Dell Technologies World 2024 | Day 1 AnalystANGLE

Clips
News
More from Dell Technologies World 2024

Bob O'Donnell

Founder & Chief Analyst

TECHnalysis Research

Bob Laliberte

Principal Analyst

theCUBE Research

How Dell integrates compute, storage and networking for seamless hybrid AIOps: theCUBE analysis

From the Tensor chips in Pixel smartphones to the recently released Microsoft Copilot+ PCs, there’s a rethinking of artificial intelligence underway that blends local, on-device computation with cloud-based solutions. For the enterprise, the same process would translate as hybrid AI.How are innovators approaching this on-premises wave and equipping companies with the requisite components to realize it?TECHnalysis’ Bob O’Donnell discusses hybrid AI with theCUBE at Dell Technologies World.“Almost all of the focus has been on doing gen AI in the cloud,” said Bob O’Donnell (pictured, second from right), president and chief analyst at TECHnalysis Research LLC

play_circle_outline Exploring Gen AI and On-Premises Implementation for Enhanced Artificial Intelligence Capabilities

play_circle_outline Importance of partnerships and ecosystem in AI implementation

play_circle_outline Dell's AI factory and extending AI capabilities across the enterprise

play_circle_outline Building Trust and Collaboration: Accelerating AI Adoption with Dell, Nvidia, and Partners

Info

Day 1 AnalystANGLE

Bob O'Donnell

Founder & Chief Analyst TECHnalysis Research

Bob Laliberte

Principal Analyst theCUBE Research

search

>> Good evening and welcome to fabulous Las Vegas, Nevada. It is day one of Dell Tech World. I am very excited to be here for three days of live coverage on theCUBE. My name is Savannah Peterson. Delighted to be joined this week by my co-host, Dave Vellante. Dave.

>> How are you doing, babe?

>> Thanks for having me in Dell Tech World. I'm so happy to be here.

>> I think it's our first show that we've ever done exclusively together.

>> I know.

>> We've done some stuff at AWS, but I'm really excited too.

>> Are you ready for it? Are you a little nervous, no?

>> No. This is like my hundredth Dell Tech World.

>> I meant more about working with me, not about them, not about Dell

>> Working with you, oh no, that's like push button, you kidding me.

>> I am super excited about the show. My first laptop was Dell. It means a lot to be here and I'm very excited that we can kick off our keynote analysis with a pair of Bobs. We have Bob Laliberte, principal Analyst at theCUBEResearch. Very delighted to have you join the team within the last two months. Welcome. Thank you for being here with us tonight.

Bob Laliberte

>> Absolutely. Thank you. Great to be here.

>> You get to hang out with me this evening as well. It's going to be a rowdy late night broadcast on theCUBE. I love that as well. And then we also have Bob O'Donnell, president of Tech Analysis. Bob, thank you so much for being here.

Bob O'Donnell

>> Happy to be here. Happy to be here. Always happy to chat with theCUBE folks. It's always fun.

>> Good to have you, man.

>> So your reputation precedes you before you got here. Dave was hyping you up, so I'm super excited. First impressions. It's the first few hours of Dell Tech World. I'm going to turn to you, Bob O'Donnell first, what did you think of the keynote? What are your first thoughts about everything happening here?

Bob O'Donnell

>> Well, look, I think the overall message from the keynote is an interesting one because what we're seeing is, and it was brought up multiple times, everybody's excited about Gen AI, but a lot of people don't know where to start. But the other interesting thing that I think came out of this, if we think about it big picture, is almost all of the focus has been on doing Gen AI in the cloud. I think the story that Dell started to tell is, hey, wait a minute. Maybe we can do some of this sort of stuff on-prem. The basic message, and by the way, Dell's not the only one that says this is, "Why move your data to the AI? Why not move the AI to your data?" And the vast majority of most organizations', data is still behind their firewall. So it just makes logical sense to do that. The question is, okay, how do we do that? Well, number one, you need the compute, you need the network, you need the storage, and you need the software and platforms.

Bob O'Donnell

And then on top of that, you start to build in things like software applications designed to work with workstations to do RAG, which is this hugely important deal as companies start to fine-tune their existing models and even AI PCs where theoretically you're going to eventually move them. So you put all those pieces together and it's like, okay, this is how companies can start to think about this idea of a hybrid AI model where you've got... Yes. I mean, I'm not saying it's going to all move to the... Obviously not all going to move on-prem, but it doesn't all have to be in the cloud either. You can just the same way we've seen hybrid cloud, there's going to be hybrid AI and the tools are becoming more widely available. The deals that Dell announced with Hugging Face and Meta on leveraging open source tools, which can be run locally is very interesting. So to me that was the big takeaway was that notion of opening up this concept of hybrid AI and making it something that a lot of organizations can consider.

>> I think that's a really great takeaway. Dave, I want to turn to you, since I know you took a, I was just going to call it Twitter, but we'll call it X poll on this specifically.

>> I did.

>> What were the results of your very informed audience?

