performances, what they were calling bare metal back then. Everybody's looking at the cloud. And so we took a look at that. And I mean, network latencies, in some cases with hypervisors were as high as 200 or 300 milliseconds. And it was a number of real challenges. And so we knew that we would have to change the way that virtualization works and get into hardware. And so in 2010, 2011, we started to look at how could I offload my network processing, my IO processing to additional hardware. And that's what we delivered our first Nitro card in 2012 and 2013. We actually offloaded all of the processing of network to a Nitro card. And that Nitro card actually had a Annapurna arm chip on it. Our Nitro 1 chip. >> For the offload? >> The offload card, yeah. And so that's when my team started to code for Arm. We started to work on our Linux works for Arm. We actually had to write our own operating system initially 'cause there weren't any operating systems available we could use. And so that's what we started this journey. And over the years, when we saw how well it worked for networking, we said, let's do it for storage as well. And then we said, Hey, we could actually improve security significantly. And by 2017, we'd actually offloaded 100% of everything we did on that server to our offload cards Leaving a 100% of the server available for customers. And we're still actually the only cloud provider that does that today. >> Just to interject, in the data center today, probably 30% of the general purpose cores are used for offloads. You're saying 0% in the cloud. >> On our nitro instances, so every instance we've launched since 2017, our C5. We use 0% of that central core. And you can actually see that in our instance types. If you look at our largest instance type, you can see that we're giving you 96 cores and we're giving you, and our largest instance, 24 terabytes of memory. We're not giving you 23.6 terabytes 'cause we need some. It's all given to you as the customer. >> So much more efficient, >> Much, much more efficient, much better, better price performance as well. But then ultimately those Nitro chips, we went through Nitro 1, Nitro 2, Nitro 3, Nitro 4. We said, Hey, could we build a general purpose server chip? Could we actually bring Arm into the cloud? And in 2018, we launched the A1 instance, which was our Graviton1 instance. And what we didn't tell people at the time is that it was actually the same chip we were using on our network card. So essentially, it was a network card that we were giving to you as a server. But what it did is it sparked the ecosystem. That's why we put it out there. And I remember before launch, some was saying, is this just going to be a university project? Are we going to see people from big universities using Arm in the cloud? Was it really going to take off? And the response was amazing. The ecosystem just grew. We had customers move to it and immediately begin to see improvements. And we knew that a year later, Graviton2 was going to come out. And Graviton2 was just an amazing chip. It continues to see incredible adoption, 40% price performance improvement over other instances. >> So this is worth calling out because I think that example of the network card, I mean, innovation can come from anywhere. This is what Jassy always would say is do the experiments. Think about the impact of what's going on here. You're focused on a mission. Let's get that processing of the lowest cost, pick up some workloads. So you're constantly tinkering with tuning the engine. New discovery comes in. Nitro is born. The chip comes in. But I think the fundamental thing, and I want to get your reaction to this 'cause we've put this out there on our post on Sunday. And I said, in every inflection point, I'm old enough, my birthday was yesterday. I'm old enough to know that. >> David: I saw that. >> I'm old enough to know that in the eighties, the client server shifts. Every inflection point where development changed, the methodology, the mindset or platforms change, all the apps went to the better platform. Who wants to run their application on a slower platform? And so, and those inflects. So now that's happening now, I believe. So you got better performance and I'm imagining that the app developers are coding for it. Take us through how you see that because okay, you're offering up great performance for workloads. Now it's cloud workloads. That's almost all apps. Can you comment on that? >> Well, it has been really interesting to see. I mean, as I said, we were unsure who was going to use it when we initially launched and the adoption has been amazing. Initially, obviously it's always, a lot of the startups, a lot of the more agile companies that can move a lot faster, typically a little bit smaller. They started experimenting, but the data got out there. That 40% price performance was a reality. And not only for specific workloads, it was broadly successful across a number of workloads. And so we actually just had SAP who obviously is an enormous enterprise, supporting enterprises all over the world, announced that they are going to be moving the S/4 HANA Cloud to run on Graviton2. It's just phenomenal. And we've seen enterprises of that scale and game developers, every single vertical looking to move to Graviton2 and get that 40% price performance. >> Now we have to, as analysts, we have to say, okay, how did you get to that 40%? And you have to make some assumptions obviously. And it feels like you still have some dry powder when you looked at Graviton2. I think you were running, I don't know, it's speculated anyway. I don't know if you guys, it's your data, two and a half, 2.5 gigahertz. >> David: Yeah. >> I don't know if we can share what's going on with Graviton3, but my point is you had some dry powder and now with Graviton3, quite a range of performance, 'cause it really depends on the workload. >> David: That's right. >> Maybe you could give some insight as to that. What can you share about how you tuned Graviton3? >> When we look at benchmarking, we don't want to be trying to find that benchmark that's highly tuned and then put out something that is, Hey, this is the absolute best we can get it to and that's 40%. So that 40% is actually just on average. So we just went and ran real world workloads. And we saw some that were 55%. We saw some that were 25. It depends on what it was, but on average, it was around the 35, 45%, and we said 40%. And the great thing about that is customers come back and say, Hey, we saw 40% in this workload. It wasn't that I had to tune it. And so with Graviton3, launching this week. Available in our C7g instance, we said 25%. And that is just a very standard benchmark in what we're seeing. And as we start to see more customer workloads, I think it's going to be incredible to see what that range looks like. Graviton2 for single-threaded applications, it didn't give you that much of a performance. That's what we meant by cloud applications, generally, multi-threaded. In Graviton3, that's no longer the case. So we've had some customers report up to 80% performance improvements of Graviton2 to Graviton3 when the application was more of a single-threaded application. So we started to see. (group chattering) >> You have to keep going, the time to market is compressing. So you have that, go ahead, sorry. >> No, no, I always want to add one thing on the difference between single and multi-threaded applications. A lot of legacy, you're single threaded. So this is kind of an interesting thing. So the mainframe, migration stuff, you start to see that. Is that where that comes in the whole? >> Well, a lot of the legacy apps, but also even some of the new apps, like single threading like video transcoding, for example, is all done on a single core. It's very difficult. I mean, almost impossible to do that multi-threaded way. A lot of the crypto algorithms as well, encryption and cryptography is often single core. So with Graviton3, we've seen a significant performance boost for video encoding, cryptographic algorithms, that sort of thing, which really impacts even the most modern applications. >> So that's an interesting point because now single threaded is where the vertical use cases come in. It's not like more general purpose OS kind of things. >> Yeah, and Graviton has already been very broad. I think we're just knocking down the last few verticals where maybe it didn't support it and now it absolutely does. >> And if an ISV then ports, like an SAP's ports to Graviton, then the customer doesn't see any, I mean, they're going to see the performance difference, but they don't have to think about it. >> David: Yeah. >> They just say, I choose that instance and I'm going to get better price performance. >> Exactly, so we've seen that from our ISVs. We've also been doing that with our AWS services. So services like EMR, RDS, Elastic Cache, it will be moving and making Graviton2 available for customers, which means the customer doesn't have to do the migration at all. It's all done for them. They just pick the instance and get the price performance benefits, and so yeah. >> I think, oh, no, that was serverless. Sorry. >> Well, Lambda actually just did launch on Graviton2. And I think they were talking about a 35% price performance improvement. >> Who was that? >> Lambda, a couple of months ago. >> So what does an ISV have to do to port to Graviton. >> It's relatively straightforward, and this is actually one of the things that has slowed customers down is the, wow, that must be a big migration. And that ecosystem that I spoke about is the important part. And today, with all the Linux operating systems being available for Arm running on Graviton2, with all of the container runtimes being available, and then slowly open source applications in ISV is being available. It's actually really, really easy. And we just ran the Graviton2 four-day challenge. And we did that because we actually had an enterprise migrate one of the largest production applications in just four days. Now, I probably wouldn't recommend that to most enterprises that we see is a little too fast, but they could actually do that. >> But just from a numbers standpoint, that's insanely amazing. I mean, when you think about four days. >> Yeah. >> And when we talked on virtually last year, this year, I can't remember now. You said, we'll just try it. >> David: That's right. >> And see what happens, so I presume a lot of people have tried