Cloud AI Journey With HPE | Episode 4: Insights From theCUBE Analysts

Clips
News
More from Cloud AI Journey With HPE

Savannah Peterson

Principal Analyst & Host

SiliconANGLE Media, Inc.

Rob Strechay

Dir./Principal Analyst & Host

theCUBE Research

Metadata matters: Inside HPE’s strategy for streamlined storage and accelerated AI adoption

Unified data platforms are reshaping the foundation of enterprise storage as AI adoption accelerates, forcing businesses to overhaul aging architectures and rethink how data is managed, scaled and consumed.Faced with mounting data sprawl and the demands of GPU-accelerated workloads, organizations are turning to storage solutions that unify block, file and object workloads under one streamlined framework. The shift is urgent: Legacy systems can’t keep pace with AI-driven environments that require flexibility, centralized control and the ability to seamlessly handle unstructured datasets across hybrid cloud deployments

play_circle_outline HPE's Alletra MP product line for scalable data platforms driving AI transformation.

play_circle_outline Importance of unified data services and cloud consumption model in storage.

play_circle_outline Impact of AI on storage architecture, virtualization, and hybrid deployments.

play_circle_outline HPE's focus on providing rich technology trajectory with GreenLake Cloud platform user experience.

Info

Episode 4: Insights From theCUBE Analysts

Savannah Peterson

Principal Analyst & Host SiliconANGLE Media, Inc.

HOST

Rob Strechay

Dir./Principal Analyst & Host theCUBE Research

HOST

search

>> Hello storage fans and welcome back to our exclusive series with HPE on the cloud AI journey. My name's Savannah Peterson, delighted to be joined by analyst Rob Strechay today to break down the footage from our Patrick Osborne interview yesterday. Rob, really excited to talk about this one with you.

>> Yeah, I mean, this is one of my favorite topics, talking about data platforms and how they're going to inform and really drive and feed AI going forward.

>> HPE is in a really unique position right now. They've been building out their offering. The platform is poised to continue to be a leader in this space. Patrick and I had the chance to talk about their unified data services a lot yesterday. Let's go ahead and run that clip back.

>> We started this journey about 18 to 24 months ago with our Alletra MP product line, and we saw on the market the need for customers, they need to scale both for performance and for capacity. And we brought out a very unique and differentiated, what we call a disaggregated storage architecture, all based on some of the modern basic building blocks in data storage, all MVME, you can scale. It's an end by end architecture, and it really allows customers to start small, scale as they grow, do hardware lifecycle management. And I think one of the key things is that it's a unified data services architecture in that it can serve multiple workloads. So we do block, we do file, and then now we have capabilities around object store with our key value store. And it's a great system. It's a great opportunity for customers to talk about data transformation. It's a hot topic right now in the world of AI, because you don't have AI without data. And I think two of the big things that differentiate it for us is we're changing the storage ownership model, so we're bringing the cloud consumption model. Everyone is pretty familiar with HPE for GreenLake. They associate that with a flexible consumption model, and we're bringing the cloud operational model with our GreenLake cloud platform. So very scaled, and a lot of customers are voting with their dollars for Alletra MP, which has been very exciting.

>> Wow. Pretty interesting. What do you think, Rob?

>> I think, again, you saw it in, actually Antonio talked about it in some of the financial statements that they had in this last quarter, that it's really, the growth has been phenomenal with the Alletra MP and having come out 18 to 24 months ago. And really what they're doing with it, I think the unified data services bringing together a full lifecycle management of the hardware and making that easy, as well as giving that cloud consumption model under GreenLake, that HPE has really become known to. This is why people like cloud and really enjoy buying in that methodology. And I think customer adoption has really been, like Patrick said, they've been voting with their dollars, which has been great to see.

>> I mean, who doesn't want a community and customers who are that excited that they're voting with their dollars? You highlighted the three things that stood out to me in that particular part of the interview as well. It's a unique value proposition that's really mindful of the experience of those developers and of their customers. One of the other things that we got to break down, and Patrick was the perfect person for this, was all of the trends in the storage landscape. Let's take a look at that footage.

