 Welcome back to HPE Discover 2021, the virtual edition. My name is Dave Vellante, and you're watching theCUBE's continuous coverage of HPE's big customer event. Patrick Morehead is here, more insights and strategies. The number one analyst in the research analyst business, Patrick, always a pleasure, great to see you. Dave, great to see you too. And I know you're up there fighting for that number one spot too. It's great to see you and it's great to see you in the meetings that we're in, but it's even more fun to be here on theCUBE and I love to be on theCUBE. And every once in a while, you'll even call me a friend of theCUBE. Oh, unquestionably, my friend. And so, and I can't wait, second half. I mean, you're traveling right now. We're headed to Barcelona to Mobile World Congress later on this month. So, we're going to see each other face to face this year, 100%. So, looking forward to that. So, you know, let's get into it. You know, before we get into HPE, let's talk a little bit about what you're seeing in the market. We've got, you know, we've finally, it feels like the on-prem guys are finally getting their cloud act together. It's maybe taken a while, but we're seeing as the service models emerge. I think it's resonating with customers. Clearly not everything's moving to the cloud. There's this hybrid model emerging. Multi-cloud is real, despite what some cloud players want to say. And then there's this edge is like jump ball. What are you seeing in the marketplace? Yeah, Dave, it's as exciting as ever in, just to put in perspective, I mean, the public cloud has been around for about 10 years and still only 20%, around 20% of the data and 20% of the applications are there. Now, albeit very important ones, and I'm certainly not a public cloud denier, I never have been, but there are some missing pieces that need to come together. And, you know, even five years ago, we were debating Dave, the hybrid cloud. And I feel like when Amazon brought out Outposts, the conversation was over, right? And now what you have is cloud native folks building out hybrid and on-prem capabilities, you know, the classic on-prem folks building out a hybrid and as a service capabilities. And I really think it boils down to two things. I mean, it's wanting to have more flexibility. And, you know, I hate to use it because it sounds like a marketing word, but agility, the ability to spin up things and spin down things in a very quick way. And, you know, what they've learned, the veterans also know, hey, let's do this in a way that doesn't lock us in too much into a certain vendor. And, you know, I've been around for a long time, Dave, and I'm a realist too. Well, you have to lock yourself into something. It just depends on what you want to lock yourself into, but super exciting. And what HPE, you know, when they threw their ax in the sea with GreenLake, I think it was four years ago, I think really started to stir the pot. You know, you mentioned the term cloud denier, but you know, and I feel like the narrative from, I like to, the term used, I think you should use the term veteran, you know. This thing, ours is the only industry, Patrick, where legacy is a pejorative. But so, but the point I want to make is I feel like there's been a lot of sort of fear from the veteran players, but I've looked at it differently. I wonder what your take is. I think I calculated that the CapEx spending by the big four public clouds, I'm including Alibaba last year, was $100 billion. That's like a gift to the world. Here we're going to spend $100 billion like the internet. Here you go, build. And so, and I feel like companies like HPE are finally saying, yeah, we're going to build, we're going to build a layer and we're going to hide the complexity and we're going to add value on top. You know, what do you think about that? Yeah, so I think it's, now I wish, I wish the on-prem folks like HPE you would have done it 10 years ago, but I don't think anybody expected the cloud to be as big as it's become over the last 10 years. I think we saw companies like Salesforce with SaaS taking off, but I think it is the right direction because there are advantages to having workloads on-prem. And if you add an as-a-service capability on top of that, and let's say you even do a colo or a managed service, it's pretty close to being similar to the public cloud with the exception that you can't necessarily swipe a credit card for a bespoke workload if you're a developer. And it is a little harder to scale out, but that is the next step in the equation, Dave, which is having these folks make capital expenditures make them in a colo facility and then put a layer to swipe a credit card and you literally have the public cloud. Yeah, so that's a great point and that's where it's headed, isn't it? So let's talk about the horses on the track. HPE, as you mentioned, I didn't realize it was four years ago. I thought it was, wow, that's amazing. So when everybody's followed suit, you see Dell has announced, Cisco has announced, Lenovo has announced, I think IBM as well. So everybody's sort of following suit there. The reality is, is it's taken some time to get this stuff standardized. What are you seeing from HPE? They've made some additional announcements at Discover. What's your take on all this? Yeah, so HPE was definitely in the rabbit here and they were first in the market. It was good to see first off, some of their announcements on how it's going. They talked about $4.8 billion in TCV that's been booked, 1200 customers, over 900 partners, and 95% retention. And I think that's important. Anybody who's in the lead, and I remember what AWS used to do with the slide with the amount of customers that would just get bigger and bigger and bigger. And that's a good way to show momentum. I like the retention part two, which is 95%. And I think that says