 Good morning AI nerds, and welcome back to fabulous Denver, Colorado. My name's Savannah Peterson, and we're here on theCUBE streaming live. This is day three of four. It's been a thrilling show. This is our analyst analysis segment to get us kicked off. I'm joined by Lisa Martin and John Furrier. Lisa, your smile is just lighting up the set this morning. It's your first super computing. What do you think? I love it. The energy is phenomenal. Especially when I was walking in this morning, it was, I got kind of re-energized about all the influence and participation of academia here. I know. It is so fun to see that. NASA's here, the NSA is here. We've got a bunch of logos behind us. We do, lots of university logos behind us. And it was just, it's so nice to see not just the inclusion of academia, but their active participation and their active use of HPC, AI, and how it's really infiltrating and influencing every other industry out there. Totally, it's applied. It's not just a conversation or a hypothetical, which I think is awesome. I was just checking out the numbers this morning. There's over, there's over 10,000 people here. You can tell. I mean, the show floor is huge. I watched, I walked over six miles yesterday just inside the building, looking at all the exciting projects. And another interesting data point here. Over 50% of the attendees are interested in quantum. So we, that while there is a huge conversation around AI, I mean, I would say 100% of the attendees are interested in AI and some capacity. I think it's cool to know that we are seeing such a saturation of people interest in both of these technologies to do more faster. John, chips, a huge conversation, both here. I know there's some news going on. Tell us. Well, the story here, super computing is all about more horsepower, high performance computing, which basically is infrastructure, chips. And we've been sitting on theCUBE all week and the theme is chips meets cloud and everything in between is going to get innovated up and refactored. Microsoft's having their big event, their conference, annual cloud conference. And the news is they're announcing more chips. Surprising, and guess what? It's AI related. So they have a couple of things. They have accelerated with their Azure stack, MacMia 100, and also the Azure Cobalt CPU. So again, this just talks about more and more what Microsoft's doing. Again, Microsoft building their own chip, co-designing inside the house. You're starting to see people go fabulous on the chips, building their own custom specialty chips to fit into the stack. So if you look at Microsoft's news, that relates to the stories here, because everyone's showing their chips, right? You got, we had GROC, we saw GROC, we saw some other folks here. Everyone sees GPUs with Nvidia. Great, get the GPUs, stack them up, do a lot of computation, do some AI. But then building around the GPUs is going to be a big theme going forward. So again, Microsoft validates what we've been saying and seeing here and reporting, and that's going to happen more and more. And then the rest of the big news out of Microsoft is obviously they're related to AI. They've got tons of new products. You're seeing AI co-pilot. You're seeing a lot of news built into Azure Fabric, AI Studio, Microsoft-powered co-pilot. All the AI tools are targeting the developers. And I think you mentioned the students earlier. And you said those are the future developers. So the theme about chips, cloud, developers, Dave Nicholson, our analysts who's usually on the segment is on a call with his students now. He mentioned them yesterday. The other ones are going to solve the problem. So you start to see this generational shift. So to me, the big story is continuing to be chips meets cloud, everything gets in between, gets disrupted and innovated. But it's a generational shift, Savannah. This is what we've been seeing. It's young and old. So from the dorm room to the board room, as you're seeing innovation, people our age, my age, and then have systems background and getting involved, this community here, you see all the old boomers are still here. You got the young guns. And it's a cool environment. It's a true melting pot. And AI is like the fountain of technology youth. And I think that to me was the big story. It continues to be the infusion of AI, the enthusiasm, the opportunities, and people building as fast as possible. So there's confidence. Yes. I totally agree. I love that you brought up the age thing. I actually saw on the show floor yesterday a book called Quantum for Babies. So it's not even just the dorm room. I guess we're going all the way to the cradle. Cradle to Grape. Yeah, yeah. Cradle to Grape. It's end to end here. And there's that many people excited about a lot of things. On the note of chips and AI, interesting announcement. I love that I'm scooping our own outlet. Shout out to the team back at the ranch at Silicon Angle. Intel disclosed that Aurora managed to run an AI model with one trillion parameters using only 64 of its 10,000 servers. Amazing. So we're able to run a lot of models through a lot of complex calculations at scale. That's really exciting. It's everything from climate science to genome sequencing. And it's so fun that we've had so many conversations. Have you had any dialogues, Lisa, either with guests or just in the hallways or after the event that have really struck you? Anything