>> So a couple of things, number one is the opportunity around AI is enormous right now. That doesn't have to be oversold here, but that puts a huge tax on customers' ability to do data preparation, data classification. They're re-architecting their storage ecosystem to be able to deal with all of the demands around scale and performance for GPUs and GPU accelerated workloads. Oftentimes, there are different storage types. So AI is putting a big focus on unstructured data, so all different types of data, different protocols. Many of the applications that you use in the AI data pipeline were born in the cloud. So they use different paradigms for accessing data, different ways to organize that data. And so it's a great opportunity for customers to step back and take a look at, do I have the right architecture in place to service those needs? Some of those applications are going to get deployed on-prem because of data privacy and regulation. Some of them are going to be hybrid in nature, so we have a set of data services that unlocks that. So, AI is a big area. And then obviously for what's going on in the market right now, people are taking a look at what they're doing for virtualization. And that's a big re-architecture of their infrastructure to take a look at, you know, am I going to continue to do things the same way? Am I going to start deploying containerized applications? So there's lots of questions for customers, all opportunities for HPE to help.

>> All right, Rob, I know you're a bit of a storage expert. Do you agree with Patrick's hot take?

>> I do. I mean, the AI's impact on storage has really put it back in focus and really focusing on the fact that organizations have to figure out, do they need to re-architect what they've already deployed? Do they actually upgrade and move to a new way of looking at this? Because GPUs and how to feed those GPUs in a very, I guess you could say, not methodical, but just force-feed those GPUs to keep them busy because it's like a factory. I think also a lot of it was about being hybrid, and I think the hybrid and on-prem deployments we see in the data that we see, that I guess we all see is the fact that the data, 85% of data being used for AI is actually on-premise these days. And so when you start to think about it, we've talked about it and had it in breaking analysis with the fact that Jamie Dimon is one of the largest competitors to OpenAI, and that's because he's got zettabytes of data and that is intellectual property. So when you start to look at this, I think that really, organizations are looking at this, and you can't get away from the fact that the virtualization shift that is happening and people trying to understand what do they do with their infrastructure. I think there had been some storage silos that had been in virtualized deployments that are actually being re-centralized. I think everything that's old is new again, in that, because we went back and forth between virtual storage and actual metal storage, and I think that's happening again as well. And I thought it was great. I thought it was dead-on.

>> I think you're absolutely right. And I think one of the other things that's cool to back that up, and I've been asking a lot of our community about this, is this new rush to AI adoption and the transformation that we're amidst, is that a vertical-specific thing? Are there types of teams or applications in which we're seeing this accelerate faster or not? Let's see what Patrick had to say about the customer landscape.

>> One of the key things I see is that customers are... They no longer want to buy on horizontal lines. I've got a compute vendor, a server vendor of choice and a second source. I've got a networking person, I've got a storage vendor I work with quite a bit. They want outcomes. They want to drive towards a very compelling TCO and obviously an ROI for them that they can show back to their business and they want workload-oriented systems. So we have this great and awesome storage and persistence layer for data, but we also build that into all of our private cloud and hybrid cloud offerings. So a customer comes to us and says, "I have a workload XYZ." Maybe it's a mode one workload virtualization or database. "I've got all these new containerized applications. HPE, help me figure out, help me architect a workload-oriented system that's going to give me SLAs, it's going to provide the best TCO, and in some cases provide me a managed service to do that, because I don't want my people doing that anymore."

>> Got to say, I'm not surprised to hear him say that everyone is looking for ROI as soon as possible. Rob, what's your thoughts on that?

>> I think also talking about the fact that people really aren't trying to buy separate compute and networking from different vendors and piece part it together. They really want solutions, not just a whole bunch of Tinker Toys that they're... Throwback to when I was a kid, I guess, I was going to say.

>> I was going to say, everything old is new again, Rob.

>> I'm just using the magic wand to come back to that. But back in the day, when you start to look at these things piecing together, people don't want to do that. In fact, most organizations aren't even set up to be able to do that anymore where they just can't hire. I was saying this that I've seen one actual ad for a network admin in the last three years, and I actually reposted it on LinkedIn because I'm like, "Wow, somebody's actually hiring for a specific network admin." To me, that's been an anomaly, where you have people who are wearing all different hats, and I think also they're looking to simplify IT because their SLAs have become so tight. And I think Patrick hit on that, and it really is about offloading IT and platform engineering that we talk about a lot. I know you and I, that's something we go deep into with that. So I think a lot of this and how TCO and ROI play together is key to what you brought up in the beginning there.