a lot. Probably the more important announcements that they made is they talked about the GA of some of their solutions on GreenLake and whether it was SAP, HANA, ML Ops, HPC with Ansys, and VDI with Citrix and NVIDIA. But they also brought more of what I would call a vertical layer. And I'm sure you've seen the verticalization of all of these cloud and as a service workloads, but what they're doing with Epic with EMR and USIS with financial payments and Splunk and Intel with data and risk analysis. And finally, a full stack for Telco 5G, one of the biggest secrets, and I covered this about five years ago, is HPE actually has a full stack that Western European carriers use, and they're now extending that to 5G. So more horizontal and more vertical. That was one of the big swipes that I saw. There was a second though, but maybe we can talk about these. Yeah, okay, and so the other piece of that, of course, is standardization, right? There was a lot of customization leading up to this, and everybody's sort of, everybody always had some kind of financial game they can play and say, hey, there's as a service model, but this is definitely more of a standardized, scalable move that HPE is making with what they call lighthouse, right? Yeah, that's exactly right. And I've talked to some Green Lake customers and they obviously gave it kudos or they wouldn't have, HPE wouldn't have served them up and they wouldn't have been buying it, but they did say it took a while, took some paperwork to get it going. It's not 100% a push button, that's partially because HPE allows you to customize the hardware. You want a one-off network adapter, HPE says yes, right? You want to integrate a different type of storage, they said yes, but with Green Lake lighthouse, it's more of a, what you see is what you get, which by the way is very much like the public cloud, right? You go to a public cloud product sheet or order sheet, you're picking from a list and you really don't know everything that's underneath the covers aside from, let's say the speed of the network, the type of the storage and the amount of the storage. You do get to pick between let's say an Intel processor, Graviton 2 or an AMD processor. You get to pick your own GPU, but that's pretty much it. And HPE lighthouse, sorry, Green Lake lighthouse is bringing, I think a simplification to Green Lake that it needs to truly scale beyond, let's say the lighthouse customers that HPE has. Well done. So, and I hear your point about, we're 10 years in, we're plus, and to me, this is like a mandate. I mean, it's okay, good job guys, about time, but if I had to sort of look at the big player, like can we have an oligopoly here in this business? It's HPE Cisco, you got Dell, you got Lenovo, you got IBM, they're all doing this. And they all have a different, little difference ways to skin a cat. And to your point about simplicity, it seems like HPE is all in, and Tony's like, okay, here's where we're going. He announced that a while ago. So, and they seem to have done a good job with Wall Street and they got a simple model. Dell's obviously bigger portfolio, much more complicated. IBM is even more complicated than that. I don't know so much about Lenovo. And then Cisco, of course, has acquired a ton of SaaS companies. And it's sort of, they've got a lot of bespoke products that they're trying to put together. So they've got, but they do have SaaS models. So each of them is coming at it from a different perspective. How do you think, and so, and to the other point, we got Lighthouse, which is sort of phase one, get product market fit. Phase two now is scale, codify, standardize, and then phase three is the moat. You know, build your unique advantage that protects your business. What do you see as HPE's sort of unique value proposition and moat that they can build longer term? That's a great, great question. And let me rattle off kind of what I'm seeing at some of these players here. So Cisco, ironically, has sales the most software of any of those players that you mentioned with the exception of IBM. And they have more. Yeah, CICS, DB2, right by the way. Yeah, and they're the number two security player to Microsoft's number one. So, and I think their valuation on the street indicate, you know, shows that. I feel like, you know, Dell Tech is a very broad play, right? Because not only do they have server storage, networking, and security, but they also have PCs and devices. So it's a scale end-to-end play with a focus on VMware solutions, not exclusively, of course. And then you've got Lenovo, who is just getting into the as-a-service game and are, gosh, they're doing great in hyperscale. They've got scale, they're vertically integrated. I don't know if too many people talk about that, but Lenovo does a lot of their own manufacturing and they actually manufacture NetApp storage solutions as well. So yeah, each of these folks brings a different game to the table. I think with HPE, what you're bringing to the table is nimble. When HPE and HPE split, the number one thing that I said was that HPE is gonna have to be so much faster that it offsets the scale that Dell technology has. And the HPE's credit, although I don't think we're getting credit for this in the stock market yet, and I know you and I are both industry folks, not financial folks, but I think their biggest thing is speed and the ability to move faster. And that is what I've seen. As it relates to the moat, which is a unique competitive advantage, quite frankly, I'm still looking for that Dave in what that is. And I think in this industry, it's nearly impossible. And I would posit that any, even the cloud folks, if you say, is there something that AWS can do that Azure can't, if it puts its mind to it or GCP, I don't think so. I think it's more of a kind of land and expand. And I think for HPE, when it comes