>>Hey, welcome back everyone. Day three of eight of us Reinvent 2022. I'm John Farmer with Dave Volante, co-host the q Dave. 10 years for us, the leader in high tech coverage is our slogan. Now 10 years of reinvent day. We've been to every single one except with the original, which we would've come to if Amazon actually marketed the event, but they didn't. It's more of a customer event. This is day three. Is the machine learning ai keynote sws up there. A lot of announcements. We're gonna break this down. We got, we got Andy Thra here, vice President, prince Constellation Research. Andy, great to see you've been on the cube before one of our analysts bringing the, bringing the, the analysis, commentary to the keynote. This is your wheelhouse. Ai. What do you think about Swami up there? I mean, he's awesome. We love him. Big fan Oh yeah. Of of the Cuban we're fans of him, but he got 13 announcements. >>A lot. A lot, >>A lot. >>So, well some of them are, first of all, thanks for having me here and I'm glad to have both of you on the same show attacking me. I'm just kidding. But some of the announcement really sort of like a game changer announcements and some of them are like, meh, you know, just to plug in the holes what they have and a lot of golf claps. Yeah. Meeting today. And you could have also noticed that by, when he was making the announcements, you know, the, the, the clapping volume difference, you could say, which is better, right? But some of the announcements are, are really, really good. You know, particularly we talked about, one of that was Microsoft took that out of, you know, having the open AI in there, doing the large language models. And then they were going after that, you know, having the transformer available to them. And Amazon was a little bit weak in the area, so they couldn't, they don't have a large language model. So, you know, they, they are taking a different route saying that, you know what, I'll help you train the large language model by yourself, customized models. So I can provide the necessary instance. I can provide the instant volume, memory, the whole thing. Yeah. So you can train the model by yourself without depending on them kind >>Of thing. So Dave and Andy, I wanna get your thoughts cuz first of all, we've been following Amazon's deep bench on the, on the infrastructure pass. They've been doing a lot of machine learning and ai, a lot of data. It just seems that the sentiment is that there's other competitors doing a good job too. Like Google, Dave. And I've heard folks in the hallway, even here, ex Amazonians saying, Hey, they're train their models on Google than they bring up the SageMaker cuz it's better interface. So you got, Google's making a play for being that data cloud. Microsoft's obviously putting in a, a great kind of package to kind of make it turnkey. How do they really stand versus the competition guys? >>Good question. So they, you know, each have their own uniqueness and the we variation that take it to the field, right? So for example, if you were to look at it, Microsoft is known for as industry or later things that they are been going after, you know, industry verticals and whatnot. So that's one of the things I looked here, you know, they, they had this omic announcement, particularly towards that healthcare genomics space. That's a huge space for hpz related AIML applications. And they have put a lot of things in together in here in the SageMaker and in the, in their models saying that, you know, how do you, how do you use this transmit to do things like that? Like for example, drug discovery, for genomics analysis, for cancer treatment, the whole, right? That's a few volumes of data do. So they're going in that healthcare area. Google has taken a different route. I mean they want to make everything simple. All I have to do is I gotta call an api, give what I need and then get it done. But Amazon wants to go at a much deeper level saying that, you know what? I wanna provide everything you need. You can customize the whole thing for what you need. >>So to me, the big picture here is, and and Swami references, Hey, we are a data company. We started, he talked about books and how that informed them as to, you know, what books to place front and center. Here's the, here's the big picture. In my view, companies need to put data at the core of their business and they haven't, they've generally put humans at the core of their business and data. And now machine learning are at the, at the outside and the periphery. Amazon, Google, Microsoft, Facebook have put data at their core. So the question is how do incumbent companies, and you mentioned some Toyota Capital One, Bristol Myers Squibb, I don't know, are those data companies, you know, we'll see, but the challenge is most companies don't have the resources as you well know, Andy, to actually implement what Google and Facebook and others have. >>So how are they gonna do that? Well, they're gonna buy it, right? So are they gonna build it with tools that's kind of like you said the Amazon approach or are they gonna buy it from Microsoft and Google, I pulled some ETR data to