>> Absolutely. Speaking of things we're bringing up, you mentioned a magic wand. I had the joy of asking Patrick what he would do if he could wave a magic wand right now to get everybody on the same page. Let's roll it back.

>> This whole market around AI and how it's being applied, I know we just made some announcements on workload-oriented systems in concert with Nvidia, like a big theme around that agentic AI workflows. People are oftentimes taking brownfield applications and adding AI services to it, but it's just the beginning. Incrementalism is not going to work in this case. They need a new way of thinking. They need to set an order of magnitude, a different mindset on performance, on scale, on data services that come along with their data. In terms of AI alone, things like scalable metadata services are becoming super important. The data about the data is almost as important as the data itself because of the volume of it, the needs from an application standpoint. So I'd say starting now and having a really good enterprise view of your data, real estate, it's all over the place. It's in the core data center, it's in colocation facilities. It's at the edge now, which is super important. So there's no better time to start than now.

>> I really appreciate Patrick humoring me on this one and playing along. One quote I want to call out, Rob, just specifically get your hot take on. He said something that's really important that I think sometimes people miss when they're starting to think about their strategy around this. He said that the data about the data is almost as important or more important than the data itself. What do you think about that?

>> Dead on. I think Patrick really hit the nail on the head with that. Metadata and the business and technical metadata, because there are differences, and we talk about it a lot here. Really, that's what he's talking about is the two different types of metadata that people are utilizing for building out these AI systems. And I think it's been critical about where HPE has seen... Their skating to the puck, and you didn't have to use his magic wand to get to that puck, but again, when you start to look at it, he was really on top of it. And enterprise data is essential. It's the core of the enterprise, it's the core of organizations, and it's their intellectual property. So I'm glad you got him to play along with you and wave his magic wand there.

>> Well, I think we talk about AI being magic, but there's a lot of work and planning and development that have gone into these platforms and these solutions to make that magic happen.

>> I think that is a key in the fact that you just hit it, is that making something simple that is as complex as these storage and data platforms is not easy in the least. Been there, done that, and I think Patrick does a good job of not getting into the weeds of how difficult it is and all the parts and pieces, because it is about making it a simple customer experience.

>> Yes, it is. And that's one of the things that Patrick actually mentions in the final clip that we get to review today. When he was giving a recap of their announcements earlier this year at Discover in Barcelona at GTC, but what that means for the user experience of their customers. Let's check it out.

>> In the past, one of the big storage paradigms in the cloud was using object storage, and it was often relegated to a very low-performant, low dollar-per-terabyte tier, archiving and what have you. But now this is like a first-class protocol and people expect it to be always on very performant, have all the data services and capabilities that you would have for any other storage architecture. So we launched our Alletra MP X10000, and at GTC we announced a number of capabilities in this area around object, around the ability to do rag and inferencing directly on the platform, so indexing and a number of capabilities specific to the AI pipeline. We've made a number of announcements for our block and file portfolio as well too, so a unified block and file for customers who need those two protocols for their various virtualization and mode one workloads. So we're making quite a bit. Then one of the cool things about our platform is that the platforms themselves have a very rich technology trajectory, but the user experience is all on the GreenLake Cloud platform. And so people don't... Oftentimes, they don't understand the scale of what we've deployed. We are literally managing millions of devices across the world, almost nine exabytes of storage globally.

>> So much going on, so much to unpack. One thing that really stood out to me there is the ability to do rag and inferencing directly in the platform. He talks a lot about how things are embedded in under that same UX, that GreenLake UX experience for all the folks that are interacting with their platform. I thought that was pretty compelling.

>> I think that, again, that to me was a... You do the work closer to the data. I mean, that to me makes so much sense. Why move it all around? And I think we're seeing a lot of people coming to that conclusion, especially with rag, where you're looking at it, the retrieval, augmented generation capabilities of AI. When you have to say you're building customer support apps and you have all of these unstructured data, and I think he talked about object being a first-class citizen and looking at how you bring it all together. People don't want to go different places for different types of protocols. They want it to all be in the same place so that they can utilize all that data as part of their rag and inference. And inference is going to have a way longer tail, as we all know.

>> Inference is what's going to make that AI real, make that magic happen that we were talking about earlier. And having a partner like HPE and a platform with such a broad offering of solutions baked into that smooth experience is so mission-critical when we're talking about transformation at the level of cost and velocity, quite frankly, that we're at. Rob, I've got to say, I'm really enjoying doing this series with you. It's been absolutely fantastic.