it. >> Well, that's my advice. It's the unknown, it's the what will it take? So take a single engineer, tell them and give them a time. Say you have one week, get this running on Graviton2, and I think the results are pretty amazing, very surprised. >> We were one of the first, if not the first to say that Arm is going to be dominant in the enterprise. We know it's dominant in the Edge. And when you look at the performance curves and the time to tape out, it's just astounding. And I don't know if people appreciate that relative to the traditional Moore's law curve. I mean, it's a style. And then when you combine the power of the CPU, the GPU, the NPU, kind of what Apple does in the iPhone, it blows away the historical performance curves. And you're on that curve. >> That's right. >> I wonder if you could sort of explain that. >> So with Graviton, we're optimizing just across every single part of AWS. So one of the nice things is we actually own that end-to-end. So when it starts with the early design of Graviton2 and Graviton3, and we obviously working on other chips right now. We're actually using the cloud to do all of the electronic design automation. So we're able to test with AWS how that Graviton3 chip is going to work long before we've even started taping it out. And so those workloads are running on high-frequency CPU's on Graviton. Actually we're using Graviton to build Graviton now in the cloud. The other thing we're doing is we're making sure that the Annapurna team that's building those CPUs is deeply engaged with my team and we're going to ultimately go and build those instances so that when that chip arrives from tapeout. I'm not waiting nine months or two years, like would normally be the case, but I actually had an instance up and running within a week or two on somebody's desk studying to do the integration. And that's something we've optimized significantly to get done. And so it allows us to get that iteration time. It also allows us to be very, very accurate with our tapeouts. We're not having to go back with Graviton. They're all A1 chips. We're not having to go back and do multiple runs of these things because we can do so much validation and performance testing in the cloud ahead of time. >> This is the epiphany of the Arm model. >> It really is. >> It's a standard. When you send it to the fab, they know what's going to work. You hit volume and it's just no fab. >> Well, this is a great thread. We'll stay on this 'cause Adam told us when we met with them for re:Invent that they're seeing a lot more visibility into use cases at the scale. So the scale gives you an advantage on what instances might work. >> And makes the economics works. >> Makes the economics work, hence the timing, the shrinking time to market, not there, but also for the apps. Talk about the scale advantage you guys have. >> Absolutely. I mean, the scale advantage of AWS plays out in a number of ways for our customers. The first thing is being able to deliver highly optimized hardware. So we don't just look at the Graviton3 CPU, you were speaking about the core count and the frequency and Peter spoke about a lot of that in his keynote yesterday. But we look at how does the Graviton3 CPU work with the rest of the instance. What is the right balance between the CPU and memory? The CPU and the Hydro. What's the performance and the drive? We just launched the Nitro SSD, which is now we've actually building our own custom SSDs for Nitro getting better performance, being able to do updates, better security, making it more cloudy. We're just saying, we've been challenged with the SSD in the parts. The other place that scales really helping is in capacity. Being able to make sure that we can absorb things like the COVID spike, or the stuff you see in the financial industry with just enormous demand for compute. We can do that because of our scale. We are able to scale. And the final area is actually in quality because I have such an enormous fleet. I'm actually able to drive down AFR. So annual failure rates, are we well below what the mathematical theoretical tenant or possibility is? So if you look at what's put on that actual sticker on the box that says you should be able to get a full percent AFR. At scale and with focus, we're actually able to get that down to significantly below what the mathematical entitlement was actually be. >> Yeah, it's incredible. I've got a great, and this is the advantage, and that's why I believe anyone who's writing applications that has includes a database, data transfer, any kind of execution of code will use the stack. >> Why would they? Really, why? We've seen this, like you said before, whether it was PC, then the fastest Pentium or somebody. >> Why would you want your app to run slower? >> Unix box, right? ISVS want it to run as fast and as cheaply as possible. Now power plays into it as well. >> Yeah, well, we do have, I agree with what you're saying. We do have a number of customers that are still looking to run on x86, but obviously customers that want windows. Windows isn't available for Arm and so that's a challenge. They'll continue to do that. And you know the way we do look at it is most law kind of died out on us in 2002, 2003. And what I'm hoping is, not necessarily bringing wars a little back, but then we say, let's not accept the 10%, 15% improvement year-over-year. There's absolutely more we can all be doing. And so I'm excited to see where the x86 world's going and they doing a lot of great stuff. Intel Ice Lakes looking amazing. Milan is really great to have an AWS as well. >> Well, I'm thinking it's fair point 'cause we certainly look what Pat's doing it at Intel and he's remaking the company. I've said he's going to follow on the Arm playbook in my mind a little bit, and which is the right thing to do. So competition is a good thing. >> David: Absolutely. >> We're excited for you and a great to see Graviton and you guys have this kind of inflection point. We've been tracking for a while, but now the world's starting to see it. So congratulations to your team. >> David: Thank you. >> Just a couple of things. You guys have some news on instances. Talk about the deprecation issue and how you guys are keeping instances alive real quick. >> Yeah, we're super customer obsessed at Amazon. And so that really drives us. And one of the worst things for us to do is to have to tell a customer that we no longer supporting a service. We recently actually just deprecated the ECG classic network. I'm not sure if you saw that and that's actually off the 10 years of continuing to support it. And the only reason we did it is we have a tiny percentage of customers still using that from back in 2012. But one of the challenges is obviously instance hardware eventually will ultimately time out and fail and have hardware issues as it gets older and older. And so we didn't want to be in a place, in EC2, where we would have to constantly go to customers and say that M1 small, that C3, whatever you were running, it's no longer supported, please move. That's just a text that customers shouldn't have to do. And if they still getting value out of an older instance, let them keep using it. So we actually just announced at re:Invent, in my keynote on Tuesday, the longevity support for EC2 instances, which means we will never come back to you again and ask you to please get off an instance, because we can actually emulate all those instances on our Nitro system. And so all of these instances are starting to migrate to Nitro. You're getting all the benefits of Nitro for now some of our older zen instances, but also you don't have to worry about that work. That's just not something you need to do to get off in all the instance. >> That's great. That's a great test service. Stay on as long as you want. When you're ready to move, move. Okay, final question for you. I know we've got time, I want to get this in. The global network, you guys are known for AWS cloud WAN serve. Gives you updates on what's going on with that. >> So Werner just announced that in his keynote and over the last two to three years or so, we've seen a lot of customers starting to use the AWS backbone, which is extensive. I mean, you've seen the slides in Werner's keynote. It really does span the world. I think it's probably one of the largest networks out there. Customers starting to use that for actually their branch office communication. So instead of going and provisioning the own international MPLS networks and that sort of thing, they say, let me onboard to AWS with VPN or direct connect, and I can actually run the AWS backbone around the world. Now doing that actually has some complexity. You got to think about transit gateways. You got to think about those inter-region peering. And AWS cloud when takes all of that complexity away, you essentially create a cloud WAN, connecting to it to VPN or direct connect, and you can even go and actually set up network segments. So essentially VLANs for different parts of the organization. So super excited to get out that out of there. >> So the ease of use is the key there. >> Massively easy to use. and we have 26 SD-WAN partners. We even partnering with folks like Verizon and Swisscom in Switzerland to telco to actually allow them to use it for their customers as well. >> We'll probably use your service someday when we have a global rollout date. >> Let's do that, CUBE Global. And then the other was the M1 EC2 instance, which got a lot of applause. >> David: Absolutely. >> M1, I think it was based on A15. >> Yeah, that's for Mac. We've got to be careful 'cause M1 is our first instance as well. >> Yeah right, it's a little confusion there. >> So it's a Mac. The EC2 Mac is with M1 silicon from Apple, which super excited to put out there. >> Awesome. >> David Brown, great to see you in person. Congratulations to you and the team and all the work you guys have done over the years. And now that people starting to realize the cloud platform, the compute just gets better and better. It's a key part of the system. >> Thanks John, it's great to be here. >> Thanks for sharing. >> The SiliconANGLE is here. We're talking about custom silicon here on AWS. I'm John Furrier with Dave Vellante. You're watching theCUBE. The global leader in tech coverage. We'll be right back with more covers from re:Invent after this break. (bright music)

