SC24 | Adam Glick, Dell Technologies & Jason Schroedl, NVIDIA

Clips
News
More from SC24

Adam Glick

Sr. Director, AI Portfolio Marketing

Dell Technologies

Jason Schroedl

Director Product Marketing, Enterprise Platforms

Google

Dell’s stock falls on weak guidance, as more customers hold off for next-gen AI servers

Weak guidance sent Dell Technologies Inc.‘s shares into a tailspin today, as the stock fell more than 10% in extended trading.The company is forecasting lower server sales, and blamed the shortfall on enterprise customers that are waiting to snap up new systems powered by Nvidia Corp.’s next-generation Blackwell graphics processing unit chips and therefore shifting to later orders, some of which won’t be reflected in the current quarter.The personal computer and server giant reported third-quarter earnings that beat analysts’ expectations, but its revenue for the period came up light. The company said it delivered earnings before certain costs such as stock compensation of $2

Dell and Nvidia aim to reshape enterprise AI amid industry transition

AI is driving a seismic shift in enterprise infrastructure, with companies racing to integrate cutting-edge technologies into their operations. Things are moving quickly, and Dell Technologies Inc. and Nvidia Corp. are aiming to change the game on the data center with Dell AI Factory solutions.The goal of Dell AI Factory is to get something that is massively scalable. From there, people can start out with whatever their early workloads are, according to Adam Glick (pictured, right), senior director of AI portfolio marketing at Dell.Jason Schroedl at Nvidia and Adam Glick at Dell talk with theCUBE during SC24

play_circle_outline Merging Dell and NVIDIA in AI Factory for Data Center, Cloud, and Edge Connectivity

play_circle_outline Importance of partnerships in bringing AI factories to enterprise customers

play_circle_outline Navigating Challenges: Dell and NVIDIA's Role in AI Infrastructure Innovation Solutions

play_circle_outline Embracing Generative AI For Efficiency and Innovation: Why Getting Started Is Key to Success

play_circle_outline Ecosystem collaboration and open partnerships for solving customer problems.

play_circle_outline Continuing to grow and innovate with Dell AI Factory and NVIDIA partnership.

Info
Transcript

Adam Glick, Dell Technologies & Jason Schroedl, NVIDIA

Adam Glick

Sr. Director, AI Portfolio Marketing Dell Technologies

Jason Schroedl

Director Product Marketing, Enterprise Platforms Google

The Dell AI Factory, in collaboration with NVIDIA, offers scalable solutions for AI workloads, integrating hardware, software, and networking for efficient deployment. The goal is to help companies quickly implement generative AI and scale from proof of concept to production use cases. Workloads include RAG models for chatbots and vision systems for medical applications, showcasing AI's impact in different sectors. Companies leveraging AI are focused on speed, performance, and cost savings. The integration of software and hardware optimizations ensures optima... Read more

explore Keep Exploring

What is the Dell AI Factory and how has it been built with NVIDIA? add

What is the goal of bringing AI factories to the enterprise and how is it being achieved through partnerships with companies like Dell and NVIDIA? add

What have we done for enterprises with the Dell AI factor, NVIDIA, and other partners in terms of reference architectures? add

What advantage are companies investing in generative AI beginning to see, according to the text provided? add

What is the value of the partnership between Dell and another company beyond just a business relationship? add

What are some plans for growing the AI factory and expanding solutions in collaboration with NVIDIA? add

bolt Powered by CUBE AI

Adam Glick, Dell Technologies & Jason Schroedl, NVIDIA

search

>> Welcome back, everyone, to theCUBE's live coverage here at SuperComputing 2024, SC24, in Atlanta, Georgia. I'm John Furrier, host of theCUBE, with Dave Vellante, also the cohost with me on theCUBE Pod every Friday. Check it out. It's our long form podcast. We got NVIDIA and Dell here in the house talking AI and AI Factory. Got Adam. Great to see you. Senior director, AI portfolio marketing at Dell. And Jason, director of product marketing enterprise platforms with NVIDIA. Welcome back. Both Cube alumnis. Thanks for coming back. Good to see you.