say, okay, who are the top companies that are showing up in terms of spending? Who's spending with whom? AWS number one, Microsoft number two, Google number three, data bricks. Number four, just in terms of, you know, presence. And then it falls down DataRobot, Anaconda data icu, Oracle popped up actually cuz they're embedding a lot of AI into their products and, and of course IBM and then a lot of smaller companies. But do companies generally customers have the resources to do what it takes to implement AI into applications and into workflows? >>So a couple of things on that. One is when it comes to, I mean it's, it's no surprise that the, the top three or the hyperscalers, because they all want to bring their business to them to run the specific workloads on the next biggest workload. As you was saying, his keynote are two things. One is the A AIML workloads and the other one is the, the heavy unstructured workloads that he was talking about. 80%, 90% of the data that's coming off is unstructured. So how do you analyze that? Such as the geospatial data. He was talking about the volumes of data you need to analyze the, the neural deep neural net drug you ought to use, only hyperscale can do it, right? So that's no wonder all of them on top for the data, one of the things they announced, which not many people paid attention, there was a zero eight L that that they talked about. >>What that does is a little bit of a game changing moment in a sense that you don't have to, for example, if you were to train the data, data, if the data is distributed everywhere, if you have to bring them all together to integrate it, to do that, it's a lot of work to doing the dl. So by taking Amazon, Aurora, and then Rich combine them as zero or no ETL and then have Apaches Apaches Spark applications run on top of analytical applications, ML workloads. That's huge. So you don't have to move around the data, use the data where it is, >>I, I think you said it, they're basically filling holes, right? Yeah. They created this, you know, suite of tools, let's call it. You might say it's a mess. It's not a mess because it's, they're really powerful but they're not well integrated and now they're starting to take the seams as I say. >>Well yeah, it's a great point. And I would double down and say, look it, I think that boring is good. You know, we had that phase in Kubernetes hype cycle where it got boring and that was kind of like, boring is good. Boring means we're getting better, we're invisible. That's infrastructure that's in the weeds, that's in between the toes details. It's the stuff that, you know, people we have to get done. So, you know, you look at their 40 new data sources with data Wrangler 50, new app flow connectors, Redshift Auto Cog, this is boring. Good important shit Dave. The governance, you gotta get it and the governance is gonna be key. So, so to me, this may not jump off the page. Adam's keynote also felt a little bit of, we gotta get these gaps done in a good way. So I think that's a very positive sign. >>Now going back to the bigger picture, I think the real question is can there be another independent cloud data cloud? And that's the, to me, what I try to get at my story and you're breaking analysis kind of hit a home run on this, is there's interesting opportunity for an independent data cloud. Meaning something that isn't aws, that isn't, Google isn't one of the big three that could sit in. And so let me give you an example. I had a conversation last night with a bunch of ex Amazonian engineering teams that left the conversation was interesting, Dave. They were like talking, well data bricks and Snowflake are basically batch, okay, not transactional. And you look at Aerospike, I can see their booth here. Transactional data bases are hot right now. Streaming data is different. Confluence different than data bricks. Is data bricks good at hosting? >>No, Amazon's better. So you start to see these kinds of questions come up where, you know, data bricks is great, but maybe not good for this, that and the other thing. So you start to see the formation of swim lanes or visibility into where people might sit in the ecosystem, but what came out was transactional. Yep. And batch the relationship there and streaming real time and versus you know, the transactional data. So you're starting to see these new things emerge. Andy, what do you, what's your take on this? You're following this closely. This seems to be the alpha nerd conversation and it all points to who's gonna have the best data cloud, say data, super clouds, I call it. What's your take? >>Yes, data cloud is important as well. But also the computational that goes on top of it too, right? Because when, when the data is like unstructured data, it's that much of a huge data, it's going to be hard to do that with a low model, you know, compute power. But going back to your data point, the training of the AIML models required the batch data, right? That's when you need all the, the historical data to train your models. And then after that, when you do inference of it, that's where you need the streaming real time data that's available