to high performance computing, and I'm not just talking about government installations, I'm talking about product development, drug development. I think that is a landing place where HPE already does pretty well, can come in and expand its footprint. You know, that's really interesting observations. So, and I would agree with you, it's kind of like a, this is a copycat industry. It's like the West Coast offense, like the NFL. Oh, yeah. And so the moat comes from, you know, brand execution. And your other point about when HPE and HPE split, that was a game changer, because all of a sudden you saw companies like Veeam, it was had a long-term relationship with HPE or HPE, but then they came out of the woodworks and started to explode. And so it really opened up opportunities. So it really is around execution, isn't it? But go ahead, please. Yeah, Dave, if I had to pick something that I think HPE needs to always be ahead in as a service. And listen, I both know, announcements don't mean delivery, but there is correlation between if you start four years ahead of somebody, that other company is gonna have to put just, I mean, they're gonna have to turn that ship and many of its competitors are really big ships to be able to get there. So I think what Antonia needs to do is run like hell. Right, because it, I think it is in the lead and as a service holistically, doesn't mean they're gonna be there forever, but they have to stay ahead. They have to add more horizontal solutions. They have to add more vertical solutions. And I believe that at some point, it does need to invest in some CapEx at somebody like an Equinex, put a credit card swiper on top of that. And Dave, you have the public cloud. You don't have all the availability zones, but you have a public cloud. Yeah, that's going to happen. I think you're right on. So we see this notion of cloud expanding. It's no longer just a remote set of services somewhere out in the cloud. Like you said, outpost was the sort of signal, okay, we're coming on-prem. Clearly the on-prem guys are connecting to the cloud. Multi-cloud exists, we know this. And then there's the edge. But that brings me to that sort of vision. Everybody's laying out of this seamless integration, hiding the complexity, log into my cloud, and then life will be good. But the edge is different, right? It's not just a retail store or a race track. I mean, there's the far edge. There's the Tesla car. There's going to be compute everywhere. And that sort of ties into the data, the data flows, the real-time inferencing at the edge, AI, new semiconductor models. You came out of the semiconductor industry, you know what inside and out. Arm is exploding, it's dominating in the edge with Apple and Amazon Alexa and things like that. That's really where the action is. So this is a really interesting cocktail and soup that we have going on. How do you see it? Dave, if the data, most data, I think one thing most everybody agrees on is that most of the data will be created on the edge, whether that's a moving edge, a car, a smartphone, or what I call an edge data center without tile flooring, like that server that's bolted to the wall of McDonald's when you drive through, you can see it, versus the Walmart, every Walmart has a raised tile floor. It's the edge, too. Economically and performance-wise, it doesn't make any sense to send all that data to the mother ships, and whether that's on-prem data center or the giant public cloud. More efficient way is to do the compute at the closest way possible. But what it does, it does bring up challenges. The first challenge is security. If I wanted to, I could walk in, and I could take that server off the McDonald's or the Shell gas station wall. So I can't do that in a big data center. So security, physical security is a challenge. The second is you don't have the people to go in there and fix stuff that are qualified. If you have a networking problem that goes wrong in McDonald's, there's nobody there that can help you fix that. So this notion of autonomy and management and not keeping hypercritical data sitting out there, and it becomes a security issue, becomes a management issue. But let me talk about the benefits, though. The benefits are lower latency. You want answers more quickly. When that car is driving down the road and it has a 5G V2X communication, cameras can't see around corners, but that car communicating ahead that ran into the stop sign can, through V2X, talk to the car behind it and say, hey, something is going on there. You can't go to the big data center in the sky to make that happen. That is to be in near real time, and that compute has to happen on the edge. So I think this is a tremendous opportunity. And ironically, the classic on-prem guys, they own this space aside from smartphones, of course. But if you look at compute on a light pole, companies like Intel have built complete architectures to do that, putting compute into 5G base stations. Heck, there was an announcement this week of Google Cloud and its gaming solution putting compute in a carrier edge to give lower latency to deliver a better experience. Yeah, so of course, there is no one edge, it's highly fragmented, but I'm interested in your thoughts on kind of whose stack actually can play at the edge. And I've been sort of poking at HPE about this. And the one thing that comes back consistently is Aruba. We can take Aruba not only to the near edge, but to the far edge. And that, do you see that as a competitive advantage? Oh gosh, yes. I mean, I would say the best acquisition that HPE has made in 10 years has been Aruba. I mean, it's fantastic. And they also managed it in the right way. I mean, it was part of HPE, but it was managed a lot more loosely than a company that might get sucked into the board. And I think that paid off tremendously. They're giving Cisco