Jason Schroedl

>> Oh.

Adam Glick

>> Yeah. It's a pleasure to be back.

>> To be here. So two of the biggest names in AI right now and super computing is really changing the game at the infrastructure layer that we're seeing massive progress, product, software and the show's floor, and then Dell, some great machines on the booth. We just did a review there. NVIDIA's at the center of all the action. Congratulations on the continuous success, Jason and the team. But the game's just getting started. So I want to get your perspective on the AI factory because that's a North Star, but now proof is hitting the table. We're starting to see some progress, and AI factory is going to change the game on the data center to connect to the cloud, connect to the edge. I mean, software's driving it and people are building their own systems. I mean, this is what we're hearing. Jensen said that on his last speech. Okay. You got to own the machine. Adam, I mean, Dell, you guys had the machine and now machines.

Adam Glick

>> We've literally... We've built the Dell AI Factory with NVIDIA and the whole idea is that you've got something that is massively scalable that people can start out with whatever their early workloads are, they can start small, and literally just stack that up, not only just within a rack, but create rack scale deployments. And so we make it super simple to be able to take the hardware. We've worked a lot with our friends at NVIDIA. There's a lot of work that's gone together to integrate these two pieces. So it's not just the hardware pieces and the machines that go together, but it's also the networking and the software stack, tested, integrated, optimized to be able to deploy that and deploy it fast and at scale to deliver solutions that our customers are looking for.

>> Jason, talk about the NVIDIA role and the partnership, because you guys obviously have such a great success with software and now with the hardware system, I should say. It's not hard... I guess it's hardware, but big hardware. Really changing the paradigm too on the data center. It's a machine and the software goes with it. Now you've got partners. What's the role in the AI factory from your perspective with Dell?

Jason Schroedl

>> Yeah. There's a completely new paradigm. We're seeing companies reinventing their AI infrastructure with concepts like the AI factory. We're thrilled to be partnered with Dell in bringing the Dell AI Factory with NVIDIA to the market. We introduced it and announced it back in March at GTC together. And we've continued to innovate ever since then with new capabilities. But ultimately the goal is really to bring AI factories to the enterprise, taking what we've done together at large scale. You've heard about the xAI up and running 19 days with a 100,000 H100 GPUs. You heard about the work we've been doing with CoreWeave, with 01 supercomputer. We've been doing this at scale for quite some time with Dell and other partners, but now we're bringing this to the enterprise. So as Adam said, we've packaged this up in a simplified solution that can help companies get up and running very quickly, leveraging all the best practice and learnings that we've done at large scale, bringing that to the enterprise, so helping them wherever they are on their journey, whether they're just getting started with generative AI or whether they're looking to go from proof of concept to proof of value and scale from maybe a few dozen users, maybe a handful of use cases, to 100s of use cases and 1000s of users across their enterprise.

>> I wrote a post at the beginning of this event called Welcome to the Era of Clustered Systems. I mean, servers, we saw that movie, rack and stack. Dell, you guys did good on that wave. But now there's servers, plural, multiple servers, racks and systems together. Dave wrote a post two weeks ago on his breaking analysis, a salacious headline to get attention, but it was a very... I loved it. Jamie Dimon is the New Competitor, Sam Altman. And what he was getting at in the post, which was an attention grabber, but the meat of the post was enterprises is going to have their own OpenAI. I mean, JPMorgan Chase has, for example, petabytes.

Dave Vellante

>> 150 petabytes, data.

>> So what are they going to do? Guess what they're going to do. They're going to use the AI factory.

Adam Glick

>> Sure.

>> They're going to need AI factory.

Adam Glick

>> And that's what we see people doing, is really looking at... They have these tremendous data sets that are valuable to their organization and they're training their own AIs. Some of them are going and building their own foundational models. Other ones are using those foundational models and augmenting them in order to create area specific data to really help with everything from healthcare, they're doing to customer service to digital sales. The use cases are tremendous. And we keep seeing customers come in with really new and interesting things, both from public sector, as well as private sector.