>> Yeah, I mean, it's like it almost writes itself. I mean, the guests have been great, and I think the roadmap and the direction that HPE is going with this in the data platform space is really dead on.

>> Yeah, absolutely. If you haven't had a chance, definitely check out our segment with OpsRamp and our analysis thereof. We've also got Brad Parks coming up. This has been a fantastic conversation about storage and everything that Patrick Osborne shared with us. Rob, I look forward to the next. Thank all of you for tuning in to our exclusive series with HPE, the Cloud AI journey. My name's Savannah Peterson, here in our Palo Alto Studios, you're watching The Cube, the leading source for enterprise tech news.

Episode 4: Insights From theCUBE Analysts

search

>> Hello storage fans and welcome back to our exclusive series with HPE on the cloud AI journey. My name's Savannah Peterson, delighted to be joined by analyst Rob Strechay today to break down the footage from our Patrick Osborne interview yesterday. Rob, really excited to talk about this one with you.

>> Yeah, I mean, this is one of my favorite topics, talking about data platforms and how they're going to inform and really drive and feed AI going forward.

>> HPE is in a really unique position right now. They've been building out their offering. The platform is poised to continue to be a leader in this space. Patrick and I had the chance to talk about their unified data services a lot yesterday. Let's go ahead and run that clip back.

>> We started this journey about 18 to 24 months ago with our Alletra MP product line, and we saw on the market the need for customers, they need to scale both for performance and for capacity. And we brought out a very unique and differentiated, what we call a disaggregated storage architecture, all based on some of the modern basic building blocks in data storage, all MVME, you can scale. It's an end by end architecture, and it really allows customers to start small, scale as they grow, do hardware lifecycle management. And I think one of the key things is that it's a unified data services architecture in that it can serve multiple workloads. So we do block, we do file, and then now we have capabilities around object store with our key value store. And it's a great system. It's a great opportunity for customers to talk about data transformation. It's a hot topic right now in the world of AI, because you don't have AI without data. And I think two of the big things that differentiate it for us is we're changing the storage ownership model, so we're bringing the cloud consumption model. Everyone is pretty familiar with HPE for GreenLake. They associate that with a flexible consumption model, and we're bringing the cloud operational model with our GreenLake cloud platform. So very scaled, and a lot of customers are voting with their dollars for Alletra MP, which has been very exciting.

>> Wow. Pretty interesting. What do you think, Rob?

>> I think, again, you saw it in, actually Antonio talked about it in some of the financial statements that they had in this last quarter, that it's really, the growth has been phenomenal with the Alletra MP and having come out 18 to 24 months ago. And really what they're doing with it, I think the unified data services bringing together a full lifecycle management of the hardware and making that easy, as well as giving that cloud consumption model under GreenLake, that HPE has really become known to. This is why people like cloud and really enjoy buying in that methodology. And I think customer adoption has really been, like Patrick said, they've been voting with their dollars, which has been great to see.

>> I mean, who doesn't want a community and customers who are that excited that they're voting with their dollars? You highlighted the three things that stood out to me in that particular part of the interview as well. It's a unique value proposition that's really mindful of the experience of those developers and of their customers. One of the other things that we got to break down, and Patrick was the perfect person for this, was all of the trends in the storage landscape. Let's take a look at that footage.

>> So a couple of things, number one is the opportunity around AI is enormous right now. That doesn't have to be oversold here, but that puts a huge tax on customers' ability to do data preparation, data classification. They're re-architecting their storage ecosystem to be able to deal with all of the demands around scale and performance for GPUs and GPU accelerated workloads. Oftentimes, there are different storage types. So AI is putting a big focus on unstructured data, so all different types of data, different protocols. Many of the applications that you use in the AI data pipeline were born in the cloud. So they use different paradigms for accessing data, different ways to organize that data. And so it's a great opportunity for customers to step back and take a look at, do I have the right architecture in place to service those needs? Some of those applications are going to get deployed on-prem because of data privacy and regulation. Some of them are going to be hybrid in nature, so we have a set of data services that unlocks that. So, AI is a big area. And then obviously for what's going on in the market right now, people are taking a look at what they're doing for virtualization. And that's a big re-architecture of their infrastructure to take a look at, you know, am I going to continue to do things the same way? Am I going to start deploying containerized applications? So there's lots of questions for customers, all opportunities for HPE to help.