on the edge an absolute run for their money. They're first with new technologies, but it's about the solution. What I love about what Aruba looks at is it's looking at entertainment solutions inside of a stadium, a information solution inside of an airport as opposed to just pushing the technology forward. And then when you integrate compute with Aruba, I think that's where the real magic happens. Most of the data on a per bit basis is actually video data. And a lot of it's for security, for surveillance, and quite frankly, people taking videos off their smartphones and downloading video. I just interviewed the Chief Network Officer of T-Mobile and their number one bit of data is video, video upload and video download. But that's where the magic happens when you put that connectivity and the compute together and you can manage it in an orderly and secure fashion. Well, I have you, we have a ton of time here, but I gotta pick your brain about Intel, the future of Intel. I know you've been following it quite closely. You always have. Intel's fighting a forefront war. They're battling AMD. They're battling, you know, arm slash Nvidia. They're taking on TSMC now and foundry. And I'll add China for the lunar threat there. So what's your prognosis for Intel? I liked Bob, the previous CEO. And I think he was doing a lot of the right things. But I really think that customers and investors and even their ecosystem wanted somebody leading the company with a high degree of technical aptitude. And Pat had a great job at VMware. I mean, he had a great run there. And I think it is a very positive move. I've never seen the energy at Intel, probably in the last 10 years that I've seen today. I actually got a chance to talk with Pat. I visited Pat last month and talked to him about pretty much everything and where he wanted to take the company, the way he looked at technology, what was important, what's not important. But I think first off, in the world of semiconductors, there are no quick fixes, OK? Intel has probably another two years before we see what the results are. And I think 2023 for them is going to be a huge year. But even with all this competition, though, Dave, they still have close to 85% market share in servers and revenue share for client computing around 90%. And they've built out their networking business. They've built out a storage business with Optane. They have the leading ADAS provider with Mobileye. And listen, I was one of Intel's biggest, I was one of Intel's biggest, I was Intel's biggest customer when I was a compact. I was their biggest competitor at AMB. So I'm not, you know, obviously not overly pushing or they've just got to wait and see. They're doing the right things. They have the right strategy. They need to execute. One of the most important things that Intel did is extend their alliance with TSMC. So in 2023, we're going to see Intel compute units, these tiles they integrate into the larger chips called SOCs, be manufactured by TSMC. Not exclusively, but we could see that. So literally, we could have AMD 3-nanometer on TSMC CPU blocks competing with Intel chips with TSMC 3-nanometer CPU blocks. And it's on. With regard to NVIDIA, I mean NVIDIA is one of these companies that just keeps going, charging, charging hard. And I'm actually meeting with Jensen Wong this week and ARM CEO Simon Seger to talk about this opportunity. And that's a company that keeps on moving. Interestingly enough, NVIDIA, if the ARM deal does go through, will be the largest chip licensee, CPU licensee, and have the largest CPU footprint on the planet. So here we have AMD, who's CPU and GPU, and buying an FPGA company called Xilinx. You have Intel, CPUs, GPUs, machine learning accelerators, and FPGAs. And then you've got ARM slash NVIDIA with everything as well. We have three massive ecosystems. They're going to be colliding here. And I think it's going to be great for competition, Dave. Competition is great. When there's not competition in CPUs and GPUs, we know what happens, right? The beat just does not go on, and we start to stagnate. And I do feel like the industry on CPUs started to stagnate when Intel had no competition. So bring it on. This is going to be great for enterprises and then customers too. And then, oh, by the way, you have the custom chip providers. AWS has created no less than 15 custom semiconductors. It started with networking and Nitro and building out an edge that surrounded the general compute. And then it moved to Inferentia for inference. Trainium is about to come out for training. Graviton and Graviton 2 for general purpose CPU. And then you've got Apple. So innovation is huge. And I love to always make fun of the softwares eating the world. I always say, yeah, but it has to run on something. And so I think the combination of semiconductors, software, and cloud is just really a magical combination. Real quick. Handicap the NVIDIA ARM acquisition. What are the odds that they'll be successful? They say it's on track. You got it at 2 to 1, 3 to 1, 10 to 1? I say 75% yes, 25% no. China has been the odd man out for the last three years. They scuttled the Qualcomm NXP deal. You just don't know what China is going to do. I think the EU with some conditions is going to let this fly. I think the US is absolutely going to let this fly. And even though the IP will still stay over in the UK, I think the US wants to see this happen. Japan and Korea, I think, will allow this. China is the odd man out. In a word, the future of HPE is blank. As a service. Patrick Morehead, always a pleasure, my friend. Great to see you. Thanks so much for coming back in theCUBE. Yeah, thanks for having me on. I appreciate that. All right, everybody, stay tuned for more great coverage from HPE Discover 21. This is Dave Vellante for theCUBE, the leader in enterprise tech coverage. We'll be right back.