to you too. You can make an inference. One of the things, what, what they also announced, which is somewhat interesting, is you saw that they have like 700 different instances geared towards every single workload. And there are some of them very specifically run on the Amazon's new chip. The, the inference in two and theran tr one chips that basically not only has a specific instances but also is run on a high powered chip. And then if you have that data to support that, both the training as well as towards the inference, the efficiency, again, those numbers have to be proven. They claim that it could be anywhere between 40 to 60% faster. >>Well, so a couple things. You're definitely right. I mean Snowflake started out as a data warehouse that was simpler and it's not architected, you know, in and it's first wave to do real time inference, which is not now how, how could they, the other second point is snowflake's two or three years ahead when it comes to governance, data sharing. I mean, Amazon's doing what always does. It's copying, you know, it's customer driven. Cuz they probably walk into an account and they say, Hey look, what's Snowflake's doing for us? This stuff's kicking ass. And they go, oh, that's a good idea, let's do that too. You saw that with separating compute from storage, which is their tiering. You saw it today with extending data, sharing Redshift, data sharing. So how does Snowflake and data bricks approach this? They deal with ecosystem. They bring in ecosystem partners, they bring in open source tooling and that's how they compete. I think there's unquestionably an opportunity for a data cloud. >>Yeah, I think, I think the super cloud conversation and then, you know, sky Cloud with Berkeley Paper and other folks talking about this kind of pre, multi-cloud era. I mean that's what I would call us right now. We are, we're kind of in the pre era of multi-cloud, which by the way is not even yet defined. I think people use that term, Dave, to say, you know, some sort of magical thing that's happening. Yeah. People have multiple clouds. They got, they, they end up by default, not by design as Dell likes to say. Right? And they gotta deal with it. So it's more of they're inheriting multiple cloud environments. It's not necessarily what they want in the situation. So to me that is a big, big issue. >>Yeah, I mean, again, going back to your snowflake and data breaks announcements, they're a data company. So they, that's how they made their mark in the market saying that, you know, I do all those things, therefore you have, I had to have your data because it's a seamless data. And, and Amazon is catching up with that with a lot of that announcements they made, how far it's gonna get traction, you know, to change when I to say, >>Yeah, I mean to me, to me there's no doubt about Dave. I think, I think what Swamee is doing, if Amazon can get corner the market on out of the box ML and AI capabilities so that people can make it easier, that's gonna be the end of the day tell sign can they fill in the gaps. Again, boring is good competition. I don't know mean, mean I'm not following the competition. Andy, this is a real question mark for me. I don't know where they stand. Are they more comprehensive? Are they more deeper? Are they have deeper services? I mean, obviously shows to all the, the different, you know, capabilities. Where, where, where does Amazon stand? What's the process? >>So what, particularly when it comes to the models. So they're going at, at a different angle that, you know, I will help you create the models we talked about the zero and the whole data. We'll get the data sources in, we'll create the model. We'll move the, the whole model. We are talking about the ML ops teams here, right? And they have the whole functionality that, that they built ind over the year. So essentially they want to become the platform that I, when you come in, I'm the only platform you would use from the model training to deployment to inference, to model versioning to management, the old s and that's angle they're trying to take. So it's, it's a one source platform. >>What about this idea of technical debt? Adrian Carro was on yesterday. John, I know you talked to him as well. He said, look, Amazon's Legos, you wanna buy a toy for Christmas, you can go out and buy a toy or do you wanna build a, to, if you buy a toy in a couple years, you could break and what are you gonna do? You're gonna throw it out. But if you, if you, if part of your Lego needs to be extended, you extend it. So, you know, George Gilbert was saying, well, there's a lot of technical debt. Adrian was countering that. Does Amazon have technical debt or is that Lego blocks analogy the right one? >>Well, I talked to him about the debt and one of the things we talked about was what do you optimize for E two APIs or Kubernetes APIs? It depends on what team you're on. If you're on the runtime gene, you're gonna optimize for Kubernetes, but E two is the resources you want to use. So I think the idea of the 15 years of technical debt, I, I don't believe that. I think the APIs are still hardened. The issue that he