>> Yeah, so the poll was, what's your take on hybrid AI? Same pace as hybrid cloud. Faster than hybrid cloud. Cloud wins the day. And 8% said same pace as hybrid cloud, very low. Faster than hybrid cloud is 23%. The vast majority, 69% said cloud wins the day. So interesting. I would say this, I think you're right, hybrid AI is going to be like hybrid cloud, but it's going to be different in that the on-prem vendors took over a decade to really get their act together to create the cloud operating model. It's not going to take that long.

Bob Laliberte

>> No.

>> Totally agree with you there.

>> And I would also add that in the 1980s, Michael Dell bet his business on the Wintel Duopoly, and in many ways, Crawford Del Prete brought this up on Twitter. When I did my last poll is Nvidia, are they Cisco or are they Google? And he said the best example is actually Wintel and now you have Nvidia, which is essentially a single monopoly. They've got the hardware and the software, and Michael Dell is betting big on Nvidia and the AI factory.

>> He saw that in the keynote.

>> Basically he's co-opting, or maybe Jensen gave it to him, the AI factory. We saw this at GTC, the Dell-Nvidia AI factory. Now it's the Dell AI factory which extends beyond Nvidia. It includes Qualcomm, AI PCs. Of course they'll give ARM, Intel-

Bob O'Donnell

>> AMD....

>> AMD, it will all be in there and it has the entire Dell portfolio. So they're basically taking over the AI factory. And I think Jensen's happy to let them do that.

>> Well, you could really sense that spirit of collaboration on stage. It felt like an AI celebrity A-lister show out there between ServiceNow and Nvidia, obviously making a cameo with Michael, but a true affection, but also that sense of needing to rely on each other to be both the OS and the hardware stack for our AI enterprise future. You can tell what's going on here and seeing all of them at each other's events over the last few months, we've seen a lot of it. It's interesting. You can tell who's playing nice with each other, that's for sure right now. Bob, I'm curious about your first impressions, what you thought of the keynote or any of your takeaways.

Bob Laliberte

>> Yeah, well, it's interesting that you bring that up because one of my takeaways from this morning, obviously Bob articulated a lot of the technology piece, but the partnership piece and the ecosystem piece, huge. That's one is a very big piece of this. There's no one who can do it all by themselves. So it's really going to be about how organizations, how these vendors can work together to provide an end-to-end solution. So I like what Dell's doing with the AI factory. I think it's a little bit of a misnomer when you think about the factory, you think maybe, hey, this is a big data center thing and what's going on. But in reality, they are talking about, as Bob mentioned, extending those AI capabilities across the entire enterprise through workstations, through laptops, out to the edge at a retail. And like you said, making sure that you're taking that inferencing and that AI to the edge because that's where the real-time analysis needs to be completed. Over the course of the day we've heard multiple times people talking about latency issues, which means distance works.

>> It doesn't matter if it doesn't feel real and it doesn't feel fast.

Bob Laliberte

>> Correct. Well, and especially-

>> It's not real-time.

Bob Laliberte

>> Right. And especially if you're trying to leverage AI to do scene analytics, things like that, quality control, it needs to be done real-time at the edge. So I think what organizations are trying to do... And it is interesting, your poll that you took, because if you think about it from a certain perspective, we're coming at it from the opposite angle. When we did hybrid cloud, everyone had everything on-prem and had to move it to the cloud. Today we're starting with all these AI models are in the cloud, and will be moving down to the data center or to the on-premises locations. So it is going to take time. I mean even the cloud, I think we tracked it for over 17 years of its adoption and going from just, hey, we're putting storage up to doing things, testing the waters. But the key point I think both of us, several of you made already, was the pace of change is just unbelievable. So when we come to these events, it's not uncommon to hear about products going from inception to production in 9 to 12 months, 9 to 12 weeks, months, things like that.

>> Absolutely.

Bob Laliberte

>> So the pace of change keeps accelerating all these new products. We heard about it from even the GPU acceleration and from Apex. Hey, how long do people sign up for Apex? Because concerned about the GPU cycles and how often they come out. So I think organizations really need to look at what their strategy is going to be and what we're hearing more and more are starting to develop their own AI strategy. And I think what Dell's trying to do from what I heard today, is provide a comprehensive system for organizations to onboard AI into their environment. So from the AI factory and the data center and the backend through the inferencing, through the workstations, through the laptops, all the way through the entire space.

>> Well, what hasn't moved into the cloud is a lot of the mission-critical workloads.

>> I was going to bring this up.

>> So the question then becomes... Right. So it's the Oracle stuff, obviously the IBM mainframe, a lot of the hardcore VMware workloads have not moved into the cloud. And so the question then becomes, okay, is that where they're going to apply AI? Now they're very critical, dangerous workloads in a way. You don't want to mess with them, although if you can inject intelligence into them, you can print money. So it's going to be really interesting. I feel like... I think people responded this way in the poll because the cloud folks have massive CapEx. They have the innovation, they have the LLM gardens and the models, and that's where all the action is today. But to your points, both of you, I see what Dell's doing is they're not waiting.