>> All right, Rob, I know you're a bit of a storage expert. Do you agree with Patrick's hot take?

>> I do. I mean, the AI's impact on storage has really put it back in focus and really focusing on the fact that organizations have to figure out, do they need to re-architect what they've already deployed? Do they actually upgrade and move to a new way of looking at this? Because GPUs and how to feed those GPUs in a very, I guess you could say, not methodical, but just force-feed those GPUs to keep them busy because it's like a factory. I think also a lot of it was about being hybrid, and I think the hybrid and on-prem deployments we see in the data that we see, that I guess we all see is the fact that the data, 85% of data being used for AI is actually on-premise these days. And so when you start to think about it, we've talked about it and had it in breaking analysis with the fact that Jamie Dimon is one of the largest competitors to OpenAI, and that's because he's got zettabytes of data and that is intellectual property. So when you start to look at this, I think that really, organizations are looking at this, and you can't get away from the fact that the virtualization shift that is happening and people trying to understand what do they do with their infrastructure. I think there had been some storage silos that had been in virtualized deployments that are actually being re-centralized. I think everything that's old is new again, in that, because we went back and forth between virtual storage and actual metal storage, and I think that's happening again as well. And I thought it was great. I thought it was dead-on.

>> I think you're absolutely right. And I think one of the other things that's cool to back that up, and I've been asking a lot of our community about this, is this new rush to AI adoption and the transformation that we're amidst, is that a vertical-specific thing? Are there types of teams or applications in which we're seeing this accelerate faster or not? Let's see what Patrick had to say about the customer landscape.

>> One of the key things I see is that customers are... They no longer want to buy on horizontal lines. I've got a compute vendor, a server vendor of choice and a second source. I've got a networking person, I've got a storage vendor I work with quite a bit. They want outcomes. They want to drive towards a very compelling TCO and obviously an ROI for them that they can show back to their business and they want workload-oriented systems. So we have this great and awesome storage and persistence layer for data, but we also build that into all of our private cloud and hybrid cloud offerings. So a customer comes to us and says, "I have a workload XYZ." Maybe it's a mode one workload virtualization or database. "I've got all these new containerized applications. HPE, help me figure out, help me architect a workload-oriented system that's going to give me SLAs, it's going to provide the best TCO, and in some cases provide me a managed service to do that, because I don't want my people doing that anymore."

>> Got to say, I'm not surprised to hear him say that everyone is looking for ROI as soon as possible. Rob, what's your thoughts on that?

>> I think also talking about the fact that people really aren't trying to buy separate compute and networking from different vendors and piece part it together. They really want solutions, not just a whole bunch of Tinker Toys that they're... Throwback to when I was a kid, I guess, I was going to say.

>> I was going to say, everything old is new again, Rob.

>> I'm just using the magic wand to come back to that. But back in the day, when you start to look at these things piecing together, people don't want to do that. In fact, most organizations aren't even set up to be able to do that anymore where they just can't hire. I was saying this that I've seen one actual ad for a network admin in the last three years, and I actually reposted it on LinkedIn because I'm like, "Wow, somebody's actually hiring for a specific network admin." To me, that's been an anomaly, where you have people who are wearing all different hats, and I think also they're looking to simplify IT because their SLAs have become so tight. And I think Patrick hit on that, and it really is about offloading IT and platform engineering that we talk about a lot. I know you and I, that's something we go deep into with that. So I think a lot of this and how TCO and ROI play together is key to what you brought up in the beginning there.

>> Absolutely. Speaking of things we're bringing up, you mentioned a magic wand. I had the joy of asking Patrick what he would do if he could wave a magic wand right now to get everybody on the same page. Let's roll it back.

>> This whole market around AI and how it's being applied, I know we just made some announcements on workload-oriented systems in concert with Nvidia, like a big theme around that agentic AI workflows. People are oftentimes taking brownfield applications and adding AI services to it, but it's just the beginning. Incrementalism is not going to work in this case. They need a new way of thinking. They need to set an order of magnitude, a different mindset on performance, on scale, on data services that come along with their data. In terms of AI alone, things like scalable metadata services are becoming super important. The data about the data is almost as important as the data itself because of the volume of it, the needs from an application standpoint. So I'd say starting now and having a really good enterprise view of your data, real estate, it's all over the place. It's in the core data center, it's in colocation facilities. It's at the edge now, which is super important. So there's no better time to start than now.