brings up that I think is relevant is it's an end situation, not an or. You can have the bag of Legos, which is the primitives and build a durable application platform, monitor it, customize it, work with it, build it. It's harder, but the outcome is durability and sustainability. Building a toy, having a toy with those Legos glued together for you, you can get the play with, but it'll break over time. Then you gotta replace it. So there's gonna be a toy business and there's gonna be a Legos business. Make your own. >>So who, who are the toys in ai? >>Well, out of >>The box and who's outta Legos? >>The, so you asking about what what toys Amazon building >>Or, yeah, I mean Amazon clearly is Lego blocks. >>If people gonna have out the box, >>What about Google? What about Microsoft? Are they basically more, more building toys, more solutions? >>So Google is more of, you know, building solutions angle like, you know, I give you an API kind of thing. But, but if it comes to vertical industry solutions, Microsoft is, is is ahead, right? Because they have, they have had years of indu industry experience. I mean there are other smaller cloud are trying to do that too. IBM being an example, but you know, the, now they are starting to go after the specific industry use cases. They think that through, for example, you know the medical one we talked about, right? So they want to build the, the health lake, security health lake that they're trying to build, which will HIPPA and it'll provide all the, the European regulations, the whole line yard, and it'll help you, you know, personalize things as you need as well. For example, you know, if you go for a certain treatment, it could analyze you based on your genome profile saying that, you know, the treatment for this particular person has to be individualized this way, but doing that requires a anomalous power, right? So if you do applications like that, you could bring in a lot of the, whether healthcare, finance or what have you, and then easy for them to use. >>What's the biggest mistake customers make when it comes to machine intelligence, ai, machine learning, >>So many things, right? I could start out with even the, the model. Basically when you build a model, you, you should be able to figure out how long that model is effective. Because as good as creating a model and, and going to the business and doing things the right way, there are people that they leave the model much longer than it's needed. It's hurting your business more than it is, you know, it could be things like that. Or you are, you are not building a responsibly or later things. You are, you are having a bias and you model and are so many issues. I, I don't know if I can pinpoint one, but there are many, many issues. Responsible ai, ethical ai. All >>Right, well, we'll leave it there. You're watching the cube, the leader in high tech coverage here at J three at reinvent. I'm Jeff, Dave Ante. Andy joining us here for the critical analysis and breaking down the commentary. We'll be right back with more coverage after this short break.

(upbeat music) >> Welcome back to theCUBE's coverage of HPE GreenLake announcements. We're seeing the transition of Hewlett Packard Enterprise as a company, yes they're going all in for as a service, but we're also seeing a transition from a hardware company to what I look at increasingly as a data management company. We're going to talk today to Vishal Lall who's GreenLake cloud services solutions at HPE and Matt Maccaux who's a global field CTO, Ezmeral Software at HPE. Gents welcome back to theCube. Good to see you again. >> Thank you for having us here. >> Thanks Dave. >> So Vishal let's start with you. What are the big mega trends that you're seeing in data? When you talk to customers, when you talk to partners, what are they telling you? What's your optic say? >> Yeah, I mean, I would say the first thing is data is getting even more important. It's not that data hasn't been important for enterprises, but as you look at the last, I would say 24 to 36 months has become really important, right? And it's become important because customers look at data and they're trying to stitch data together across different sources, whether it's marketing data, it's supply chain data, it's financial data. And they're looking at that as a source of competitive advantage. So, customers were able to make sense out of the data, enterprises that are able to make sense out of that data, really do have a competitive advantage, right? And they actually get better business outcomes. So that's really important, right? If you start looking at, where we are from an analytics perspective, I would argue we are in maybe the third generation of data analytics. Kind of the first one was in the 80's and 90's with data warehousing kind of EDW. A lot of companies still have that, but think of Teradata, right? The second generation more in the 2000's was around data lakes, right? And that was all about Hadoop and others, and really the difference between the first and the second generation was the first generation was more around structured data, right? Second became more