>>from >>around the globe. It's the Cube with digital coverage of AWS reinvent 2020 sponsored by Intel, AWS and our community partners. >>Hello and welcome back to the Cube Coverage of eight of us reinvent 2020 Virtual. I'm John for your host of the Cube. Normally we're in person this year. It's a virtual event. It is reinvent and cube virtual here. We got great interview here. Segment with VM ware and A W s. Two great guests. Keep both Cube alumni. Marc Lemire, senior vice president, general manager, The Cloud Services Business Unit VM Ware and Dave Brown, Vice president Elastic Compute Cloud easy to from Amazon Web services Gentlemen, great to see you guys. Thanks for coming on. >>Great. Thank you. Good to be back. >>Thanks. Great to be back. >>So you know, Dave, we love having you on because ec2 obviously is the core building block of a device. Once the power engine, it's the core product. And Mark, we were just talking a few months ago at VM World of momentum you guys have had on the business front. It's even mawr accelerated with co vid on the pandemic. Give us the update The partnership three years ago when Pat and Andy in San Francisco announced the partnership has been nothing but performance. Business performance, technical integration. Ah, lots happened. What's the update here for reinvent? >>Yeah, I guess the first thing I would say is look, you know, the partnership has has never been stronger. You know, as you said, uh, we announced the partnership and delivered the initial service three years ago. And I think since then, both companies have really been focused on innovating rapidly on behalf of our customers bringing together the best of the VM, or portfolio, and the best of, you know, the entire AWS. A set of capabilities. And so we've been incredibly pleased to be able to deliver those that value to our joint customers. And we look forward to continue to work very closely together. You know, across all aspects of our two companies toe continue to deliver more and more value to our joint customers. >>Well, I want to congratulate you guys at VM where, you know, we've been following that story from day one. I let a lot of people skeptical on the partnership. We were pretty bullish on it. We saw the value. It's been just been great Synergy day. I want to get your thoughts because, you know, I've always been riffing about enabling technologies and and the way it works is enabling technologies. Allow your partners to make more money, too. Right? So you guys do that with the C two, and I know that for a fact because we're doing well with our virtual event cloud, but are easy to bills are up, but who cares? We're doing well. This is the trend you guys are enabling partners, and VM Ware in particular, has a lot of customers that are on AWS. What's your perspective on all this? >>You know the part. The part maker system is so important for us, right? And we get from our customers. We have many customers who, you know, use VM ware in their own environment. They've been using it for years and years, um, true for many other software applications as well and other technologies. Andi, when they moved to AWS there very often. When you use those tools on those services on AWS is well and so you know, we we partner with many, many, many, many companies, and so it's a high priority for us. The VM Ware partnership, I think, is being sort of role model for us in terms of, you know, sitting out outside Sana goal back in 2016. I think it waas and, you know, delivering on that. Then continue to innovate on features over the last three years listening to our customers, bringing larger customers on board, giving them more advanced networking features, improving. You know that the instance types of being whereas utilizing to deliver value to their customers and most recently, obviously, with Outpost AWS outposts and parking with VM ware on VM are enabled outposts and bringing that to our customers and their own data centers. So we see the whole partner ecosystem is critically important. Way were spent a lot of time with VM and other partners on something that our customers really value. >>Mark, I want to get your thoughts on this because I was just riffing with Day Volonte about this. Um, heightened awareness with that covert 19 in the pandemic has kind of created, which is an accelerant of the value. And one >>of the >>things that's a parent is when you have this software driven and software defined kind of environment, whether it's in space or on premise or in the cloud. Um, it's the software that's driving everything, but you have to kind of components. You have the how do you operate something, And then how does the software works? So you know, it's the hand in the glove operators and software in the cloud really is becoming kind of the key things. You guys have been very successful as a company with I t operations, and now you're moving into the cloud. Can you share your thoughts on how VM Ware cloud on AWS takes that next level for your customers? So I think that's a key point that needs to be called that. What's your What's your thoughts on that? >>Yeah, I think you hit the nail on the head, and I think, you know, look, every company is on a journey to transform the level of capability they're able to offer to their customers and their employees, right? And a big part of that is how do they modernize their application environment? How do they how do they deliver new applications and services? And so this has been underway for for a while now. But if if anything, I think Cove, it has only accelerated. Um, the need for customers to be able to continue to go down that path. And so, you know, between VM ware in AWS, um, you know, we're looking to provide those customers a platform that allows them to accelerate their path to application, modernization and new services and capabilities. And, um, you know, Dave talked about the ecosystem and the importance of the ecosystem that AWS and I think you know, together. What we've been able to do if you sort of think about it, is, you know, bringing together this rich set of VM Ware services and capabilities. Um, that we've talked about before, as well as new VM Ware capabilities, for example, the ability to enable kubernetes based applications and services on top of this Corby, um or platform with Tan Xue. Right. So customers can get access to all of that is they go down this modernization path. But, you know, right next door in the same ese is 375 native AWS services that they can use together in conjunction, uh, with that environment. And so if you think about accelerating that journey right Being ableto rapidly migrate those VM ware based workloads into the AWS cloud. When you're in the AWS cloud, be able to modernize that environment using the VM Ware Tansu capability, the native AWS services and then the infrastructure that needs to come together to make that possible, for example, the network connectivity that needs to be enabled, um, to take advantage of some of those services together. Um, you know, we're really we're trying to accelerate our delivery of those capabilities so that we can help our customers accelerate the delivery of that application value thio to their customers. >>David want to get your thoughts on the trends If you speak to the customers out there at VM Ware, customers that are on the cloud because you know the sphere, for instance, very popular on the Ws Cloud with VM Ware Cloud as well as these new modern application trends like Tan Xue, Project Monterey is coming around the corner that was announced that VM world what trends do you see from the two perspective that you could share to the VM ware eight of his customers? What's the key wave right now that they should be riding on. >>Yeah, I think a few things, you know, we definitely are seeing an acceleration in customers Looking Thio looking to utilize humor on AWS You know, there was a lot of interest early on, really, over the last year, I think we've seen 140% growth in the service, which has been incredibly exciting for both of us and really shows that we we're providing customers with the service that works. You know, I think one of the key things that Mark called out just talking previously was just how simple it is for customers to move. You know, often moving to the cloud gets muddled with modernization, and it takes a long time because customers to kind of think about how do they actually make this move? Or are they stuck within their own facility on data center or they need to modernize? We moved to a different hyper visor with PM on AWS. You literally get that same environment on AWS, and so whether it's a a migration because you want to move out of your on premise facility, whether it's a migration because you want to grow and expand your facility without needing to. You know, build more data centers yourself Whether you're looking to build a d. R site on AWS on whether you looking just, you know, maybe build a new applications tank that you wanna build in a modern way, you know, using PMR in Tanzania and all the AWS services, all of those a positive we're seeing from customers. Um, you know, I think I think as the customers grow, the demand for features on being were in AWS grows as well. And we put out a number of important features to support customers that really, really large scale. And that's something that's being exciting. It's just some of the scale that we're seeing from very, very large being, we customers moving over to AWS. And so I think you know a key messages. If you have a Vienna installation today and you're thinking about moving to the cloud, it's really a little that needs to stop you in starting to move. It is is very simple to set up, and very little you have to do to your application stack to actually move it over. >>Mark, that's a great point. I want to get your thoughts on that in reaction toe. What? Dave just said Because this is kind of what you guys had said many years ago and also a VM world when we were chatting, disrupting operations just to stand up the clubs shouldn't be in place. It should be easy on you. Heard what Dave said. It's like you got >>a >>lot of cultures that are operating large infrastructure and they want to move to the cloud. But they got a mandate toe make everything. Is a services more cloud native coming. So, yeah, you gotta check off the VM where boxes and keep things running. But you gotta add more modern tooling mawr application pressure there. So there's a lot of pressure from the business units and the business models to say We gotta take