Bob O'Donnell

>> Right.

>> They're going hard after this. They're making investments. They're not only buying GPUs, they're doing a lot of dog fooding internally.

>> They are.

>> Now that said, they have a mandate to do that internally. They're probably better at on-prem than a lot of customers.

>> Hardware supports that. I mean, there's a lot of reasons for them to err on that side, but I think... I'm really glad you brought up the edge point as well as this core infrastructure point because a lot of the hyper-personalization we're going to experience for AIs on highly sensitive data, not the type of data we're going to be keeping in the cloud in a lot of these systems as much as possible for the sake of security and governance. So I do think there is this kind of interesting dance where I kind of disagree with your poll. I think it is going to be a split system. I think it's going to be a very balanced hybrid. We're going to have cloud situation for stuff that is lower weight and less compute versus things that are really impactful and gritty.

Bob O'Donnell

>> In some ways, though, I'm not terribly surprised by the poll, even though I disagree with the results, I'm not terribly surprised.

>> Twitter followers, they have no idea.

Bob O'Donnell

>> It's all good.

>> You should have seen how they blew the security question at RSA.

Bob O'Donnell

>> I think part of the problem honestly is people just don't know. Most people presume you can't even run any of these models locally or if they know that you can, they don't know how to do it, right? Even ChatGPT, theoretically you can license it, but it's not talked about. It's not well understood. So I think that's part of it. But I wanted to follow up on a point Bob made earlier, which is the partnerships thing and tying in what you just said, Savannah as well, the partnership even extends to their own customers. Because we are so early in this process, what Dell is seeing and doing is sharing with these customers, hey, this is what we're doing because all these customers are coming to Dell and saying, "Hey, what are you guys doing by the way? We want to figure it out." So there's very much of a information sharing level of a partnership that's just helping drive this discussion because it's kind of like, hey, we're all in this together. We're all trying to figure it out. Dell's put some policies and principles in place. They're like, we've got to know what we're doing here so that we can explain it to others. They've taken on that responsibility. So they're leveraging that as a means for, I think, gaining more response and more, what's the word I'm thinking of, trust from their customers to say, hey, look, we'll tell you what we did right? We'll tell you what we did wrong and help save you some money along the way. So I think that's part of it, and I view that as another type of partnership. Just to follow up on.

>> Well, you just mentioned there trust was another big cornerstone of the keynote. Jensen said there's no one he trusts more than Michael Dell, which is a real-

>> Oh, Bill McDermott said that.

>> Oh yeah, was Bill, sorry, excuse me.

>> After Jensen said, "There's nobody at GTC, there's nobody better at end-to-end systems." So he had the imprimatur of Jensen was like the godfather of all this stuff, and Bill McDermott who's like Mr. Smooth up there and they say-

>> My God,

>> "... You're the best at this and I trust you more than anything." I mean, its wow.

>> What an endorsement, the social proof. What I think is really interesting about what we're seeing both from GTC and from this keynote, from these events, is the way these collaborations are out front and center. I feel like sometimes some of the partnerships are a little behind doors as people roll out a new system or a new stack or a new piece of hardware, whatever it might be. Here they're all saying, hi, we're all working with the most important and smartest people in the industry. Look at us working with the most important people in the industry. This is how you know can trust us to build your solution, multi-billion dollar enterprise who's looking to find their permanent solution.

Bob O'Donnell

>> Well, go ahead-

Bob Laliberte

>> Sorry I was just going to say, and plus if you're trying to accelerate the adoption of a technology, the last thing you want to do is be given a recipe and try and figure out how to put it all together. Here's your GPUs, here's your server, here's someone else's storage, here's someone else's networking, good luck. To really accelerate that and to drive the adoption and accelerate the value of these solutions I think a lot of these organizations are looking for that validated blueprint design. Help me understand how to put all this together for my specific needs. And I think we heard a lot about that today with organizations, the Dell organization talking about finding the specific use cases and whether they're going to be horizontal, whether they're going to be specific to verticals, et cetera. So I think we're seeing organizations really start that process, and that's what Dell's really trying to accelerate, and that's what a lot of the stories we're hearing are. Look, I think on stage, someone mentioned today, "If you're not using this in your contact center, you're already be behind."

>> "You're already behind." That was Michael. Yeah, he said that towards the end.

Bob Laliberte

>> "You're already behind." And so the key is really to get people getting the enterprises to really just get started and that's showing that they can help them.

>> That's that train analogy.

>> There's something we have to talk about here and you, I think, know this space and that's developers. So you just came off of, let's see, we're at Red Hat Summit, you were at KubeCon.

>> KubeCon.