Dave Vellante

>> And so how are customers keeping up with the velocity that you guys are driving? Every time Jensen talks keynote or a podcast, I learn something new. That's what I really like about the way he presents is because there's always a new little tidbit in there. And one of the things I heard recently was in the old days, 286, 386, 486 Pentium, et cetera, Excel would run faster, but now he mentioned the AI infrastructure is accelerating, but also the software, it's not the same software. It's not static anymore. It's not just running faster. You were able to do new things. Agentic is... Last year was co-pilots. Now it's agents, swarm of agents. How are customers keeping up?

Adam Glick

>> So there's a number of pieces that go into that. First and foremost, NVIDIA's been creating some incredible software and the pace of innovation of the software they're building is truly impressive with the NIMs and the blueprints. And then we've been co-engineering with them. There's been over 350,000 hours of engineering work going to make sure that those things are optimized and work really well on the hardware so that our customers can go down and they can simply work on the things they want to work on versus focusing on how do they keep up on all these pieces. They don't have to. It's made super easy because there are these blueprints, "Hey. I want to do something where I want a customer service bot. Great. There's a blueprint for that." You deploy that with the NIM. We have a way to just put that right on the machine. And you can actually start from that spot rather than starting from, "What are the machines I need? What are the accelerators? Hey. What's my technical software stack going to be? How do I optimize it?" Making that as easy as possible and then layering onto that, we have a whole services division. And they work and they help build these things as we come up with our DRDs, our designs for these things. And so they're able to go out and work with customers and share that knowledge with them, help them come along and say, "Hey. Here's how we get you to be able to go faster with these things," because what we see is speed is incredibly important and the companies that are starting on this early and getting in. If you don't do it, you're getting left behind because competitors, you just see it moving so rapidly. And so we're working with them, with the software that's integrated, with the hardware and the services to help make sure that people can stay up to date on that.

>> It's interesting. It's almost like... I don't want to use the word onboarding the enterprise, but if you look at the evolution of, say, on the consumer side, the training and the inference, they're well ahead, but then the Llama models are getting better price performance. To the point about the Jamie Dimon versus Sam Altman, the intellectual property and the data, okay, is huge. So no one's going to just let those jewels leave the building, so to speak, or the on-premise, but you want to connect to the cloud, so you're seeing the system. So I have to ask, to the software question. Every new architecture in this wave presents new software. Transformers leveled up the old algorithms from the '80s when I went to school. Everyone knew AI, but no one really could do anything with it. But transformers changed the game. And then architectures changed... And then the software came out. Now, new architectures, this is that Intel way that Dave talks about. This is where you're at. So you're bringing the factory. Now, are you onboarding them? Kind of. I mean, a factory, start with a factory. The whole .

Adam Glick

>> Oh, absolutely. We get in there with them and we help them with it because some people know what they need. They just need the hardware. But a lot of the enterprises, hey, they're asking, "Help us. Help us pick the right use cases. Help us look at what we can. What are the best things for us to go do first that will deliver the most value for our organization?" And we're there to do that because we understand the software, the hardware, how it's integrated, and we can make that happen for

>> What have you guys learned, Jason? Because the NVIDIA has a lot of expertise. And the other thing, I'll quote Jensen, well, looked at Jensen's speeches, he says things like, "We have experience at scale." Dell, you have experience at scale with customers on your side. How do you take that scale and bring it to the factory? Because I'm just putting my use cases in now, get the low hanging fruit. And it's not just GPUs. They're key. It's networking's got to work too. All these things in the factory got to work. Take us through the progression of how this unfolds with that context of what you've learned and what are customers actually doing. Is it RAG? Is that the first use case? Is it consolidating stacks? Take us through the progression.