advantage of the modern applications. How do you How do you look at that? >>Yeah, yeah, I mean, I think Look, making this a simple is possible is obviously a really important aspect of what we're trying Thio enable for our customers. Also, I think the speed is important, right? How you know, how can we enable them? Thio accelerate their ability to move to the cloud, but then also accelerate their ability Thio, um, deliver new services and capabilities that will differentiate their business. And then how do we, uh, kind of take some of the heavy lifting off the customers plate in terms of what it actually takes to operate and run the infrastructure and do so in a highly available way that they could depend upon for their business? And of course, delivering that full capabilities of service is a big part of that. You know, one of my when my favorite customer examples eyes a company called Stage Coach, uh, European based transportation company. And they run a network of Busses and trains, etcetera, and they actually decided to use VM. Tosto run one of their most mission critical applications, which is involved with basically scheduling, scheduling those systems right in the people that they know, the bus drivers in the train conductors etcetera. And so if you think about that application right, its's a mission critical application for them. It's also one that they need to be able to iterate involved and improve very quickly, and they were able to take advantage of a number of fairly unique capabilities of the joint service we built together to make that possible. Um, you know, the first thing that they did is they took advantage of something called stretch clusters. The M we're cloud on AWS stretch clusters Where, uh, we basically take that VM Ware environment and we stretch it. We stretch the network across to aws availability zones in the same region, Onda. Then they could basically run their applications on top of that that environment. And this is a really powerful capability because it ensures the highest levels of s L. A. For that application for four nines. In this case, if anything happens, Thio fail in one of those, uh, Aziz, we can automatically fail over and restart the application in the second ese on DSO provides this high level of availability, but they're also able to take advantage of that without on day one. Talk about keeping it simple without on day one, requiring any changes to the application of myself because that application knew how to work in the sphere. And so you know that I work in the sphere in the cloud and it can fail over on the sphere in the cloud on dso they were able to get there quickly. They're able Thio enable that application and now they're taking the next step. Which is how do I enhance and make that application even better, you know, leveraging some of the VM or capabilities also looking to take advantage of some of the native AWS capabilities. So I think that sort of speed, um you know that simplicity that helps helps customers down that path to delivering more value to their employees and their customers. That and we're really excited that were ableto offer that your customers >>just love the philosophy that both companies work back from the customer customer driven kind of mentality certainly key here to this partnership, and you can see the performance. But I think one of the differentiations that I love is that join integration thing engineering that you guys were doing together. I think that's a super valuable, differentiated VM where Dave, this is a key part of the relationship. You know, when I talked to Pat Gelsinger and and again back three years ago and he had Raghu from VM, Ware was like, This is different engineering together. What's your perspective from the West side when someone says, Yeah. Is that Riel? You know, it is easy to really kind of tied in there and his Amazon really doing joint engineering. What do you say to that? >>Oh, absolutely. Yeah, it's very real. I mean, it's been an incredible, incredible journey together, Right? Right, Right from the start, we were trying to work out how to do this back in 2016. You know, we were using some very new technology back then that we hadn't honestly released yet. Uh, the nitrous system, right? We started working with family and the nitrous system back in late 2016, and we only launched our first nitrous system enabled instance that reinvent 2017. And so we were, you know, for a year having being a run on the nitrous system, internally making sure that, you know, we would support their application and that VM Ware ran well on BC around. Well, on aws on, that's been ongoing. And, you know, the other thing I really enjoy about the relationship is learning how to best support each other's customers on on AWS and being where, and Mark is talking about stretch clusters and are being whereas, you know, utilizing the availability zones. We've done other things in terms of optimizing placement with across, you know, physical reaction in data centers. You know, Mark and the team have put forward requirements around, you know, different instance types and how they should perform invest in the Beamer environment. We've taken that back into our instance type definition and what we've released there. So it happens in a very, very low level. And I think it's both teams working together frequently, lots of meetings and then, you know, pushing each other. You know, honestly. And I think for the best experience or at the end of the day, for our joint customers. So it's been a great relationship. >>It helps when both companies are very fluent technically and pushing the envelope with technology. Both cultures, I know personally, are very strong technically, but they also customer centric. Uhm, Mark, I gotta put you on the spot on this question because this comes up every year this year more than ever. Um, is the question around VM ware on A W S and VM ware in general, and it's more of a general industry theme. But I wanna ask you because I think it relates to the US Um vm ware cloud on aws. Um, the number one question we get is how can I automate my I t operations? Because it's kind of a no brainer. Now it's kind of the genes out of the bottle. That's a mandate. But it's not always easy. Easy as it sounds to dio, you still got a lot to dio. Automation gets you level set to take advantage of some of these higher level services, and all customers want to get there fast. Ai i o t a lot of goodness in the cloud that you kinda gotta get there through kinda automating the based up first. So how did how are your customers? How are you guys helping customers automate their infrastructure operations? >>Yeah, I mean, Askew articulated right? This is a huge demand. The requirement from our customer base, right? Uh, long gone are the days that you wanna manually go into a u I and click around here, click there to make things happen, right? And so, um, you know, obviously, in addition to the core benefit of hey, we're delivering this whole thing is a service, and you don't have to worry about the hardware, the software, the life cycle all of that, Um you know, at a higher level of the stack, we're doing a lot of work to basically expose a very rich set of AP eyes. We actually have enabled that through something called the VM, or Cloud Developer center, where you can go and customer could go and understand all of the a p i s that we make available to that they can use to build on top of to effectively automated orchestrate their entire VM or cloud on AWS based infrastructure. And so that's an area we've we've invested a lot in. And at the end of the day, you know we want Thio. Both enable our customers to take their existing automation tooling that they might have been using on their VM ware based environment in their own data center. Obviously, all of that should continue to work is they bring that into the emcee aws. Um but now, once we're in AWS and we're delivering, this is a service in AWS. There's actually a higher level of automation, um that we can enable, and so you know everything that you can do through the VM or cloud console. Um, you can do through a P. I s So we've exposed roughly a piece that allow you to add or remove instance capacity ap eyes that allow you to configure the network FBI's that allow you toe effectively. Um, automate all aspects of sort of how you want Thio configure and pull together that infrastructure. Onda. You know, as Dave said, a lot of this, you know, came from some of those early just customer discussions where that was a very, very clear expectations. So, you know, we've we've been working hard. Thio make that possible. >>So can customers integrate native Cloud native technologies from AWS into APS running on VM ware cloud on any of us? >>Yeah. I mean, I'll give you one example for so we you know, we've been able to support for cloud formation right on top of the M C. Mehta best. And so that's, you know, one way that you can leverage these 80 best tools on top of on top of the m. C at best. Um and you know, as we talked about before, uh, you know everything on the VM ware in the VM ware service. We're exposing through those AP eyes. And then, of course, everything it best does has been built that way from the start. And so customers can work. Um, you know, seamlessly across those two environments. >>Great stuff. Great update. Final question for both of you. Uh, Dave will start with you. What's the unique advantages? When you people watching? That's gonna say, OK, I get it. I see the momentum. I've now got a thing about post pandemic growth strategies. I gotta fund the projects, so I'm either gonna retool while I'm waiting for the world to open up. Two. I got a tail wind. This is good for my business. I'm gonna take advantage of this. How do they modernize our application? What? The unique things with VM Ware Cloud on AWS. What's unique? What would you say? I >>mean, I think the big thing for me eyes the consistency, um, the other way that were built This between the the sphere on prime environment and the the sphere that you get on aws with BMC on aws. Um you know, when I think about modernization and honestly, any project that I do, we do it Amazon I don't like projects that required enormous amount of planning and then tooling. And then, you know, you've this massive waterfall stock project before you do anything meaningful. And what's so great about what we built here is you can start that migration almost immediately, start