>> I really appreciate Patrick humoring me on this one and playing along. One quote I want to call out, Rob, just specifically get your hot take on. He said something that's really important that I think sometimes people miss when they're starting to think about their strategy around this. He said that the data about the data is almost as important or more important than the data itself. What do you think about that?

>> Dead on. I think Patrick really hit the nail on the head with that. Metadata and the business and technical metadata, because there are differences, and we talk about it a lot here. Really, that's what he's talking about is the two different types of metadata that people are utilizing for building out these AI systems. And I think it's been critical about where HPE has seen... Their skating to the puck, and you didn't have to use his magic wand to get to that puck, but again, when you start to look at it, he was really on top of it. And enterprise data is essential. It's the core of the enterprise, it's the core of organizations, and it's their intellectual property. So I'm glad you got him to play along with you and wave his magic wand there.

>> Well, I think we talk about AI being magic, but there's a lot of work and planning and development that have gone into these platforms and these solutions to make that magic happen.

>> I think that is a key in the fact that you just hit it, is that making something simple that is as complex as these storage and data platforms is not easy in the least. Been there, done that, and I think Patrick does a good job of not getting into the weeds of how difficult it is and all the parts and pieces, because it is about making it a simple customer experience.

>> Yes, it is. And that's one of the things that Patrick actually mentions in the final clip that we get to review today. When he was giving a recap of their announcements earlier this year at Discover in Barcelona at GTC, but what that means for the user experience of their customers. Let's check it out.

>> In the past, one of the big storage paradigms in the cloud was using object storage, and it was often relegated to a very low-performant, low dollar-per-terabyte tier, archiving and what have you. But now this is like a first-class protocol and people expect it to be always on very performant, have all the data services and capabilities that you would have for any other storage architecture. So we launched our Alletra MP X10000, and at GTC we announced a number of capabilities in this area around object, around the ability to do rag and inferencing directly on the platform, so indexing and a number of capabilities specific to the AI pipeline. We've made a number of announcements for our block and file portfolio as well too, so a unified block and file for customers who need those two protocols for their various virtualization and mode one workloads. So we're making quite a bit. Then one of the cool things about our platform is that the platforms themselves have a very rich technology trajectory, but the user experience is all on the GreenLake Cloud platform. And so people don't... Oftentimes, they don't understand the scale of what we've deployed. We are literally managing millions of devices across the world, almost nine exabytes of storage globally.

>> So much going on, so much to unpack. One thing that really stood out to me there is the ability to do rag and inferencing directly in the platform. He talks a lot about how things are embedded in under that same UX, that GreenLake UX experience for all the folks that are interacting with their platform. I thought that was pretty compelling.

>> I think that, again, that to me was a... You do the work closer to the data. I mean, that to me makes so much sense. Why move it all around? And I think we're seeing a lot of people coming to that conclusion, especially with rag, where you're looking at it, the retrieval, augmented generation capabilities of AI. When you have to say you're building customer support apps and you have all of these unstructured data, and I think he talked about object being a first-class citizen and looking at how you bring it all together. People don't want to go different places for different types of protocols. They want it to all be in the same place so that they can utilize all that data as part of their rag and inference. And inference is going to have a way longer tail, as we all know.

>> Inference is what's going to make that AI real, make that magic happen that we were talking about earlier. And having a partner like HPE and a platform with such a broad offering of solutions baked into that smooth experience is so mission-critical when we're talking about transformation at the level of cost and velocity, quite frankly, that we're at. Rob, I've got to say, I'm really enjoying doing this series with you. It's been absolutely fantastic.

>> Yeah, I mean, it's like it almost writes itself. I mean, the guests have been great, and I think the roadmap and the direction that HPE is going with this in the data platform space is really dead on.

>> Yeah, absolutely. If you haven't had a chance, definitely check out our segment with OpsRamp and our analysis thereof. We've also got Brad Parks coming up. This has been a fantastic conversation about storage and everything that Patrick Osborne shared with us. Rob, I look forward to the next. Thank all of you for tuning in to our exclusive series with HPE, the Cloud AI journey. My name's Savannah Peterson, here in our Palo Alto Studios, you're watching The Cube, the leading source for enterprise tech news.