>> Right. So you know the developers, the role that the developers play here. Dell is not known for having a big developer ecosystem, neither is Cisco. IBM paid $34 billion and got a great developer ecosystem. So they're, I think, sitting in a pretty good position there. I guess my open question would be how important will the developers be in that on-prem rollout? I would think that infrastructure as a code, you get that in cloud AI as natural language. Maybe that makes it easier for those on-prem companies that don't have a developer ecosystem to attract them. I don't know. What are you hearing from the developer?

>> It's a great question, Dave, and especially coming off of Google as well as KubeCon and talking to the CNCF, very big developer ecosystems. They mentioned on stage, everyone's always talking about, it doesn't matter what hype curve we're talking about, but Michael mentioned it specifically, they're building the easy button for AI. Now, I think that's a generous statement at this stage. I think AI is still very hard for everyone no matter your stage. But I think what's going to matter is what tools have been built with developers in mind and what tools have had the developer experience in mind. Because you can see the people throwing spaghetti against the wall and saying, "Oh, this is a tool for X, Y or Z," but that might get 10 people who really need the solution right now or 1,000 people. But the truly adopted software or tools or agents or any of that stuff is going to be the companies that iterate with developers. So to your question, anyone who wants to win in this space, in my opinion, especially if they're trying to have a full-stack solution, really needs to play ball with the developer side of things as much as they are being friendly with all their big favorite partners right now. Because if the developers don't like it, they're not going to use it and they're going to...

>> Does GenAI make it easier for the guys who don't have a developer ecosystem to get one or attract them?

Bob Laliberte

>> Yeah, I mean that's certainly,

>> Developers still got to develop though. I think, no is my answer. No. I think it trims some of the fat and the boring parts of people's days. Developers only spend 27% of their day creating the rest of its admin and cleanup and all this other stuff. So I think we'll see that percentage grow, but it's not going to eliminate any of that.

>> You got CapEx, you got innovation, you got developers, and I think that's why 69% of the people said-

Bob O'Donnell

>> Well, yeah, I mean look, here's-

>> It's interesting. I'm curious to see what they say in six months. We're going to have to redo this poll every couple of months just to pulse it out a little bit.

Bob O'Donnell

>> Well, another other thing I was going to throw out there, a different angle between the Dell and Nvidia stuff that you brought up, Dave. Is in an interesting way it feels like Nvidia is able to use Dell as a channel because let's not forget they can sell plenty of GPUs, but they have a whole software system-

>> Yeah that partnership is real cozy....

Bob O'Donnell

>> they have been developing and they do not have-

>> A lot of reason for that to work out....

Bob O'Donnell

>> these connections. So Dell has got all of those connections so they can walk into enterprises and carry along all this Nvidia software, which has been a big part of these combined offerings is the Nvidia software piece of it. And that's where a lot of people don't understand that part of Nvidia.

>> That's a really great point, Bob, glad you brought that up.

Bob O'Donnell

>> So we'll see what happens. But I think that's part of what Nvidia sees as the benefit with Dell is the ability to .

>> You heard Jensen touting it today with NIMS.

>> Totally.

Bob O'Donnell

>> That's why I'm saying the stuff that they showed at GTC I felt like went way under the radar and nobody really kind of got it.

>> That's a great point.

Bob O'Donnell

>> I think there's a lot potentially there.

Bob Laliberte

>> That was big.

>> It was.

Bob Laliberte

>> A big part of that was for inferencing, which Dell was also playing a lot of attention to conversation-

>> A lot of conversation-...

Bob Laliberte

>> saying a lot of ....

>> much more. I'm glad you just... And we'll wrap it up here in just a second, but I'm really glad you brought up inferencing. I feel like that's one of the less sexy topics sometimes in this space, but it is one of the things that's going to make the difference in what experiences are real. And I was excited to hear how much it was brought up as a true hardware nerd those types of things really delight my soul. So I was happy to hear it. All right, last question for you guys lightning round because we'll wrap this up, but I suspect we'll see both of you on this stage very much with Dave and I. What are you most excited about this week? Take off your analyst hat for two seconds. What are you most excited to learn? And it could involve your analyst's brain, but is there any cool conversation, a company announcing anything, any of Dell's new stuff they've talked about that you're really keen to check out?

Bob Laliberte

>> That's a good question. I don't probably have anything that I'm totally excited about. I was definitely interested and intrigued. I've been learning a lot more about their networking play and what they're doing with Sonic and how that can scale and do things like that. So I'm going to be digging into that a little bit more this week and I'll report on that later when I blog at the end of the week about the event.

>> And what we look forward to your analysis in our closing segment, Bob.

Bob O'Donnell

>> So for me it's the AI PCs. I've been a PC guy forever, and so this is like the dawn of a new era. Now it's not only Dell, but they're a big player here and they had five different AI PC systems. That's more than anybody else.

>> I know they flashed them up there. We all got to see it today.

Bob O'Donnell

>> They did. So to me, I'm very excited to see what happens, how we think about on-device, AI. All of these kinds of things I think are super cool. So anyway, we'll see.

>> For me, Dell's stock price is up 180% since last summer. So that's got my attention and I'm trying to figure out how much more do they have to go?