bringing a few applications over. And when you do that, you can start saying, Okay, where do we want to make improvements? But just by moving over to aws NBN were on AWS, you start to reap the benefits of being in the child right from day one. Many of the things Mark called out about infrastructure management and that sort of thing. But then you get to modernize off to that as well. And so just the richness in terms of, you know, being where a tan xue and then the you know, I think it's more than 200 AWS services. Now you get to bring all that into your application stack, but at a time at a at a at a cadence or time that really matters to you. But you could get going immediately, and I think that's the thing that customers ready need to do if you find yourself in a situation you know, with just how much the world's changed in the last year. Looking Thio. Modernize your applications deck, Looking for the cost benefits. Looking to maybe get out of the data center. Um, it's a relatively easy both forward and just put in a couple of engineers a couple of technicians on to actually starting to do the process. I think you'll be very surprised at how much progress you can actually make in a short amount of time. >>Mark, you're in charge of the Cloud Services business unit at VM Ware CPM. Where cloud on AWS successful more to do a lot of action kubernetes cloud native automation and the list goes on and on. What are the most unique advantages that you guys have? What would you say? >>Yeah, I mean, I would maybe just build on Dave's comments a bit. I think you know, if you look at it through the customer lens three ability to reiterate and the ability to move quickly and not being forced into sort of a one size fits all model, right? And so there may be certain applications that they run into VM, and they want to run into VM forever. Great. We could enable that there might be other applications that they want to move from a VM into a container, remove into kubernetes and do that in a very seamless way. And we can enable that with, uh, with Tan Xue, right? By the way, they may wanna actually many applications. They're gonna require, uh, complex composite applications that have some aspects of it running in communities, other aspects running on VMS. You know, other aspects connecting to some native AWS services. And so, you know, we could enable those types of, you know, incremental value that's delivered very, very quickly that allows them at the end of the day to move, move fast on behalf of their own customers and deliver more about it to them. So I think this this sort of philosophy, right that Dave talked about I think is is one of the really important things we've tried to focus on, um, together. But, you know, on behalf of our joint customers and you know that that sort of capabilities just gets richer and richer. Overtime right. Both of us are continuing to innovate, and both of us will continue to think about how we bring those services together as we innovate in our respective areas and how they need to link together as part of this This intense solution. Um, so, uh, you know that I think that you're gonna see us continue to invest, continue to move quickly. Um, continue to respond to what our customers together are asking us. Thio enable for them. >>Well, really appreciate the insight. Thanks for coming on this cube virtual, um, segment. Um, virtualization has hit the cube where we have multiple virtual stages out there at reinvent on the site. Obviously, it's a virtual event over three weeks, so it's a little bit not four days or three days. It's three weeks. So, um, if you're watching this, check out the site. Tons of good V o D. The executive leaderships Check out the keynotes that air there. It's awesome. Big news. Of course. Check out the cube coverage, but I have one final final question is you guys are leaders in the industry and within your companies, and we're virtual this year. You gotta manage your teams. You still gotta go to work every day. You gotta operate your business is a swell as work with customers. What have you guys learned? And can you share any, um, advice or observations of how to be effective as a leader, a za manager, and as a customer interface point for your companies? >>Well, I I think, uh, let me go first, then Mark Mark and had some things, you know, I think we're moving to certainly in the last year, specifically with covert. You know, we've we've we've just passed out. I think we just passed out seven months off, being remote now on, obviously doing reinvent as well. Um, it zits certainly taken some adjusting. I think we've done relatively well, um, with, you know, going virtual. We were well prepared at Amazon to go virtual, but from a leadership point of view, you know, making sure that you have been some positives, right? So for one, I have I have teams all over the world, and, uh, being virtually actually helped a lot with that. You know, everybody is virtually all on the same stage. It's not like we have a group of us in Seattle and a few others scattered around the world. Everybody's on the same cold now. on that has the same you know, be able to listen to in the same way. But I better think a lot about sort of just my own time. Personally, in the time that my team spends, I think it's been very easy for us. Thio run a little too hot waken start a little too early and run a little too late in the evenings on DSO, making sure that we protect that time. And then, obviously, from a customer point of view, you know, we found that customers are very willing to engage virtually as well around the world s Oh, that's something we've been able to utilize very well to continue to have. You know what we call our executive briefing center and do those sorts of things customer meetings on in some ways. You know, without the plane trip on either side to the other side of the world, you're able to do more of those and stay even more in contact with your customers. So it's been it's been a lot of adjustment for us. I think we've done well. I think you know, a zay said. We've had a look at Are we keeping it balanced because I think it's very easy to get out of balance and just from a time point of view. But I think I'm sure it'll show. It'll change again as the world goes back to normal. But in many ways, I think we've learned a lot of valuable lessons that I hope in some cases don't go away. I think well will probably be more virtual going forward. So that's what a bit of from my side >>creating. Yeah. Confronting hot people run hard. You can, you know, miss misfire on that and burnout gonna stay, Stay tuned. Mark your thoughts. Is leader customers defeating employees? Customers? >>Yeah. I mean, in many ways, I would say similar experience. I think, uh, I mean, if you sort of think back, right, uh, it's in many ways amazing that within the course of literally a week, right, I think about some of the BMR experience we went from, uh, you know, 90 95% of our employees, at least in the US, working in an office right to immediately all working from home. And, uh, you know, I think having the technology is available to make that possible and really? For the most part, without skipping a beat. Um, it is pretty pretty amazing, right? Um and then, you know, I think from a productivity perspective, in many ways, you know, it z increased productivity. Right? Um, they have mentioned the ability engage customers much more easily you think about in the past, you would have taken a flight to Europe to maybe meet with, you know, 5 to 10 customers and spent an entire week. And now you can do that in, you know, in the morning, right? Um, and the way we sort of engaged our teams, I think in many ways, um, sort of online, uh, can create a very, very rich experience, right? In a way to bring people together across many locations in a much more seamless way than if maybe part of the team is there in the office. And some other part of the team is trying toe connect in through resume or something else. A little bit of a fragmented experience. But if everyone's on the same platform, regardless of where you are e think we've seen some benefits from that. >>It's interesting. You see virtualization. What that did to the servers created cloud, you know. Hey, Productivity. >>You also have to be careful. You don't run those servers too hot. You >>gotta have a cooling. You got the cooling Eso I You know, this is really an interesting, you know, social, uh, equation Global phenomenon of productivity Cloud. Combined with this notion of virtual changes, the workloads, the work flows, the workplace and the workforce, right, The future work. So I think, you know, we're watching this closely. I know you guys have both had great success from the pandemic with this new pressure on the cloud, because it's a new model, a new way to do things, So we'll keep watching it. Thanks for the insight. Thanks for coming on and and enjoy the rest of reinvent. >>Great. Thank >>you. Great to be here. >>Okay, this the cubes coverage. I'm John for your host of Cuban, remember? Go to the reinvent site. Three weeks of great virtual content over this month, Of course. Cube coverage for three weeks. Stay tuned off. All the analysis and a lot of great thought leadership in the industry commentary. Stay with us throughout the month. Thank you. Yeah,