about unstructured data, but you really couldn't run transactions on that data. And I would say, now we are entering this third generation, which is about data lake houses, right? Customers what they want really is, or enterprises, what they want really is they want structured data. They want unstructured data altogether. They want to run transactions on them, right? They want to use the data to mine it for machine learning purposes, right? Use it for SQL as well as non-SQL, right? And that's kind of where we are today. So, that's really what we are hearing from our customers in terms of at least the top trends. And that's how we are thinking about our strategy in context of those trends. >> So lake house use that term. It's an increasing popular term. It connotes, "Okay, I've got the best of data warehouse "and I've got the best of data lake. "I'm going to try to simplify the data warehouse. "And I'm going to try to clean up the data swamp "if you will." Matt, so, talk a little bit more about what you guys are doing specifically and what that means for your customers. >> Well, what we think is important is that there has to be a hybrid solution, that organizations are going to build their analytics. They're going to deploy algorithms, where the data either is being produced or where it's going to be stored. And that could be anywhere. That could be in the trunk of a vehicle. It could be in a public cloud or in many cases, it's on-premises in the data center. And where organizations struggle is they feel like they have to make a choice and a trade-off going from one to the other. And so what HPE is offering is a way to unify the experiences of these different applications, workloads, and algorithms, while connecting them together through a fabric so that the experience is tied together with consistent, security policies, not having to refactor your applications and deploying tools like Delta lake to ensure that the organization that needs to build a data product in one cloud or deploy another data product in the trunk of an automobile can do so. >> So, Vishal I wonder if we could talk about some of the patterns that you're seeing with customers as you go to deploy solutions. Are there other industry patterns? Are there any sort of things you can share that you're discerning? >> Yeah, no, absolutely. As we kind of hear back from our customers across industries, I think the problem sets are very similar, right? Whether you look at healthcare customers. You look at telco customers, you look at consumer goods, financial services, they're all quite similar. I mean, what are they looking for? They're looking for making sense, making business value from the data, breaking down the silos that I think Matt spoke about just now, right? How do I stitch intelligence across my data silos to get more business intelligence out of it. They're looking for openness. I think the problem that's happened is over time, people have realized that they are locked in with certain vendors or certain technologies. So, they're looking for openness and choice. So that's an important one that we've at least heard back from our customers. The other one is just being able to run machine learning on algorithms on the data. I think that's another important one for them as well. And I think the last one I would say is, TCO is important as customers over the last few years have realized going to public cloud is starting to become quite expensive, to run really large workloads on public cloud, especially as they want to egress data. So, cost performance, trade offs are starting to become really important and starting to enter into the conversation now. So, I would say those are some of the key things and themes that we are hearing from customers cutting across industries. >> And you talked to Matt about basically being able to essentially leave the data where it belongs, bring the compute to data. We talk about that all the time. And so that has to include on-prem, it's got to include the cloud. And I'm kind of curious on the edge, where you see that 'cause that's... Is that an eventual piece? Is that something that's actually moving in parallel? There's lot of fuzziness as an observer in the edge. >> I think the edge is driving the most interesting use cases. The challenge up until recently has been, well, I think it's always been connectivity, right? Whether we have poor connection, little connection or no connection, being able to asynchronously deploy machine learning jobs into some sort of remote location. Whether it's a very tiny edge or it's a very large edge, like a factory floor, the challenge as Vishal mentioned is that if we're going to deploy machine learning, we need some sort of consistency of runtime to be able to execute those machine learning models. Yes, we need consistent access to data, but consistent access in terms of runtime is so important. And I think Hadoop got us started down this path, the ability to very efficiently and cost-effectively run large data jobs against large data sets. And it attempted to work into the source ecosystem, but because of the monolithic deployment, the tightly coupling