got you extra excited? I think what I love talking about is with some of these startups who are in there. They've been working for a while. They're seeing the market dynamics, the market changes. And they're listening to customers in terms of this is what we need to do in order to deliver new products and services, differentiate ourselves, stay ahead of the competition. I always appreciate talking with new companies. I get to do that again today. Understand, what did you see from a vision perspective or gap in the market perspective? From an HPC perspective, AI perspective, enterprise AI that you knew that's a problem, we can solve it. And so there's a lot of innovation going on. And as John pointed out yesterday, it's overnight. Everything is real-time these days and it has to be. For every industry, whether we're talking about healthcare and saving lives or financial services or manufacturing, it has to be now. The one thing I haven't seen much talk of, and it's kind of implied, but it's not front and center, is security. And if you remember, a couple of years ago, they had that so-called news that was turned out to be fake, but there was a supply chain hack in the hardware and Amazon was accused of having vulnerability backdoor from the Chinese and Russians. And so that comes up a lot. So we're not seeing a lot of security conversations on the queue. I know when I talk to the Dell folks, they're hardcore on the security and they have that built into their gear. So I think, is it a good thing that we're not having security conversations because is it okay or is it getting overplayed by the hype of the AI? So that's what the one thing I would say, I haven't seen a lot of emphasis on, is what's the security angle? Specifically the silicon and the hardware. That's going to be an area that's going to be targeted. Yeah, I think that's actually, I think that's a really good point. I mean, and we're not hearing as many, I know I'm obsessed with it and I say it every day, but we're not hearing as many conversations about quantum as we were last year. And I think you hit the nail on the head there, John. I think it is that the conversation and the excitement from the crowd is so loud right now around AI and it's such a great use case for all of the hardware buzzing in this building that I think that's just been where people have focused their energy in their words. I do think that it very much matters. We're going to have more data to secure than ever. These models are going to be looking at things that are very sensitive when they are learning and doing these computations for researchers or whatever their application is. So it is going to matter. I just think, I mean, we talked about security a lot in Chicago last week at CNCF, but I think this is, I think it's just not as sexy of a conversation. Well, there's news out there. So Intel had a vulnerability. They just patched a bug effectively, virtually all modern Intel CPUs that lets code run inside the VM that crashes the hypervisor, a risk to all the cloud providers, a huge issue. So it's out there. So these zero days are out there, you have vulnerabilities. It's going to be one of those things where it's like, it's never going to go away, but it's data-driven. That's where I think AI can come into play. So again, all our other CUBE events we go to, it comes up that security will be helped by AI because of the data involved. So, and I think that's one piece. Savannah, the other thing that's interesting in the news I want to get you got thoughts on is that OpenAI and ChatGPT, since their dev day, shut down new registrations because of demand. And they can't, they don't have the capacity. They need high performance computing. So that highlights two things. One, HPC meets AI. And two, the potential impact to sustainability. Which is a big part of the conversation here. We had folks on here talking about that. So that to me is like, that's going to be a big discussion. How much power is being used by all these new open source models and who's got a good solution and where does it end? You talked a little bit yesterday about the vision of sustainable AI with Dell and I think Denver data works. And it's something that's absolutely, not only on their radar, but they're actively working towards it. Because you bring up a great point. You talked about ChatGPT shutting down new registrations because of capacity. That's what one of the main major themes we're talking about here is capacity. But looking at it from a sustainability perspective it's going to be absolutely critical. You did that interview. We also did one with Dell and at North. Correct. Again, the Nordic angle there. Yeah. And that's going to have to be a solution. Yeah. I mean. Well, it's interesting to see that dialogue was particularly interesting. So in terms of the sustainability not only is that a problem because we're talking about an incredible amount of power required, but it was cool to get the take of a Nordic company on this. That's another thing that's a big theme here. It's very global. There's a lot of different players from all over the world. Not only just of different ages and from different backgrounds, but talking to at North from the Nordics a historically sustainable place. There's a lot of renewable energy so it's front of mind for them. And I thought it was so easy to digest and such a great illustration of how you can