>> Announcer: From around the globe it's theCUBE with digital coverage of smart data marketplaces brought to you by Io Tahoe. >> Hi everybody, this is Dave Vellante. And welcome back. We've been talking about smart data. We've been hearing Io Tahoe talk about putting data to work and keep heart of building great data outcomes is the Cloud of course, and also Cloud native tooling. Stuti Deshpande is here. She's a partner solutions architect for Amazon Web Services and an expert in this area. Stuti, great to see you. Thanks so much for coming on theCUBE. >> Thank you so much for having me here. >> You're very welcome. So let's talk a little bit about Amazon. I mean, you have been on this machine learning journey for quite sometime. Take us through how this whole evolution has occurred in technology over the period of time. Since the Cloud really has been evolving. >> Amazon in itself is a company, an example of a company that has gotten through a multi year machine learning transformation to become the machine learning driven company that you see today. They have been improvising on original personalization model using robotics to all different women's centers, developing a forecasting system to predict the customer needs and improvising on that and reading customer expectations on convenience, fast delivery and speed, from developing natural language processing technology for end user infraction, to developing a groundbreaking technology such as Prime Air jobs to give packages to the customers. So our goal at Amazon With Services is to take this rich expertise and experience with machine learning technology across Amazon, and to work with thousands of customers and partners to handle this powerful technology into the hands of developers or data engineers of all levels. >> Great. So, okay. So if I'm a customer or a partner of AWS, give me the sales pitch on why I should choose you for machine learning. What are the benefits that I'm going to get specifically from AWS? >> Well, there are three main reasons why partners choose us. First and foremost, we provide the broadest and the deepest set of machine learning and AI services and features for your business. The velocity at which we innovate is truly unmatched. Over the last year, we launched 200 different services and features. So not only our pace is accelerating, but we provide fully managed services to our customers and partners who can easily build sophisticated AI driven applications and utilizing those fully managed services began build and train and deploy machine learning models, which is both valuable and differentiating. Secondly, we can accelerate the adoption of machine learning. So as I mentioned about fully managed services for machine learning, we have Amazon SageMaker. So SageMaker is a fully managed service that are any developer of any level or a data scientist can utilize to build complex machine learning, algorithms and models and deploy that at scale with very less effort and a very less cost. Before SageMaker, it used to take so much of time and expertise and specialization to build all these extensive models, but SageMaker, you can literally build any complex models within just a time of days or weeks. So to increase it option, AWS has acceleration programs just in a solution maps. And we also have education and training programs such as DeepRacer, which are enforces on enforcement learning and Embark, which actually help organization to adopt machine learning very readily. And we also support three major frameworks such as TensorFlow five charge, or they have separate teams who are dedicated to just focus on all these frameworks and improve the support of these frameworks for a wide variety of workloads. And finaly, we provide the most comprehensive platform that is optimized for machine learning. So when you think about machine learning, you need to have a data store where you can store your training sets, your test sets, which is highly reliable, highly scalable, and secure data store. Most of our customers want to store all of their data and any kind of data into a centralized repository that can be treated at the central source of fraud. And in this case from the Amazon Esri data store to build and endurance machine learning workflow. So we believe that we provide this capability of having the most comprehensive platform to build the machine learning workflow from internally. >> Great. Thank you for that. So I wanted, my next question is, this is a complicated situation for a lot of customers. You know, having the technology is one thing, but adoption is sort of everything. So I wonder if you could paint a picture for us and help us understand, how you're helping customers think about machine learning, thinking about that journey and maybe give us the context of what the ecosystem looks like? >> Sure. If someone can put up the belt, I would like to provide a picture representation of how AWS and fusion machine learning as three layers of stack. And moving on to next bill, I can talk about the bottom there. And bottom there as you can see over this screen, it's basically for advanced technologists advanced data scientists who are machine learning practitioners who work at the framework level. 90% of data scientists use multiple frameworks because multiple frameworks are adjusted and are suitable for multiple and different kinds of workloads. So at this layer, we provide support for all of the different types of frameworks. And the bottom layer is only for the advanced scientists and developers who are actually actually want to build, train and deploy these machine learning models by themselves and moving onto the next level, which is the middle layer. This layer is only suited for non-experts. So here we have SageMaker where it provides a fully managed service there you can build, tune, train and deploy your machine learning models at a very low cost and with very minimal efforts and at a higher scale, it removes all the complexity, heavy lifting and guesswork from this stage of machine learning and Amazon SageMaker has been the scene that will change. Many of our customers are actually standardizing on top off Amazon SageMaker. And then I'm moving on to the next layer, which is the top most layer. We call this as AI services because this may make the human recognition. So all of the services mentioned here such as Amazon Rekognition, which is basically a deep learning service optimized for image and video analysis. And then we have Amazon Polly, which can do the text to speech conversion and so on and so forth. So these are the AI services that can be embedded into the application so that the end user or the end customer can build AI driven applications. >> Love it. Okay. So you've got the experts at the bottom with the frameworks, the hardcore data scientists, you kind of get the self driving machine learning in the middle, and then you have all the ingredients. I'm like an AI chef or a machine learning chef. I can pull in vision, speech, chatbots, fraud detection, and sort of compile my own solutions that's cool. We hear a lot about SageMaker studio. I wonder if you could tell us a little bit more, can we double click a little bit on SageMaker? That seems to be a pretty important component of that stack that you just showed us. >> I think that was an absolutely very great summarization of all the different layers of machine unexpected. So thank you for providing the gist of that. Of course, I'll be really happy to talk about Amazon SageMaker because most of our customers are actually standardizing on top of SageMaker. That is spoken about how machine learning traditionally has so many complications and it's very complex and expensive and I traded process, which makes it even harder because they don't know integrated tools or if you do the traditional machine learning all kind of deployment, there are no integrated tools for the entire workflow process and deployment. And that is where SageMaker comes into the picture. SageMaker removes all the heaviness thing and complexities from each step of the deployment of machine learning workflow, how it solves our challenges by providing all of the different components that are optimized for every stage of the workflow into one single tool set. So that models get to production faster and with much less effort and at a lower cost. We really continue to add important (indistinct) leading to Amazon SageMaker. I think last year we announced 50 cubic litres in this far SageMaker being improvised it's features and functionalities. And I would love to call out a couple of those here, SageMaker notebooks, which are just one thing, the prominent notebooks that comes along with easy two instances, I'm sorry for quoting Jarvin here is Amazon Elastic Compute Instances. So you just need to have a one thing deployment and you have the entire SageMaker Notebook Interface, along with the Elastic Compute Instances running that gives you the faster time to production. If you're a machine, if you are a data scientist or a data engineer who worked extensively for machine learning, you must be aware about building training datasets is really complex. So there we have on his own ground truth, that is only for building machine learning training data sets, which can reduce your labeling cost by 70%. And if you perform machine learning and other model technology in general, there are some workflows where you need to do inferences. So there we have inference, Elastic Inference Incense, which you can reduce the cost by 75% by adding a little GP acceleration. Or you can reduce the cost by adding managed squad training, utilizing easy to spot instances. So there are multiple ways that you can reduce the costs and there are multiple ways there you can improvise and speed up your machine, learning deployment and workflow. >> So one of the things I love about, I mean, I'm a prime member who is not right. I love to shop at Amazon. And what I like about it is the consumer experience. It kind of helps me find things that maybe I wasn't aware of, maybe based on other patterns that are going on in the buying community with people that are similar. If I want to find a good book. It's always gives me great reviews and recommendations. So I'm wondering if that applies to sort of the tech world and machine learning, are you seeing any patterns emerge across the various use cases, you have such scale? What can you tell us about that? >> Sure. One of the battles that we have seen all the time is to build scalable layer for any kind of use case. So as I spoke before that as much, I'm really looking to put their data into a single set of depository where they have the single source of truth. So storing of data and any kind of data at any velocity into a single source of would actually help them build models who run on these data and get useful insights out of it. So when you speak about an entry and workflow, using Amazon SageMaker along bigger, scalable analytical tool is actually what we have seen as one of the factors where they can perform some analysis using Amazon SageMaker and build predictive models to say samples, if you want to take a healthcare use case. So they can build a predictive model that can victimize the readmissions of using Amazon SageMaker. So what I mean, to say is, by not moving data around and connecting different services to the same set of source of data, that's tumor avoid creating copies of data, which is very crucial when you are having training data set and test data sets with Amazon SageMaker. And it is highly important to consider this. So the