of the compute and the data, it never achieved that cloud native vision. And so what as role in HPE through GreenLake services is delivering with open source-based Kubernetes, open source Apache Spark, open source Delta lake libraries, those same cloud native services that you can develop on your workstation, deploy in your data center in the same way you deploy through automation out at the edge. And I think that is what's so critical about what we're going to see over the next couple of years. The edge is driving these use cases, but it's consistency to build and deploy those machine learning models and connect it consistently with data that's what's going to drive organizations to success. >> So you're saying you're able to decouple, to compute from the storage. >> Absolutely. You wouldn't have a cloud if you didn't decouple compute from storage. And I think this is sort of the demise of Hadoop was forcing that coupling. We have high-speed networks now. Whether I'm in a cloud or in my data center, even at the edge, I have high-performance networks, I can now do distributed computing and separate compute from storage. And so if I want to, I can have high-performance compute for my really data intensive applications and I can have cost-effective storage where I need to. And by separating that off, I can now innovate at the pace of those individual tools in that opensource ecosystem. >> So, can I stay on this for a second 'cause you certainly saw Snowflake popularize that, they were kind of early on. I don't know if they're the first, but they certainly one of the most successful. And you saw Amazon Redshift copied it. And Redshift was kind of a bolt on. What essentially they did is they teared off. You could never turn off the compute. You still had to pay for a little bit compute, that's kind of interesting. Snowflakes at the t-shirt sizes, so there's trade offs there. There's a lot of ways to skin the cat. How did you guys skin the cat? >> What we believe we're doing is we're taking the best of those worlds. Through GreenLake cloud services, the ability to pay for and provision on demand the computational services you need. So, if someone needs to spin up a Delta lake job to execute a machine learning model, you spin up that. We're of course spinning that up behind the scenes. The job executes, it spins down, and you only pay for what you need. And we've got reserve capacity there. So you, of course, just like you would in the public cloud. But more importantly, being able to then extend that through a fabric across clouds and edge locations, so that if a customer wants to deploy in some public cloud service, like we know we're going to, again, we're giving that consistency across that, and exposing it through an S3 API. >> So, Vishal at the end of the day, I mean, I love to talk about the plumbing and the tech, but the customer doesn't care, right? They want the lowest cost. They want the fastest outcome. They want the greatest value. My question is, how are you seeing data organizations evolve to sort of accommodate this third era of this next generation? >> Yeah. I mean, the way at least, kind of look at, from a customer perspective, what they're trying to do is first of all, I think Matt addressed it somewhat. They're looking at a consistent experience across the different groups of people within the company that do something to data, right? It could be a SQL users. People who's just writing a SQL code. It could be people who are writing machine learning models and running them. It could be people who are writing code in Spark. Right now they are, you know the experience is completely disjointed across them, across the three types of users or more. And so that's one thing that they trying to do, is just try to get that consistency. We spoke about performance. I mean the disjointedness between compute and storage does provide the agility, because there customers are looking for elasticity. How can I have an elastic environment? So, that's kind of the other thing they're looking at. And performance and DCU, I think a big deal now. So, I think that that's definitely on a customer's mind. So, as enterprises are looking at their data journey, those are the at least the attributes that they are trying to hit as they organize themselves to make the most out of the data. >> Matt, you and I have talked about this sort of trend to the decentralized future. We're sort of hitting on that. And whether it's in a first gen data warehouse, second gen data lake, data hub, bucket, whatever, that essentially should ideally stay where it is, wherever it should be from a performance standpoint, from a governance standpoint and a cost perspective, and just be a node on this, I like the term data mesh, but be a node on that, and essentially allow the business owners, those with domain context to you've mentioned data products before to actually build data products, maybe air quotes, but a data product is something that can be monetized. Maybe it cuts costs. Maybe it adds value in other ways. How do you see HPE fitting into that long-term vision which we know is going to take some time to play