take the warm water from liquid cooling and they're using it to heat homes in Denmark. Yeah, that's amazing. And that is, it really shifted my thinking around it. Cause I think when we think about sustainability I'm still thinking in the back of my mind like, yeah, right guys. It's still a server farm out there somewhere sucking a metric F ton of power. And the reality is like that's just the truth. That's one of the biggest challenges. I love tearing examples like that. Because that truly does make it a full ecosystem and makes this a holistic movement rather than just some run for tech money. No, that's it. It's an ecosystem. Yeah. I mean the thing about this ecosystem is going to look a lot like clouds. So what's interesting is this whole, I won't say full stack, but from the silicon infrastructure bare metal hardware to the app is going to be the innovation. What's interesting in some of the conversations where they talk about composability, de disaggregation, these are concepts that kind of bring back the on premises kind of old school data center thinking in a new way. So you're seeing a lot of kind of the old traditional enterprise suppliers get totally stoked because they're like, they know that world. Now it's not yesterday's data center. It's a new architecture. So, and it's not taking away from the cloud growth. So there's some interesting power dynamics in the industry going on around cloud growth. Are people taking their cloud workloads and bringing it back on premise? It's called repatriation. That data is not actually in their numbers. Dave Vellante has got great data from theCUBE research that show that there's not a lot of repatriation, but there's a lot of net new buildout on premises and edge. That's going to come from the AI buildout and that's why NVIDIA is such the darling right now with the GPU. So okay, if that's going to happen, what happens next? We're going to have some other vendors on early HPE and others going to talk about what that means for the customer. It's going to be a multi vendor world, multi chip world. Multi cloud. Multi cloud. Multi everything. And with supercomputing, you got super cloud, super app, super chips. It's super. Super exciting, Savannah. It is just super here on theCUBE, as usual. Speaking of super things, you know, I'm curious. I didn't notice a llama walking to set today like we did yesterday to see it. I did actually. You did. I saw it on the corner and I said hello to the llama and the grok folks. The llama's still here. Yeah, I got a photo. I'm like, pull it up, John. We'll get an ISO on that. That's fun. I love it. I missed it. I'm honestly a little jealous. The next segment, oh my gosh. The next segment that I... It's fantastic. The next segment that I have. It has given me a look, too. Side eye. You got a side eye for the llama. That's a nice grin. I think that's a super warm, good morning welcome to the world. Llama is... Agree, agree. Happy llama. You know it's a hot market when you got nice things like llamas and all these kind of like attention gimmicks out there to get people's attention. Very clever by grok. I think that was such a good call. I give props to them. It's a good play there. This community is hard to market to. That's one of the conversations that we've had to these folks really definitely see through the hype and really are looking for application. And I think that's why, to your earlier great point, Lisa, that that's why we see so many projects that are actually being realized. I mean, it's unfortunate folks can't see it back home, but I can see it in an eye shot right behind you. There are actually published research papers and diagrams, so this really is the intersection of research and enterprise. Yeah, and it's, yeah. Well, we've got a great day lined up. We have Kimberly coming back on from Broadcom and Steen Graham from Scaleless AI, hot company. And then we're going to have CMO come in from Volter. We're going to have Renny Raman, who was a stealth company, but now we can disclose it's Teresa. He's into this whole specialty cloud vision. We're going to hear from him. He's been a thought leader on that. And we've got Dell coming back on. We're talking about HCI, hyperscale-like infrastructure coming to the enterprise. Dell should do very well on this new trend of this new revitalized renaissance in the data center. And then just a great day to continue to bring people in. IBM's going to come in here. We're going to have TAC come in. Tommy from TAC, who was at the community event that Dell put on. They got a real strong view of the future of AI, so looking forward to hearing from that. But it's just another great day in paradise. It is a great day in paradise. And John, can you remind us how do you feel about Dell? I'm very impressed with Dell. The moves they're making, great products. Dell has survived every single transformation. When the web came, they were mail order. And since mail was going away, they became dominant on the web. And I'm expecting them to be dominant with AI. So we'll see, Savannah. We'll see, and I can't wait for the conversations we all get to have today. John and Lisa, thank you so much for joining me here in Denver. And thank all of you for tuning in to our four days of coverage live from Denver, Colorado. We're here at Supercomputing 2023. My name's Savannah Peterson, and you're watching theCUBE, the leading source for emerging tech news.