>> Yeah, well, when you tell me, just whisper it in my ear so that I can invest accordingly.

>> I was shouting from the rooftops when they were trading at 25 cents in the dollar. I

>> I feel ridiculous today.

>> There you go.

>> I must've missed that shout. I won't miss that shout next time. I'm really excited for all of our guests. We've got J.J. Davis on next with us, we're going to be talking about AI sustainability, which is super important. 30 different segments this week between us Dave, very excited. Thank you for having me here. Bob and Bob, thank you so much for joining us.

Bob O'Donnell

>> Thank you.

>> And thank all of you for tuning in to our fabulous three days of live coverage here at Dell Tech World. My name's Savannah Peterson. You're watching theCUBE, the leading source for enterprise tech news.

Day 1 AnalystANGLE

search

>> Good evening and welcome to fabulous Las Vegas, Nevada. It is day one of Dell Tech World. I am very excited to be here for three days of live coverage on theCUBE. My name is Savannah Peterson. Delighted to be joined this week by my co-host, Dave Vellante. Dave.

>> How are you doing, babe?

>> Thanks for having me in Dell Tech World. I'm so happy to be here.

>> I think it's our first show that we've ever done exclusively together.

>> I know.

>> We've done some stuff at AWS, but I'm really excited too.

>> Are you ready for it? Are you a little nervous, no?

>> No. This is like my hundredth Dell Tech World.

>> I meant more about working with me, not about them, not about Dell

>> Working with you, oh no, that's like push button, you kidding me.

>> I am super excited about the show. My first laptop was Dell. It means a lot to be here and I'm very excited that we can kick off our keynote analysis with a pair of Bobs. We have Bob Laliberte, principal Analyst at theCUBEResearch. Very delighted to have you join the team within the last two months. Welcome. Thank you for being here with us tonight.

Bob Laliberte

>> Absolutely. Thank you. Great to be here.

>> You get to hang out with me this evening as well. It's going to be a rowdy late night broadcast on theCUBE. I love that as well. And then we also have Bob O'Donnell, president of Tech Analysis. Bob, thank you so much for being here.

Bob O'Donnell

>> Happy to be here. Happy to be here. Always happy to chat with theCUBE folks. It's always fun.

>> Good to have you, man.

>> So your reputation precedes you before you got here. Dave was hyping you up, so I'm super excited. First impressions. It's the first few hours of Dell Tech World. I'm going to turn to you, Bob O'Donnell first, what did you think of the keynote? What are your first thoughts about everything happening here?

Bob O'Donnell

>> Well, look, I think the overall message from the keynote is an interesting one because what we're seeing is, and it was brought up multiple times, everybody's excited about Gen AI, but a lot of people don't know where to start. But the other interesting thing that I think came out of this, if we think about it big picture, is almost all of the focus has been on doing Gen AI in the cloud. I think the story that Dell started to tell is, hey, wait a minute. Maybe we can do some of this sort of stuff on-prem. The basic message, and by the way, Dell's not the only one that says this is, "Why move your data to the AI? Why not move the AI to your data?" And the vast majority of most organizations', data is still behind their firewall. So it just makes logical sense to do that. The question is, okay, how do we do that? Well, number one, you need the compute, you need the network, you need the storage, and you need the software and platforms.

Bob O'Donnell

And then on top of that, you start to build in things like software applications designed to work with workstations to do RAG, which is this hugely important deal as companies start to fine-tune their existing models and even AI PCs where theoretically you're going to eventually move them. So you put all those pieces together and it's like, okay, this is how companies can start to think about this idea of a hybrid AI model where you've got... Yes. I mean, I'm not saying it's going to all move to the... Obviously not all going to move on-prem, but it doesn't all have to be in the cloud either. You can just the same way we've seen hybrid cloud, there's going to be hybrid AI and the tools are becoming more widely available. The deals that Dell announced with Hugging Face and Meta on leveraging open source tools, which can be run locally is very interesting. So to me that was the big takeaway was that notion of opening up this concept of hybrid AI and making it something that a lot of organizations can consider.

>> I think that's a really great takeaway. Dave, I want to turn to you, since I know you took a, I was just going to call it Twitter, but we'll call it X poll on this specifically.

>> I did.

>> What were the results of your very informed audience?

>> Yeah, so the poll was, what's your take on hybrid AI? Same pace as hybrid cloud. Faster than hybrid cloud. Cloud wins the day. And 8% said same pace as hybrid cloud, very low. Faster than hybrid cloud is 23%. The vast majority, 69% said cloud wins the day. So interesting. I would say this, I think you're right, hybrid AI is going to be like hybrid cloud, but it's going to be different in that the on-prem vendors took over a decade to really get their act together to create the cloud operating model. It's not going to take that long.

Bob Laliberte

>> No.

>> Totally agree with you there.