out? >> I think what's important for organizations to realize is that they don't have to go to the public cloud to get that experience they're looking for. Many organizations are still reluctant to push all of their data, their critical data, that is going to be the next way to monetize business into the public cloud. And so what HPE is doing is bringing the cloud to them. Bringing that cloud from the infrastructure, the virtualization, the containerization, and most importantly, those cloud native services. So, they can do that development rapidly, test it, using those open source tools and frameworks we spoke about. And if that model ends up being deployed on a factory floor, on some common X86 infrastructure, that's okay, because the lingua franca is Kubernetes. And as Vishal mentioned, Apache Spark, these are the common tools and frameworks. And so I want organizations to think about this unified analytics experience, where they don't have to trade off security for cost, efficiency for reliability. HPE through GreenLake cloud services is delivering all of that where they need to do it. >> And what about the speed to quality trade-off? Have you seen that pop up in customer conversations, and how are organizations dealing with that? >> Like I said, it depends on what you mean by speed. Do you mean a computational speed? >> No, accelerating the time to insights, if you will. We've got to go faster, faster, agile to the data. And it's like, "Whoa, move fast break things. "Whoa, whoa. "What about data quality and governance and, right?" They seem to be at odds. >> Yeah, well, because the processes are fundamentally broken. You've got a developer who maybe is able to spin up an instance in the public cloud to do their development, but then to actually do model training, they bring it back on-premises, but they're waiting for a data engineer to get them the data available. And then the tools to be provisioned, which is some esoteric stack. And then runtime is somewhere else. The entire process is broken. So again, by using consistent frameworks and tools, and bringing that computation to where the data is, and sort of blowing this construct of pipelines out of the water, I think is what is going to drive that success in the future. A lot of organizations are not there yet, but that's I think aspirationally where they want to be. >> Yeah, I think you're right. I think that is potentially an answer as to how you, not incrementally, but revolutionized sort of the data business. Last question, is talking about GreenLake, how this all fits in. Why GreenLake? Why do you guys feel as though it's differentiable in the market place? >> So, I mean, something that you asked earlier as well, time to value, right? I think that's a very important attribute and kind of a design factor as we look at GreenLake. If you look at GreenLake overall, kind of what does it stand for? It stands for experience. How do we make sure that we have the right experience for the users, right? We spoke about it in context of data. How do we have a similar experience for different users of data, but just broadly across an enterprise? So, it's all about experience. How do you automate it, right? How do you automate the workloads? How do you provision fast? How do you give folks a cloud... An experience that they have been used to in the public cloud, on using an Apple iPhone? So it's all about experience, I think that's number one. Number two is about choice and openness. I mean, as we look at GreenLake is not a proprietary platform. We are very, very clear that the design, one of the important design principles is about choice and openness. And that's the reason we are, you hear us talk about Kubernetes, about Apaches Spark, about Delta lake et cetera, et cetera, right? We're using kind of those open source models where customers have a choice. If they don't want to be on GreenLake, they can go to public cloud tomorrow. Or they can run in our Holos if they want to do it that way or in their Holos, if they want to do it. So they should have the choice. Third is about performance. I mean, what we've done is it's not just about the software, but we as a company know how to configure infrastructure for that workload. And that's an important part of it. I mean if you think about the machine learning workloads, we have the right Nvidia chips that accelerate those transactions. So, that's kind of the last, the third one, and the last one, I think, as I spoke about earlier is cost. We are very focused on TCO, but from a customer perspective, we want to make sure that we are giving a value proposition, which is just not about experience and performance and openness, but also about costs. So if you think about GreenLake, that's kind of the value proposition that we bring to our customers across those four dimensions. >> Guys, great conversation. Thanks so much, really appreciate your time and insights. >> Matt: Thanks for having us here, David. >> All right, you're welcome. And thank you for watching everybody. Keep it right there for more great content from HPE GreenLake announcements. You're watching theCUBE. (upbeat music)