Jason Schroedl

>> I mean, you're absolutely right. I mean, we've learned to do this at scale with many of our largest customers and we built out reference architectures, cloud provider reference architectures that we've published and made available to our partners and our customers. We've done the same thing now for enterprises with the Dell AI factor with NVIDIA and other partners. We have new enterprise reference architectures that actually are based on those cloud provider reference architectures, but done at a different scale so that organizations that are just getting started with their initial use cases that are building out clusters that are enterprise skills fitting in their existing enterprise rack scale architecture... Whether it's with an initial use case for something like a digital assistant or whether they're building out a larger deployment using a RAG, agentic RAG, we can meet them where they are, leverage those reference architectures that include all of the compute with our GPUs, with our gray CPUs, with networking, with Spectrum-X, networking, with the software that we've talked about with NVIDIA Enterprise, the NIM inference microservices, the new agentic capabilities we have now with the AI blueprints together with all the capabilities that Dell brings with Dell PowerEdge servers, Dell PowerScale storage, the services we talked about, even though the workstations, with Precision workstations, all that packed together to be able to say, "Here is a reference architecture and a Dell reference design that customers can then take and deploy and get up and running with those use cases as quickly as possible." And in fact, we're seeing dramatic time savings and faster time to market leveraging our reference architectures and the Dell reference designs that get those customers up and running quickly, get them from proof of concept to having actual production use cases where they're delivering value.

>> What are some of the product development innovations that you guys are doing? Because this is, again... It's a starting point, but it's already built into the company. So it's got a lot of stuff in there, GPU's all these clustered systems and here, but networking's got to work too. So all this is happening. Where's the innovation? If you had to describe product innovation on the Dell and NVIDIA side with the AI factory, what is the key product innovation of the AI factory?

Jason Schroedl

>> . Oh, boy. How much time you got? -

>> Liquid cooling jumps out at me.

Adam Glick

>> I mean, liquid cooling's a big thing, as you think about the liquid that's going on. Also, just the density, if you think about the new processors and the new accelerators that are coming out and how we can get more and more dense racks. Also, how do you ship those pieces at scale so it's not just you're buying pieces parts, but you're actually buying whole systems? You're buying an integrated rack that has all these pieces built in and that takes in the advantages of what we built with the systems, what the networking is and there's specialty networking between the accelerators, as well as the networking between the machines. And so have we built systems and racks that can integrate those things and make doing that quick and easy for people to do? And that's just on hardware side. And then you move over to the software side and say... It's not just, "Hey. We've got software and it runs, x86 infrastructures and CUDA and accelerators, but has it been optimized for those systems?" And so you can get away from doing all the, "How do we tune this? And how do we fit it together?" But how do we actually make sure this is ready to go, so when you have it, you set it up and you start running it?

Jason Schroedl

>> And not only that, you get the best possible performance too, because I mean, obviously speed... Time to first token is really important for all our enterprise customers, but so is performance and so is efficiency and so is total cost of ownership. So we're working together to make sure... Leveraging the reference architectures, the reference designs we've built together, make sure we have the optimal performance for those customers so that, again, they're getting the best value for .

>> I mean, you're provisioning the infrastructure at the factory. I mean, if you look at IT, over the years, they bought a lot of servers, a lot of gear, racks, and boil that together, now here's the new thing. But now the workloads are different. So what are those characteristics? I mean, we talk about end-to-end workloads. What workloads are running on the factory what you guys are seeing starting today? And how do you see a steady-state workload?

Adam Glick

>> Sure. So to your point, the deployment and setup pieces of it, we built automation integration that can speed up that deployment of that software and getting it up and running by like 86% a reduction in that time. So I mean, a real difference in how long it takes to really get out there and getting it done quickly. And when you look at those workloads that you're asking about, we see people doing a range of things. So certainly RAG and using RAG as part of chat bots is a common one that we see people starting with, especially in things like support or even internal systems like HR. We're also seeing things like vision systems where people are using them in medical situations. Northwest Medicine is using it to analyze radiological X-rays, look at the X-ray and then provide first feedback on that. Then it uses an LLM and RAG to be able to write up the report for the doctor, so the doctor spends more time doing what doctors do best, which is not typing on keyboards and writing reports. So that'd be two good examples I could think of .