>> And I would also add that in the 1980s, Michael Dell bet his business on the Wintel Duopoly, and in many ways, Crawford Del Prete brought this up on Twitter. When I did my last poll is Nvidia, are they Cisco or are they Google? And he said the best example is actually Wintel and now you have Nvidia, which is essentially a single monopoly. They've got the hardware and the software, and Michael Dell is betting big on Nvidia and the AI factory.

>> He saw that in the keynote.

>> Basically he's co-opting, or maybe Jensen gave it to him, the AI factory. We saw this at GTC, the Dell-Nvidia AI factory. Now it's the Dell AI factory which extends beyond Nvidia. It includes Qualcomm, AI PCs. Of course they'll give ARM, Intel-

Bob O'Donnell

>> AMD....

>> AMD, it will all be in there and it has the entire Dell portfolio. So they're basically taking over the AI factory. And I think Jensen's happy to let them do that.

>> Well, you could really sense that spirit of collaboration on stage. It felt like an AI celebrity A-lister show out there between ServiceNow and Nvidia, obviously making a cameo with Michael, but a true affection, but also that sense of needing to rely on each other to be both the OS and the hardware stack for our AI enterprise future. You can tell what's going on here and seeing all of them at each other's events over the last few months, we've seen a lot of it. It's interesting. You can tell who's playing nice with each other, that's for sure right now. Bob, I'm curious about your first impressions, what you thought of the keynote or any of your takeaways.

Bob Laliberte

>> Yeah, well, it's interesting that you bring that up because one of my takeaways from this morning, obviously Bob articulated a lot of the technology piece, but the partnership piece and the ecosystem piece, huge. That's one is a very big piece of this. There's no one who can do it all by themselves. So it's really going to be about how organizations, how these vendors can work together to provide an end-to-end solution. So I like what Dell's doing with the AI factory. I think it's a little bit of a misnomer when you think about the factory, you think maybe, hey, this is a big data center thing and what's going on. But in reality, they are talking about, as Bob mentioned, extending those AI capabilities across the entire enterprise through workstations, through laptops, out to the edge at a retail. And like you said, making sure that you're taking that inferencing and that AI to the edge because that's where the real-time analysis needs to be completed. Over the course of the day we've heard multiple times people talking about latency issues, which means distance works.

>> It doesn't matter if it doesn't feel real and it doesn't feel fast.

Bob Laliberte

>> Correct. Well, and especially-

>> It's not real-time.

Bob Laliberte

>> Right. And especially if you're trying to leverage AI to do scene analytics, things like that, quality control, it needs to be done real-time at the edge. So I think what organizations are trying to do... And it is interesting, your poll that you took, because if you think about it from a certain perspective, we're coming at it from the opposite angle. When we did hybrid cloud, everyone had everything on-prem and had to move it to the cloud. Today we're starting with all these AI models are in the cloud, and will be moving down to the data center or to the on-premises locations. So it is going to take time. I mean even the cloud, I think we tracked it for over 17 years of its adoption and going from just, hey, we're putting storage up to doing things, testing the waters. But the key point I think both of us, several of you made already, was the pace of change is just unbelievable. So when we come to these events, it's not uncommon to hear about products going from inception to production in 9 to 12 months, 9 to 12 weeks, months, things like that.

>> Absolutely.

Bob Laliberte

>> So the pace of change keeps accelerating all these new products. We heard about it from even the GPU acceleration and from Apex. Hey, how long do people sign up for Apex? Because concerned about the GPU cycles and how often they come out. So I think organizations really need to look at what their strategy is going to be and what we're hearing more and more are starting to develop their own AI strategy. And I think what Dell's trying to do from what I heard today, is provide a comprehensive system for organizations to onboard AI into their environment. So from the AI factory and the data center and the backend through the inferencing, through the workstations, through the laptops, all the way through the entire space.

>> Well, what hasn't moved into the cloud is a lot of the mission-critical workloads.

>> I was going to bring this up.

>> So the question then becomes... Right. So it's the Oracle stuff, obviously the IBM mainframe, a lot of the hardcore VMware workloads have not moved into the cloud. And so the question then becomes, okay, is that where they're going to apply AI? Now they're very critical, dangerous workloads in a way. You don't want to mess with them, although if you can inject intelligence into them, you can print money. So it's going to be really interesting. I feel like... I think people responded this way in the poll because the cloud folks have massive CapEx. They have the innovation, they have the LLM gardens and the models, and that's where all the action is today. But to your points, both of you, I see what Dell's doing is they're not waiting.

Bob O'Donnell

>> Right.

>> They're going hard after this. They're making investments. They're not only buying GPUs, they're doing a lot of dog fooding internally.

>> They are.

>> Now that said, they have a mandate to do that internally. They're probably better at on-prem than a lot of customers.

>> Hardware supports that. I mean, there's a lot of reasons for them to err on that side, but I think... I'm really glad you brought up the edge point as well as this core infrastructure point because a lot of the hyper-personalization we're going to experience for AIs on highly sensitive data, not the type of data we're going to be keeping in the cloud in a lot of these systems as much as possible for the sake of security and governance. So I do think there is this kind of interesting dance where I kind of disagree with your poll. I think it is going to be a split system. I think it's going to be a very balanced hybrid. We're going to have cloud situation for stuff that is lower weight and less compute versus things that are really impactful and gritty.

