 Good morning, high-performance computing fans, and welcome to Denver, Colorado. We're here at Supercomputing 2023. My name is Savannah Peterson, joined by my fellow analysts, and one of my favorite co-hosts of all time, Lisa Martin. Good morning, sweetheart. How are you doing? Good morning, great. This is my first SC. I'm so excited to be here. I feel lucky to get to have the first time with you. I know. What do you think so far? What's your first impression? It's massive. It's massive. So much bigger than you would expect, right? Yes, yes. Now, you were here last year. Is it bigger than last year? I'm hearing it is. It certainly feels bigger than last year. Last year was very much buzzing, but you can tell what the rise in AI, one of the key themes of the show. There's a lot of activity. John, I want to bring you into the show. You mentioned you saw a llama walking in. Tell us about that. Well, obviously the AI. A real llama. The AI is the gift that's just giving more and more energy and hype to this market, of technical people, by the way, which are skeptical hype, but the hype is matched by people working on stuff. The llama was walking around the streets outside in downtown Denver. Just cruising. People, it was wrong with the grok sign. It was grok was the company promoting it. Llama and AI, synonymous. But that's the theme of the show so far, is the impact of AI to high-performance computing, and the intersection of the semiconductor players, and the cloud is happening, and everything in between is going to be up for grabs, up for innovation. It's going to be a good market for kind of rethinking, taking the exoflop mindset, and chips, and interconnects, and networking, and bringing that up to this performance of this next-gen cloud and edge. So, I mean, it's going to be a perfect storm of innovation. So this is going to be a changing fast ecosystem. Couldn't agree with you more. David, you and I had the pleasure of co-hosting in Dallas last year. What do you think is the biggest difference between the vibe there, and what you've felt already since you've been here? I think that, although AI was already front and center, and people's minds were last year, we definitely were talking about AI, but not as much as I think we'll be talking AI this year, for sure. Now, one thing I want to say, I'm not 100% sure that the llama thing wasn't just a coincidence, based on walking around downtown Denver last night. There may be very- There are cowboys here, maybe they're wrangling llamas. There could be a breeding population of llamas downtown. That's possibility. Could be a endangered species. Yeah. We have rock on the show, don't we? Maybe we did a llama on the show. Ooh. We'll find out. Yeah, I mean, the whole tech scene's here. But yeah, 100%, it's more going to be, okay, what is this intersection between high-performance computing and AI? And I saw some presentations yesterday where it was going back to sort of old school, super computing, looking at fluid dynamics. Okay, great. What about AI? So I think that we're going to be pulling the AI story out of guests we get to interview, and they're going to be sharing with us their intersections. Well, the explosion from a year ago with ChatGPT and GenAI has been huge. We've seen such, everyone's talking about it. Everyone's really transfixed on it. So I'm looking forward to your point, Dave, to teasing that out of the conversations that we're having with lots of different companies, including a lot of hot startups. And you know, I got to give a shout out to Dell Technologies who makes this possible. And they had the vision last year to see this coming. And I attended their community session with Dave yesterday, their HPC community day, and they had everyone there. We had all the top players you had Boeing talking about how they're going from 800,000 hours to stand up the kind of analysis with cores and with faster cloud action and how Dell is taking advantage of that. This is an opportunity for the Dells of the world to get back into the game because a lot of on-premise things going to happen. You're going to see that. And that's already kind of conversations on the hallways so far, Dave, about the net new, not repatriation from the cloud, net new workloads running on-premise, on hardware. Bare metal, that's going to have a fabric over the top. So kind of Dave, now a shift with the architecture. So I think it's going to be an on-premise wave, cloud operations wave, and it's going to be a performance wave. So that's new. And I think Dell saw that last year. So props to them for helping us sponsor this event. And I'm excited we're going to have many guests from Dell on the show today and throughout the course of the week. John, you mentioned something that I want to highlight there. You mentioned customer use cases. And I'm really hoping that we hear a lot more customer stories from some of our guests so that we can actually bridge the application portion of the AI conversation. It's one thing to talk about large language models and machine learning. It's an entirely different thing to implement it at scale in a way that matters to people outside of our nerd bubble. And I really hope that we get to see that. That's my plan for a lot of the conversations. It's really to understand that real world application regardless of industry. Yeah, Dave and I were talking at lunch yesterday, again at that Dell community kickoff and about end-to-end, the kind of end-to-end workload architecture conversations, not just the point model or AI specifically. And I think you're going to see the conversation shift to, okay, there's going to be some new things we're going to do in the workflows. Or the, as we talked about a lot yesterday, when the job's complete, which has been their world when something happens, which is part of the processing, you're going to see faster concepts. And with iteration, that's the theme of AI. You iterate and you got to do testing. So I saw in the Dell and other vendors, they have these new environments where you can go in and test and baseline your AI and then repeat it and see if that, set that up. That's new. That's the new dynamic you're going to see where AI is going to bring a discipline around. How do you instrument it? What's the iteration playbook? Which is going to kind of go old school back to the old POC days. You go in