pattern that we have seen is to utilize a central source of depository of data, which could be Amazon Extra. In this scenario, scalable analytical layer along with SageMaker. I would have to code at Intuit for a success story over here. I'm using sandwich, a Amazon SageMaker Intuit had reviews the machine learning deployment time by 90%. So I'm quoting here from six months to one week. And if you think about a healthcare industry, there hadn't been a shift from reactive to predictive care. So utilizing predictive models to accelerate research and discovery of new drugs and new treatments. And you've also observed that nurses were supported by AI tools increase their, their productivity has increased by 50%. I would like to say that one of our customers are really diving deep into the AWS portfolio of machine learning and AI services and including transcribed medical, where they are able to provide some insights so that their customers are getting benefits from them. Most of their customers are healthcare providers and they are able to give some into insights so that they can create some more personalized and improvise patient care. So there you have the end user benefits as well. One of the patterns that I have, I can speak about and what we have seen as well, appearing a predictive model with real time integration into healthcare records will actually help their healthcare provider customers for informed decision making and improvising the personalized patient care. >> That's a great example, several there. And I appreciate that. I mean, healthcare is one of those industries that is just so right for technology ingestion and transformation, that is a great example of how the cloud has really enabled really. I mean, I'm talking about major changes in healthcare with proactive versus reactive. We're talking about lower costs, better health, longer lives is really inspiring to see that evolve. We're going to watch it over the next several years. I wonder if we could close in the marketplace. I've had the pleasure of interviewing Dave McCann, a number of times. He and his team have built just an awesome capability for Amazon and its ecosystem. What about the data products, whether it's SageMaker or other data products in the marketplace, what can you tell us? >> Sure. Either of this market visits are interesting thing. So let me first talk about the AWS marketplace of what, AWS marketplace you can browse and search for hundreds of machine learning algorithms and machine learning, modern packages in a broad range of categories that this company provision, fixed analysis, voice answers, email, video, and it says predictive models and so on and so forth. And all of these models and algorithms can be deployed to a Jupiter notebook, which comes as part of the SageMaker that form. And you can integrate all of these different models and algorithms into our fully managed service, which is Amazon SageMaker to Jupiter notebooks, Sage maker, STK, and even command as well. And this experience is followed by either of those marketplace catalog and API. So you get the same benefits as any other marketplace products, the just seamless deployments and consolidate it. So you get the same benefits as the products and the invest marketplace for your machine learning algorithms and model packages. And this is really important because these can be darkly integrated into our SageMaker platform. And I don't even be honest about the data products as well. And I'm really happy to provide and code one of the example over here in the interest of cooler times and because we are in unprecedented times over here we collaborated with our partners to provide some data products. And one of them is data hub by tablet view that gives you the time series data of phases and depth data gathered from multiple trusted sources. And this is to provide better and informed knowledge so that everyone who was utilizing this product can make some informed decisions and help the community at the end. >> I love it. I love this concept of being able to access the data, algorithms, tooling. And it's not just about the data, it's being able to do something with the data and that we've been talking about injecting intelligence into those data marketplaces. That's what we mean by smart data marketplaces. Stuti Deshpande, thanks so much for coming to theCUBES here, sharing your knowledge and tell us a little bit about AWS. There's a pleasure having you. >> It's my pleasure too. Thank you so much for having me here. >> You're very welcome. And thank you for watching. Keep it right there. We will be right back right after this short break. (soft orchestral music)

>> Announcer: From around the globe it's theCUBE with digital coverage of smart data marketplaces brought to you by Io Tahoe. >> Hi everybody, this is Dave Vellante. And welcome back. We've been talking about smart data. We've been hearing Io Tahoe talk about putting data to work and keep heart of building great data outcomes is the Cloud of course, and also Cloud native tooling. Stuti Deshpande is here. She's a partner solutions architect for Amazon Web Services and an expert in this area. Stuti, great to see you. Thanks so much for coming on theCUBE. >> Thank you so much for having me here. >> You're very welcome. So let's talk a little bit about Amazon. I mean, you have been on this machine learning journey for quite sometime. Take us through how this whole evolution has occurred in technology over the period of time. Since the Cloud really has been evolving. >> Amazon in itself is a company, an example of a company that has gotten through a multi year machine learning transformation to become the machine learning driven company that you see today. They have been improvising on original personalization model using robotics to all different women's centers, developing a forecasting system to predict the customer needs and improvising on that and reading customer expectations on convenience, fast delivery and speed, from developing natural language processing technology for end user infraction, to developing a groundbreaking technology such as Prime Air jobs to give packages to the customers. So our goal at Amazon With Services is to take this rich expertise and experience with machine learning technology across Amazon, and to work with thousands of customers and partners to handle this powerful technology into the hands of developers or data engineers of all levels. >> Great. So, okay. So if I'm a customer or a partner of AWS, give me the sales pitch on why I should choose you for machine learning. What are the benefits that I'm going to get specifically from AWS? >> Well, there are three main reasons why partners choose us. First and foremost, we provide the broadest and the deepest set of machine learning and AI services and features for your business. The velocity at which we innovate is truly unmatched. Over the last year, we launched 200 different services and features. So not only our pace is accelerating, but we provide fully managed services to our customers and partners who can easily build sophisticated AI driven applications and utilizing those fully managed services began build and train and deploy machine learning models, which is both valuable and differentiating. Secondly, we can accelerate the adoption of machine learning. So as I mentioned about fully managed services for machine learning, we have Amazon SageMaker. So SageMaker is a fully managed service that are any developer of any level or a data scientist can utilize to build complex machine learning, algorithms and models and deploy that at scale with very less effort and a very less cost. Before SageMaker, it used to take so much of time and expertise and specialization to build all these extensive models, but SageMaker, you can literally build any complex models within just a time of days or weeks. So to increase it option, AWS has acceleration programs just in a solution maps. And we also have education and training programs such as DeepRacer, which are enforces on enforcement learning and Embark, which actually help organization to adopt machine learning very readily. And we also support three major frameworks that just tensive no charge, or they have separate teams who are dedicated to just focus on all these frameworks and improve the support of these frameworks for a wide variety of workloads. And finaly, we provide the most comprehensive platform that is optimized for machine learning. So when you think about machine learning, you need to have a data store where you can store your training sets, your test sets, which is highly reliable, highly scalable, and secure data store. Most of our customers want to store all of their data and any kind of data into a centralized repository that can be treated at the central source of fraud. And in this case from the Amazon Esri data store to build and endurance machine learning workflow. So we believe that we provide this capability of having the most comprehensive platform to build the machine learning workflow from internally. >> Great. Thank you for that. So I wanted, my next question is, this is a complicated situation for a lot of customers. You know, having the technology is one thing, but adoption is sort of everything. So I wonder if you could paint a picture for us and help us understand, how you're helping customers think about machine learning, thinking about that journey and maybe give us the context of what the ecosystem looks like? >> Sure. If someone can put up the belt, I would like to provide a picture representation of how AWS and fusion machine learning as three layers of stack. And moving on to next bill, I can talk about the bottom there. And bottom there as you can see over this screen, it's basically for advanced technologists advanced data scientists who are machine learning practitioners who work at the framework level. 