Jason Schroedl

>> Yeah. No. Those are all great use cases. I mean, you mentioned one of the examples is the digital assistant, and I think there's a great customer story that we can showcase is the city of Amarillo, which is a real compelling story where this is a city that has like 62 different languages and dialects in the small city with maybe less than 24% of the population that doesn't speak English as first language. So they have very diverse population and they just wanted a way to very easily interact with their government information services. And this is now they're creating a digital assistant that can speak to and interact with those citizens in their own dialect and language and answer simple questions, like whether it's about the city services or community planning or educational services, and get that instantaneously in their own language. And that's just one example that we've seen with the Dell AI Factory with NVIDIA with the digital assistant. But we're seeing lots of other examples like Adam talked about.

Dave Vellante

>> Guys, what's so fascinating to me about this whole AI conversation is we follow very closely companies like Dell and NVIDIA and the Mag 7 and trillion dollar valuations. Dell's valuations tripled or quadrupled in the last year. Awesome. But the amazing value that's going to be created is your customers. And so I think about, well, how... If everybody's going to have an AI factory... Which I buy into. Accelerated computing. We're going to see a massive transformation of infrastructure. It's going to democratize AI. You're going to have organizations driving AI from top down command and control. You're going to have individuals leveraging AI, AI PCs, is AI everywhere. You think about how is that going to change the structure of industries, not just the technology business, but all your customers' industries, financial services and healthcare and all these others. And it's interesting, you were mentioning the 150 petabytes that JPMC has. So they've got this proprietary advantage. On the other hand, they have a lot of baggage too. So you're going to have AI native companies that are doing the same work with 1/10 of the people, driving agents, and you're going to see this, I think, really interesting collision of new structure industries battling existing industries with tons of proprietary data. You guys both have moats, so you're like, "Great. Let's see what happens," and all built upon your infrastructure. When you think about the future, how do you think it will play out? Not that it's... I'm not asking for a crystal ball, but when you talk to customers, you must be seeing both. You're seeing the big five LLM vendors doing things that were impossible before. You see existing customers maybe taking smaller bites. What's that landscape look like?

Jason Schroedl

>> Yeah. Mean, I'd say today what we're seeing is that companies that are investing in generative AI are beginning to see tremendous advantage. And I think the difference that you'll see is the existing companies that are leveraging generative AI deploying those use cases are going to be the ones that win in their industry. There will be new entrants that'll also disrupt those industries, and they may take share, but the companies that don't get started today with generative AI, they're going to be left behind. And I think that's really the message I think that Jensen and NVIDIA have tried to share with the broader market, is that generative AI is disrupting every industry. It is revolutionizing every industry. And so every company needs to get started today and they need to move quickly to leverage this technology to drive efficiency, to drive business innovation. And the companies that harness that innovation are the ones that you're going to be able to succeed, whether they're existing players in that particular industry or market or whether they're new entrants.

Dave Vellante

>> And Dell's an interesting example, the stuff that you guys are doing internally. It's amazing actually. John and I... Well, actually you weren't there, but we were out there a couple of weeks ago, got a little deep dive on some of the internal things that you're doing. It's 100 billion dollars company really driving AI because you don't want to get caught blindsided by some disrupter.

Adam Glick

>> The AI transformation that we're seeing is really a Cambrian explosion. You're talking about the people who are on the forefront or not. It's really you need to avoid being left behind because what you can do with AI is just amazing. If I just look at what my team is able to accomplish just with the people that we have, and we've looked at it as like, "Oh, my God," now the goals for the team just continue to grow because what one person can get done is so amazing. And we're just seeing that not only internally what we're seeing in terms of our own AI usage, the own models we have, all the... We've got something like 200 different projects that we're using internally. My team is working on five or six of them. I mean, it's everywhere. It's spread throughout the company. But we've seen the companies on the forefront that dive in to that, what a transformational effect it can have on the organization. And I think you're just going to see that multiply, that it's more and more organizations realize how much this can benefit them, can benefit their customers, can benefit their constituents, can benefit their employees. You're going to see greater and greater adoption. And the organizations that get in sooner, they're going to get the greatest advantage out of that because everything has a learning curve. And as much as we try and integrate it, make it fast for people, everyone goes through that curve. We try and flatten that curve for folks, but ultimately you got to get on the curve in order to start get that benefit.