and you test up for a few days and I think Dell had a program where you can come in for, in two weeks, stand up, X to flop kind of performance for those HPC workloads. Again, we're back to the old school kind of concepts in the modern way. So I think that's going to change the ecosystem. Yeah, and I, as you know, I spend a couple of mornings a week in my Wharton classes with the CTO program, digital transformation program. So I'm in these virtual rooms with dozens of CIOs and CTOs and they're relaying their real time stories around this very subject. And I'm definitely going to be probing on some of these questions. Not so, you know, as you mentioned, it's not so much a story about repatriation, but I'm definitely hearing the lament from CIOs in very large organizations that they no longer have the sort of data center muscle necessary to build things that they know they could have built on their own just a few years ago. Now, for fear of missing out, they feel compelled to run in the direction of cloud service providers, essentially. And so it'll be interesting to hear the perspectives of folks who are providing hardware and software infrastructure into both of those worlds, both the on-prem and the cloud world. So I'm going to be channeling the questions that I've been given by my students this week. I'm really looking forward to it. Interesting that you're hearing them actually succumb to external pressures at the C level. Well, they're lamenting it at least, but they are. But yes, they're saying, you know, we really believe that we could create differentiated value if we still had these people in our data center who could do these things differently, but we don't because we got rid of those people or we moved those people into new roles. Well, who did that? Well, I did. Alrighty then. So now what are you going to do? Well, we're going to use service providers. And so, you know, this whole interesting dynamic, especially with Dell in the mix, the question of when do you need something like, you know, NVIDIA GPUs versus TPUs versus CPUs? What are the different requirements of training versus inference in an AI environment? So we're really, we're going to get our knuckle dragging hardware people knuckles scraped up this week. You must be, you're a self-proclaimed hardware nerd. I am a hardware nerd. I heard you were outed last year. Yeah, yeah, yeah, I mean, anyone who knows me knows that I'm a hardware nerd. And even when we talk about cloud or we talk about software, there's hardware behind every piece of that. So I think it's really exciting to walk a show floor and see enormous racks and cooling systems, very different type of show floor here. David, I love that you brought up students that you're going to bring some of your students' questions to the floor here. There's a lot of universities represented and a lot of projects. This is really an intersection, not only of enterprise and startups, which we talk about often on theCUBE, but also academics and academia. I think it's going to take a lot of unique minds of great diversity to write the algorithms for our collective future. And I love that we're here. I can't wait to hear what some of those conversations are going to be. I got to bring it up because no one's mentioned it yet and it was a hot topic last year, quantum. Do you think that we are going to be talking about quantum as much as we were last year, quantum hybrid? That was a big chunk of our conversation. Or is AI going to overpower that? Obviously, those two things are interconnected, but I'm curious. Well, I mean, my opinion is clear. AI is going to overpower quantum on the hype side, for sure. Quantum's had its hype cycle a day. However, I was kind of skeptical on quantum coming into here in terms of still seeing that visibility into those use cases. And I thought AI, you see faster time to value. You mentioned some of your students yesterday. This people did some pragmatic quick AI hits you can get right now out of the gate that are quick wins to kind of go to the next level. But quantum, you're starting to see that. Yesterday, Boeing pointed out specifically in the presentation that there are visibility into use cases coming quickly in traditional computing architectures where quantum will add value. Molecule work and also wing design. So in wing design, they have, they layer the fiber and the composite, whatever the material is, and they can simulate that at scale. That is amazing. So now you're starting to see the beginning of the areas where you can apply the quantum. And we've been waiting for that. So I think that's going to start the domino effect of more quantum conversations, where real scientists, as they get more horsepower, more exoflops, fix the networking, get the interconnects in there, I think you're going to see another architecture level move up in traditional HPC to embrace that quantum piece. So, yeah, so Savannah, so just to be clear, I will be talking about quantum computing and I will not be talking about quantum computing. And those two things will happen simultaneously. In different places. I'm clear, but if your mind is implode yet. And what will be the result of those two complex calculations happening simultaneously? Both. Fantastic. Well, I look forward to that hybrid model, that quantum hybrid experience. The user experience is amazing on that algorithm. You know, you've got to make the classical with the future in certain cases. Well, chip implants are coming next. That's what I'm waiting for. You know, AI chips, full AGI implanted. I mean, there are, I have met some of them at conferences, this is the detour, but who cares? There are definitely people who are biohacking and there are quite a few people in the cyborg community. There's also, yeah, there's some wild stuff. I bet there's actually people on this show floor that very much identify. Maybe we'll find out who our closet biohackers are on the show. Maybe that's like a low key challenge for each one of us. I like that, I like that. This is a neuroculture. I mean, first of all, they're going to love your Rubik's cubes. And by the way, one solve. Yes, yes. Okay, we saw that yesterday. Also for the OCD. Oh, don't do that. Don't do that. Oh, does that stress you out when they move? One was solved. One was solved, one was unsolved. Yesterday they were both unsolved. There's so many