90% of data scientists use multiple frameworks because multiple frameworks are adjusted and are suitable for multiple and different kinds of workloads. So at this layer, we provide support for all of the different types of frameworks. And the bottom layer is only for the advanced scientists and developers who are actually actually want to build, train and deploy these machine learning models by themselves and moving onto the next level, which is the middle layer. This layer is only suited for non-experts. So here we have seen Jamaica where it provides a fully managed service there you can build, tune, train and deploy your machine learning models at a very low cost and with very minimal efforts and at a higher scale, it removes all the complexity, having a thing and guess guesswork from this stage of machine learning and Amazon SageMaker has been the scene that will change. Many of our customers are actually standardizing on top off Amazon SageMaker. And then I'm moving on to the next layer, which is the top most layer. We call this as AI services because this may make the human recognition. So all of the services mentioned here such as Amazon Rekognition, which is basically a deep learning service optimized for image and video analysis. And then we have Amazon Polly, which can do the text to speech from Russian and so on and so forth. So these are the AI services that can be embedded into the application so that the end user or the end customer can build AI driven applications. >> Love it. Okay. So you've got the experts at the bottom with the frameworks, the hardcore data scientists, you kind of get the self driving machine learning in the middle, and then you have all the ingredients. I'm like an AI chef or a machine learning chef. I can pull in vision, speech, chatbots, fraud detection, and sort of compile my own solutions that's cool. We hear a lot about SageMaker studio. I wonder if you could tell us a little bit more, can we double click a little bit on SageMaker? That seems to be a pretty important component of that stack that you just showed us. >> I think that was an absolutely very great summarization of all the different layers of machine unexpected. So thank you for providing the gist of that. Of course, I'll be really happy to talk about Amazon SageMaker because most of our customers are actually standardizing on top of SageMaker. That is spoken about how machine learning traditionally has so many complications and it's very complex and expensive and I traded process, which makes it even harder because they don't know integrated tools or if you do the traditional machine learning all kind of deployment, there are no integrated tools for the entire workflow process and deployment. And that is where SageMaker comes into the picture. SageMaker removes all the heaviness thing and complexities from each step of the deployment of machine learning workflow, how it solves our challenges by providing all of the different components that are optimized for every stage of the workflow into one single tool set. So that models get to production faster and with much less effort and at a lower cost. We really continue to add important (indistinct) leading to Amazon SageMaker. I think last year we announced 50 cubic litres in this far SageMaker being improvised it's features and functionalities. And I would love to call out a couple of those here, SageMaker notebooks, which are just one thing, the prominent notebooks that comes along with easy two instances, I'm sorry for quoting Jarvin here is Amazon Elastic Compute Instances. So you just need to have a one thing deployment and you have the entire SageMaker Notebook Interface, along with the Elastic Compute Instances running that gives you the faster time to production. If you're a machine, if you are a data scientist or a data engineer who worked extensively for machine learning, you must be aware about building training datasets is really complex. So there we have on his own ground truth, that is only for building machine learning training data sets, which can reduce your labeling cost by 70%. And if you perform machine learning and other model technology in general, there are some workflows where you need to do inferences. So there we have inference, Elastic Inference Incense, which you can reduce the cost by 75% by adding a little GP acceleration. Or you can reduce the cost by adding managed squad training, utilizing easy to spot instances. So there are multiple ways that you can reduce the costs and there are multiple ways there you can improvise and speed up your machine, learning deployment and workflow. >> So one of the things I love about, I mean, I'm a prime member who is not right. I love to shop at Amazon. And what I like about it is the consumer experience. It kind of helps me find things that maybe I wasn't aware of, maybe based on other patterns that are going on in the buying community with people that are similar. If I want to find a good book. It's always gives me great reviews and recommendations. So I'm wondering if that applies to sort of the tech world and machine learning, are you seeing any patterns emerge across the various use cases, you have such scale? What can you tell us about that? >> Sure. One of the battles that we have seen all the time is to build scalable layer for any kind of use case. So as I spoke before that as much, I'm really looking to put their data into a single set of depository where they have the single source of truth. So storing of data and any kind of data at any velocity into a single source of would actually help them build models who run on these data and get useful insights out of it. So when you speak about an entry and workflow, using Amazon SageMaker along bigger, scalable analytical tool is actually what we have seen as one of the factors where they can perform some analysis using Amazon SageMaker and build predictive models to say samples, if you want to take a healthcare use case. So they can build a predictive model that can victimize the readmissions of using Amazon SageMaker. So what I mean, to say is, by not moving data around and connecting different services to the same set of source of data, that's tumor avoid creating copies of data, which is very crucial when you are having training data set and test data sets with Amazon SageMaker. And it is highly important to consider this. So the pattern that we have seen is to utilize a central source of depository of data, which could be Amazon Extra. In this scenario, scalable analytical layer along with SageMaker. I would have to code at Intuit for a success story over here. I'm using sandwich, a Amazon SageMaker Intuit had reviews the machine learning deployment time by 90%. So I'm quoting here from six months to one week. And if you think about a healthcare industry, there hadn't been a shift from reactive to predictive care. So utilizing predictive models to accelerate research and discovery of new drugs and new treatments. And you've also observed that nurses were supported by AI tools increase their, their productivity has increased by 50%. I would like to say that one of our customers are really diving deep into the AWS portfolio of machine learning and AI services and including transcribed medical, where they are able to provide some insights so that their customers are getting benefits from them. Most of their customers are healthcare providers and they are able to give some into insights so that they can create some more personalized and improvise patient care. So there you have the end user benefits as well. One of the patterns that I have, I can speak about and what we have seen as well, appearing a predictive model with real time integration into healthcare records will actually help their healthcare provider customers for informed decision making and improvising the personalized patient care. >> That's a great example, several there. And I appreciate that. I mean, healthcare is one of those industries that is just so right for technology ingestion and transformation, that is a great example of how the cloud has really enabled really. I mean, I'm talking about major changes in healthcare with proactive versus reactive. We're talking about lower costs, better health, longer lives is really inspiring to see that evolve. We're going to watch it over the next several years. I wonder if we could close in the marketplace. I've had the pleasure of interviewing Dave McCann, a number of times. He and his team have built just an awesome capability for Amazon and its ecosystem. What about the data products, whether it's SageMaker or other data products in the marketplace, what can you tell us? >> Sure. Either of this market visits are interesting thing. So let me first talk about the AWS marketplace of what, AWS marketplace you can browse and search for hundreds of machine learning algorithms and machine learning, modern packages in a broad range of categories that this company provision, fixed analysis, voice answers, email, video, and it says predictive models and so on and so forth. And all of these models and algorithms can be deployed to a Jupiter notebook, which comes as part of the SageMaker that form. And you can integrate all of these different models and algorithms into our fully managed service, which is Amazon SageMaker to Jupiter notebooks, Sage maker, STK, and even command as well. And this experience is followed by either of those marketplace catalog and API. So you get the same benefits as any other marketplace products, the just seamless deployments and consolidate it. So you get the same benefits as the products and the invest marketplace for your machine learning algorithms and model packages. And this is really important because these can be darkly integrated into our SageMaker platform. And I don't even be honest about the data products as well. And I'm really happy to provide and code one of the example over here in the interest of cooler times and because we are in unprecedented times over here we collaborated with our partners to provide some data products. And one of them is data hub by tablet view that gives you the time series data of phases and depth data gathered from multiple trusted sources. And this is to provide better and informed knowledge so that everyone who was utilizing this product can make some informed decisions and help the community at the end. >> I love it. I love this concept of being able to access the data, algorithms, tooling. And it's not just about the data, it's being able to do something with the data and that we've been talking about injecting intelligence into those data marketplaces. That's what we mean by smart data marketplaces. Stuti Deshpande, thanks so much for coming to theCUBES here, sharing your knowledge and tell us a little bit about AWS. There's a pleasure having you. >> It's my pleasure too. Thank you so much for having me here. >> You're very welcome. And thank you for watching. Keep it right there. We will be right back right after this short break. (soft orchestral music)