Jason Schroedl

>> And those are the learnings we're bringing together with the Dell AI Factory with NVIDIA, is bringing that out to enterprise customers to help them get up and running and get started quickly, to be able to innovate faster, to compete in their markets, to drive business value, to help improve their customer experience, to create new revenue opportunities. And we're seeing that across the board with all of our enterprise customers .

>> Your enthusiasm is awesome. And the confidence is showing now at the show here. I love that comment, Dave, about industries. On our last pod, we said productivity will change the economics. You talk about industry impact, the economics are going to shift because productivity is changing.

Adam Glick

>> Productivity's going to go through the roof. Everyone I know that starts using these tools just sees it. We see it from internal data, what I see from my own teams, what I see from other teams, and we see it from our customers. The ones that go in and implement these things, they just talk about what a Samsung SDS does and how much faster they're able to do what they need to do because their meeting transcriptions are all done with AI and the email summarization that they do. And so someone's not sitting there typing that out because unless you're a stenographer, your job isn't to type on a keyboard. The keyboard is just an interface for you to work at what the real value is, which are your ideas and what you deliver for a customer.

Dave Vellante

>> We haven't had a-

>> Killer app.

Dave Vellante

>> We haven't had a decade-long productivity boom since the PC era in the '90s and before that it was the '60s with the consumer boom. And so the AI promises that we're going to have a decade of above average productivity growth. Personally, I think it's coming and we all-

>> The other theme too we've been hearing is that it feels like the '90s, but during that wave of open systems kicked in, you saw categories emerge, new categories, networking, I mean, open networking. Hello, Cisco. Hello, TCPIP. Again, software change. Internet comes on. Boom. Here we're seeing the same thing. We're now three months into our New York Stock Exchange studio. Our first of two sets is running. So I've been doing a lot of startup interviews down there. And the common theme is, "I can't get enough GPUs, so..." NVIDIA . Number two is the startups, the entrepreneurial activity is booming because they're seeing these new opportunities. And surprisingly enough, maybe it's just New York, but there are all these young guns in their 20s. They're not doing consumer. They're doing enterprise. Why? Because the enterprise has gone mainstream. It's the whole company and it's complicated. The data's . So all the top talent is solving enterprise. You look at the startup ecosystem, it's mostly enterprise. So entrepreneurship is booming.

Jason Schroedl

>> Yeah. .

Adam Glick

>> Oh. And yeah. And you see AI shows up in all of that one. I mean, I just look at the startup communities that I engage with, and they're just... AI shows up in everything because people realize the value that it delivers. You can do so much more.

>> Yeah. Value. Bingo. Again, let's get to the value because you hit the nail on the head, because that's where the value is because we're past the ZIRP era, those interest rates. Now like, "Show me the money." Don't be profitable, but have a product that works. I think that's the value. But in the enterprise, the value is business. So the ecosystem becomes huge. Now, Dell, you guys know what the ecosystem is. NVIDIA, you're now expanding out your partnerships. A connected ecosystem is not just APIs anymore. It's data. So how is the ecosystem impacted by the AI factory? Because I mean, the AI factory has to work, but has to be open to all potentially. How do you guys view the ecosystem conversation, the relationship, how they work now? Is it more engineering-oriented? Is it more business? I mean, has it changed? Can you share your ecosystem thoughts?

Adam Glick

>> So if I think about our partnership, I mean, certainly there's a business relationship there, but it's much more than that. I talked about how many engineering hours, over 350,000 hours, that we spent. It's more than just, "Hey. We've decided to do business together," because we look at how do we solve customer problems? And it's going in and saying not just, "We want to sell you some servers or storage or networking or accelerator ," but, "How do we put that all together? And how do we solve a problem that you have?" And so when you think about the ecosystem, there are a lot of different pieces that come together. When we solve a solution and we go and we integrate it and we run it for a customer, sure, there's accelerators in there, there's PCs, there's storage, but there's also lots of other pieces. There's the software stack that sits on top of that. There's applications that sit in that. Those integrate with other systems. And it's that whole piece that gets brought together. Customers benefit because what they want is they want to solve their problem and we help them solve the problem. It moves everything faster and it creates a bigger ecosystem. One of the biggest strengths I think we have is that the ecosystem we bring together, all the different companies that we have very unique partnerships with that we bring that value to the customer when they help buy a solution like the Dell AI Factory with NVIDIA.