Bob O'Donnell

>> In some ways, though, I'm not terribly surprised by the poll, even though I disagree with the results, I'm not terribly surprised.

>> Twitter followers, they have no idea.

Bob O'Donnell

>> It's all good.

>> You should have seen how they blew the security question at RSA.

Bob O'Donnell

>> I think part of the problem honestly is people just don't know. Most people presume you can't even run any of these models locally or if they know that you can, they don't know how to do it, right? Even ChatGPT, theoretically you can license it, but it's not talked about. It's not well understood. So I think that's part of it. But I wanted to follow up on a point Bob made earlier, which is the partnerships thing and tying in what you just said, Savannah as well, the partnership even extends to their own customers. Because we are so early in this process, what Dell is seeing and doing is sharing with these customers, hey, this is what we're doing because all these customers are coming to Dell and saying, "Hey, what are you guys doing by the way? We want to figure it out." So there's very much of a information sharing level of a partnership that's just helping drive this discussion because it's kind of like, hey, we're all in this together. We're all trying to figure it out. Dell's put some policies and principles in place. They're like, we've got to know what we're doing here so that we can explain it to others. They've taken on that responsibility. So they're leveraging that as a means for, I think, gaining more response and more, what's the word I'm thinking of, trust from their customers to say, hey, look, we'll tell you what we did right? We'll tell you what we did wrong and help save you some money along the way. So I think that's part of it, and I view that as another type of partnership. Just to follow up on.

>> Well, you just mentioned there trust was another big cornerstone of the keynote. Jensen said there's no one he trusts more than Michael Dell, which is a real-

>> Oh, Bill McDermott said that.

>> Oh yeah, was Bill, sorry, excuse me.

>> After Jensen said, "There's nobody at GTC, there's nobody better at end-to-end systems." So he had the imprimatur of Jensen was like the godfather of all this stuff, and Bill McDermott who's like Mr. Smooth up there and they say-

>> My God,

>> "... You're the best at this and I trust you more than anything." I mean, its wow.

>> What an endorsement, the social proof. What I think is really interesting about what we're seeing both from GTC and from this keynote, from these events, is the way these collaborations are out front and center. I feel like sometimes some of the partnerships are a little behind doors as people roll out a new system or a new stack or a new piece of hardware, whatever it might be. Here they're all saying, hi, we're all working with the most important and smartest people in the industry. Look at us working with the most important people in the industry. This is how you know can trust us to build your solution, multi-billion dollar enterprise who's looking to find their permanent solution.

Bob O'Donnell

>> Well, go ahead-

Bob Laliberte

>> Sorry I was just going to say, and plus if you're trying to accelerate the adoption of a technology, the last thing you want to do is be given a recipe and try and figure out how to put it all together. Here's your GPUs, here's your server, here's someone else's storage, here's someone else's networking, good luck. To really accelerate that and to drive the adoption and accelerate the value of these solutions I think a lot of these organizations are looking for that validated blueprint design. Help me understand how to put all this together for my specific needs. And I think we heard a lot about that today with organizations, the Dell organization talking about finding the specific use cases and whether they're going to be horizontal, whether they're going to be specific to verticals, et cetera. So I think we're seeing organizations really start that process, and that's what Dell's really trying to accelerate, and that's what a lot of the stories we're hearing are. Look, I think on stage, someone mentioned today, "If you're not using this in your contact center, you're already be behind."

>> "You're already behind." That was Michael. Yeah, he said that towards the end.

Bob Laliberte

>> "You're already behind." And so the key is really to get people getting the enterprises to really just get started and that's showing that they can help them.

>> That's that train analogy.

>> There's something we have to talk about here and you, I think, know this space and that's developers. So you just came off of, let's see, we're at Red Hat Summit, you were at KubeCon.

>> KubeCon.

>> Right. So you know the developers, the role that the developers play here. Dell is not known for having a big developer ecosystem, neither is Cisco. IBM paid $34 billion and got a great developer ecosystem. So they're, I think, sitting in a pretty good position there. I guess my open question would be how important will the developers be in that on-prem rollout? I would think that infrastructure as a code, you get that in cloud AI as natural language. Maybe that makes it easier for those on-prem companies that don't have a developer ecosystem to attract them. I don't know. What are you hearing from the developer?