smart people here. Someone will fix my earrings. All right, so if you're watching here and you have to show, come by and check out the solving, the problem will have a little off-site little session. Absolutely. Yeah, yeah, this is your chance to get a segment on theCUBE. The faster you go, unfortunately, the shorter time you get. However, we will allow you to come solve my earrings. I have made a pact with myself to learn how to do it by KubeCon Paris because it's embarrassing. Savannah, the thing I want to get your thoughts on is this neuroculture here, Dave, you too. You've seen the audience here. They're technical and they're snarky, but they're also skeptical of height. So it's interesting to see how they react to the AI height because there's a varied element of, this is totally, look at the lambs on the streets. You're bringing a lot of height into a scientific anti-hype, anti-marketing kind of vibe. What do you guys see the reaction there? Because I'm seeing not a lot of blowback, but definitely skeptical posture. You brought up something that I think is really interesting. At KubeCon last week, it's largely a community of developers. We always say in that community, you can't sell to them. You can't get salesy with developers. They're not going to drink the Kool-Aid. I think in this community, in the high-performance computing community, they really separate and distill the hype from the myths and the reality of what the true application is. And I think that that's interesting. I think that people are here to have a good time. I saw people in light-up cowboy hats. I hope to find one this year like I had last year. But I do think that conventional marketing stunts aren't going to go as far as you might think. It's not going to be intriguing. I have to say, and you can't quite see them here on the set, but just next to us over here, there are actually published papers and research algorithms and a lot of information exchange up on booths that you would not see outside of an academic environment anywhere. So what I can say about this community to your point, I think they can see the hype through the forest of chaos right now and the noise, whatever the analogy might be there. But I do think also everyone is here to learn and everyone is here to think about how they can collaborate to be the companies that emerge in this next wave of our economy. Do you think they're rejecting the hype or accepting it? No way. No, I'm not buying the rejecting the hype. I don't think they're rejecting it. I think that, I mean, no, no, it's a great question. You know, when we talked about it a little bit last night, I think there's a bit of a celebration of the return to attention to hardware and the sexiness of hardware. I mean, granted I'm biased as a hardware nerd, but I do think that when applied applicably and without, you know, fear mongering or some chaos around what our quantum AI future could be, it's not all Skynet, I think that there actually is a celebration for everyone here to where more people in their mainstream life may actually know what they do and why that is important. So I'm excited. What do you think, David? I think Lisa and I are going to be the BS detectors on this one, if anyone comes in and says- Oh, I love that. If anyone comes in and says, oh, AI is hype. Come on, we just saw an example of- I love it. Cambridge University in the UK, casually mentioning that their cluster, their data center that they use for work that they do, has 100 million pounds worth of, I don't know if it's pounds or Euro, I think 60 pounds. 100 million pounds worth of hardware. None of it is more than four years old. If you're running an operation like that- That's a wild amount of stuff. And you don't understand that you can get another 200 million pounds if you simply say, it's for AI, then you're foolish. So there is a very, very practical, cynical reason to embrace AI if you're in the HPC-SC community. And I think that the savvy people will be doing that. Yeah, and I want to see where it is in the hype cycle versus becoming a business imperative, a business reality across industries. Well, the guy from TAC yesterday said in the Dell session, AI vindicates the HPC way. AI hardware is going to dominate. AI needs these new interconnects. And so AI is feeding into the system. So I would agree with you that there's maybe a hype in the general consumer market, which has GPT, but the nerds are embracing it as a lift. Yeah, I think it's more of an awareness and it's going to democratize. I mean, we were talking to Johnny Dallas last night and he was saying that he's seeing engineering teams of three have the same level of efficacy as engineering teams of 500. That changes the whole ball game for folks in this space. So I think we're seeing a lot of movement. I think it's really exciting. And I can't wait to highlight you too as our AI BS experts for the rest of the week. I think that's going to be awesome. David, I just want to hit on one last point that you mentioned there. I happen to be one of the judges for the Consumer Electronics Show. And I'm judging the UN Human Security for all category. So I got to see almost 150 projects that will be announced at the show already a few weeks ago. And I could not believe how many of them misappropriated the use of the term AI. Really? Oh, yeah. Everyone just threw that on there, Jack, or in their video or whatever. And for those of us who have been around hardware nerds, OGs in the tech space, we can see when someone's bluffing on that front. So I'm looking forward to what I am confident is an incredibly sophisticated and deep discourse around all of high performance computing and what it actually is. John and I plan to enjoy a delicious steak with the famous AI steak sauce. And do you plan to bring this many a good dad jokes to the rest of the week? Yeah, I'll be here, I'll be here until Thursday. Fantastic. On that note, John Furrier, David Nicholson, thank you so much for being here. Lisa Martin, I cannot wait to hear your thoughts over the course of your first super community. It's going to be fantastic. I'm pumped, I'm pumped. Yeah, I can feel your energy. I'm very excited. And thank all of you for tuning in to our live coverage here from the Mile High City. My name is Savannah Peterson and you're watching theCUBE. The leading source for emerging tech news.