Jason Schroedl

>> I mean, there's a broad ecosystem of software companies, companies that are building agentic AI solutions that we're partnering together with Dell on to bring onto the Dell AI Factor with NVIDIA. So for a variety of different use cases, again, things like from digital assistants to co-generation to digital twins, building out specific solutions for particular use cases. Those ISVs and those new solutions and software capabilities leveraging NVIDIA enterprise software with our NVIDIA NIMs and our AI blueprints can now run on Dell AI factory with NVIDIA, bring those to our customers again and get them up and running very quickly with each of those different use cases.

>> We always run late because the NVIDIA, Dell conversations are great. Love to do a deeper dive. Certainly we're following the AI factory story. It's a huge concept that's happening. Final question for you guys to break is what's the next step in the partnership with you guys? Is it continue to move the needle on AI factory? Are there specific objectives on the roadmap and the relationship? Can you share what's next in the partnership?

Adam Glick

>> Well, we can share some of it.

>> Come on.

Adam Glick

>> We're doing a lot of work together. But it's about how do we continue to grow that AI factory. There's a lot more things we can do in terms of form factors, in terms of size, in terms of scale, but also in terms of the software and the solutions that are being built. NVIDIA has done an incredible job at invading at breakneck speed when it comes to the NIMs and the blueprints, and we've been matching them on what we're doing in the integrations with that and creating validated designs. So what you're going to see is more and more solutions that solve more and more customer problems built on the Dell AI Factory with NVIDIA. And I think you'll see more of that as we move forward and as we meet in the future.

Dave Vellante

>> I mean, NVIDIA's roadmap is pretty transparent. It's not like you're hiding it. You're sharing it with everybody. It's...

Jason Schroedl

>> Right. Yeah. They're clearly... There's work to we're doing together up and down the stack. So we're innovating, collaborating closely on obviously the GPU infrastructure, particularly with Blackwell and where we're headed next, and also continued innovation with Hopper. We just announced availability this week with the H200 MVL, which is our PCI form factor for enterprises deploying to Hopper at scale. And we're continuing to work together with Spectrum-X from a networking standpoint, continue to accelerate that and working not only with Spectrum switches, also our BlueField-3 GPUs, delivering value together from compute to networking and then all the way up to software with the work we're doing together with NIMs and agentic blueprints.

>> It's a great wave. Or it's like Dave said. It's like the old processor days. The next one's coming and it gets better every time, but it still works together. It's not like it's... It's open. It's not like something replaces something. This is the beautiful thing.

Jason Schroedl

>> And that's the key is making sure we help our customers be future-proofed. So as they continue to scale their deployments, as we continue to come out with new innovations, because it's moving so quickly, as we've accelerated our roadmap, whether it's new GPUs, whether it's networking capabilities, new software capabilities, new storage and server capabilities or services, we're all working together to help make sure our customers are successful.

Adam Glick

>> And that builds into the Dell AI Factory with NVIDIA, so people that build their factories now will be able to expand that when the new innovations come out together to be able to continue to grow that and drive the innovation in their organizations.

>> Well. I've always loved the AI factory. Dave knows I love the name. And then I love the GTC. Michael Dell's in the front row. He gets the shout-out, waves. The whole Dell team's there. We knew, saw that coming, but it really is a great positioning. It makes a lot of sense because this is the next level. I mean, the old era's closed, is opening up a whole nother era and just the beginning. So thanks for coming on and sharing.

Adam Glick

>> Thank you so much. It's great to be here.

Jason Schroedl

>> .

>> . AI factory's real. It's happening. And it's happening very fast. And of course theCUBE factory is doing its job, pumping out the content as best as we can. And of course, we're going to be... Got re:Invent coming up. We got GTC next year. All the top events are happening here on theCUBE. Thanks for watching.