>> It's a great question, Dave, and especially coming off of Google as well as KubeCon and talking to the CNCF, very big developer ecosystems. They mentioned on stage, everyone's always talking about, it doesn't matter what hype curve we're talking about, but Michael mentioned it specifically, they're building the easy button for AI. Now, I think that's a generous statement at this stage. I think AI is still very hard for everyone no matter your stage. But I think what's going to matter is what tools have been built with developers in mind and what tools have had the developer experience in mind. Because you can see the people throwing spaghetti against the wall and saying, "Oh, this is a tool for X, Y or Z," but that might get 10 people who really need the solution right now or 1,000 people. But the truly adopted software or tools or agents or any of that stuff is going to be the companies that iterate with developers. So to your question, anyone who wants to win in this space, in my opinion, especially if they're trying to have a full-stack solution, really needs to play ball with the developer side of things as much as they are being friendly with all their big favorite partners right now. Because if the developers don't like it, they're not going to use it and they're going to...

>> Does GenAI make it easier for the guys who don't have a developer ecosystem to get one or attract them?

Bob Laliberte

>> Yeah, I mean that's certainly,

>> Developers still got to develop though. I think, no is my answer. No. I think it trims some of the fat and the boring parts of people's days. Developers only spend 27% of their day creating the rest of its admin and cleanup and all this other stuff. So I think we'll see that percentage grow, but it's not going to eliminate any of that.

>> You got CapEx, you got innovation, you got developers, and I think that's why 69% of the people said-

Bob O'Donnell

>> Well, yeah, I mean look, here's-

>> It's interesting. I'm curious to see what they say in six months. We're going to have to redo this poll every couple of months just to pulse it out a little bit.

Bob O'Donnell

>> Well, another other thing I was going to throw out there, a different angle between the Dell and Nvidia stuff that you brought up, Dave. Is in an interesting way it feels like Nvidia is able to use Dell as a channel because let's not forget they can sell plenty of GPUs, but they have a whole software system-

>> Yeah that partnership is real cozy....

Bob O'Donnell

>> they have been developing and they do not have-

>> A lot of reason for that to work out....

Bob O'Donnell

>> these connections. So Dell has got all of those connections so they can walk into enterprises and carry along all this Nvidia software, which has been a big part of these combined offerings is the Nvidia software piece of it. And that's where a lot of people don't understand that part of Nvidia.

>> That's a really great point, Bob, glad you brought that up.

Bob O'Donnell

>> So we'll see what happens. But I think that's part of what Nvidia sees as the benefit with Dell is the ability to .

>> You heard Jensen touting it today with NIMS.

>> Totally.

Bob O'Donnell

>> That's why I'm saying the stuff that they showed at GTC I felt like went way under the radar and nobody really kind of got it.

>> That's a great point.

Bob O'Donnell

>> I think there's a lot potentially there.

Bob Laliberte

>> That was big.

>> It was.

Bob Laliberte

>> A big part of that was for inferencing, which Dell was also playing a lot of attention to conversation-

>> A lot of conversation-...

Bob Laliberte

>> saying a lot of ....

>> much more. I'm glad you just... And we'll wrap it up here in just a second, but I'm really glad you brought up inferencing. I feel like that's one of the less sexy topics sometimes in this space, but it is one of the things that's going to make the difference in what experiences are real. And I was excited to hear how much it was brought up as a true hardware nerd those types of things really delight my soul. So I was happy to hear it. All right, last question for you guys lightning round because we'll wrap this up, but I suspect we'll see both of you on this stage very much with Dave and I. What are you most excited about this week? Take off your analyst hat for two seconds. What are you most excited to learn? And it could involve your analyst's brain, but is there any cool conversation, a company announcing anything, any of Dell's new stuff they've talked about that you're really keen to check out?

Bob Laliberte

>> That's a good question. I don't probably have anything that I'm totally excited about. I was definitely interested and intrigued. I've been learning a lot more about their networking play and what they're doing with Sonic and how that can scale and do things like that. So I'm going to be digging into that a little bit more this week and I'll report on that later when I blog at the end of the week about the event.

>> And what we look forward to your analysis in our closing segment, Bob.

Bob O'Donnell

>> So for me it's the AI PCs. I've been a PC guy forever, and so this is like the dawn of a new era. Now it's not only Dell, but they're a big player here and they had five different AI PC systems. That's more than anybody else.

>> I know they flashed them up there. We all got to see it today.

Bob O'Donnell

>> They did. So to me, I'm very excited to see what happens, how we think about on-device, AI. All of these kinds of things I think are super cool. So anyway, we'll see.

>> For me, Dell's stock price is up 180% since last summer. So that's got my attention and I'm trying to figure out how much more do they have to go?

>> Yeah, well, when you tell me, just whisper it in my ear so that I can invest accordingly.

>> I was shouting from the rooftops when they were trading at 25 cents in the dollar. I

>> I feel ridiculous today.

>> There you go.

>> I must've missed that shout. I won't miss that shout next time. I'm really excited for all of our guests. We've got J.J. Davis on next with us, we're going to be talking about AI sustainability, which is super important. 30 different segments this week between us Dave, very excited. Thank you for having me here. Bob and Bob, thank you so much for joining us.

Bob O'Donnell

>> Thank you.

>> And thank all of you for tuning in to our fabulous three days of live coverage here at Dell Tech World. My name's Savannah Peterson. You're watching theCUBE, the leading source for enterprise tech news.