 Hello everyone, and welcome back to theCUBE, where we're live from Dallas, Texas, here at Supercomputing 2022. I am joined by my co-host, David Nicholson. Thank you so much for being here with me and putting up with my trashy jokes all day. Thanks for having me. Yeah, yes. We are going to be talking about AI this morning, and I'm very excited that our guest has set the stage for us here quite well. Please welcome Andrea Booker. Andrea, thank you so much for being here with us. I'm really excited to be here. How's your show going so far? It's been really cool. I think being able to actually see people in person, but also be able to see the latest technologies and have the live dialogue that connects us in a different way than we have been able to virtually. Oh yeah, no, it's all about that human connection and then driving towards our first question. So as we were just chet chatting, you said you're excited about making AI real and humanizing that. Absolutely. What does that mean to you? So I think when it comes down to artificial intelligence, it means so many different things to different people. Absolutely. I was talking to my father the other day, for context, he's in his late 70s, right? And I'm like, oh, artificial intelligence, this or that. And he's like, machines taking over the world, right? Very much the dark side of it. A little bit Terminator. And I'm like, well, not so much. So that was a fun discussion. And then you flip it to the other side and I'm talking to my 11-year-old daughter and she's like, Alexa, make sure you know my song preferences, right? And that's the other very real way in which it's kind of impacting our lives, right? There's so many different use cases that I don't think everyone understands how that resonates, right? It's the simple things from, you know, recommendation engines when you're on Amazon and it suggests just a little bit more. I'm a little bit get to that one, right? To stuff that's more impactful in regards to getting faster diagnosis from your doctors, right? Such peace of mind being able to actually hear that answer faster, know how to go tackle something, right? Right point, yeah. And, you know, what's even more interesting is from a business perspective, you know, the projections are over the next five years, about 90% of customers are going to use AI applications in some fashion, right? And the reason why that's interesting is because if you look at it today, only about 15% of them are doing so, right? So when we're talking growth and the opportunity, it's amazing. Yeah, I can imagine. So when you're talking to customers, what are they, are they excited? Are they nervous? Are you educating them on how to apply Dell technology to advance their AI? Where are they off at? Because we're so early. Yeah, well, I think they're figuring out what it means to them, right? Because there's so many different customer applications of it, right? You have those in which, you know, are on the highest end in which that our new XE products are targeting that when they think of it, you know, I like to break it down in this fashion in which artificial intelligence can actually save human lives, right? And this is those extreme workloads that I'm talking about. We actually can develop a COVID vaccine faster, right? Pandemic tracking, you know, with global warming that's going on. And we had these extreme weather events with hurricanes and tsunamis and all these things to be able to get advanced notice to people to evacuate, to move. I mean, that's a pretty profound thing. And it is, you know, so it could be used in that way to save lives, right? Which is, it's the natural outgrowth of the speeds and feeds discussions that we might have internally. It's like, oh, speed doubled. Okay, didn't it double last year? Yeah, it doubled last year too, so it's 4x now. What does that mean to your point? Yeah, yeah. Able to deliver faster insight, insights that are meaningful within a timeframe when otherwise they wouldn't be meaningful. If I tell you, within a two month window, whether it's going to rain this weekend, that doesn't help you in hindsight. We did the calculation and we figured out it's going to be 40 degrees at night last Thursday. Knowing it was going to completely freeze here in Dallas to our definition in Texas, we prepare better to back to bring clothes. We were talking to NASA about that yesterday too. I mean, I think it must be fascinating for you to see your technology deployed in so many of these different use cases as well. Absolutely, absolutely. It's got to be part of one of the more exciting questions. And not all of them are extreme, right? There's also examples of natural language processing and what it does for us. The fact that it can break down communication barriers because we're global, right? We're all in a global environment. So if you think about conference calls in which we can actually clearly understand each other and what the intent is and the messaging brings us closer in different ways as well, which is huge, right? You don't want things lost in translation, right? So it helps on so many fronts. You're familiar with the Turing test idea of whether or not the test is if you can't discern within a certain number of questions that you're interacting with an AI versus a real human, then it passes the Turing test. I think there should be a natural language processing test where basically I say, fine. And you tell me what- Did you see if it was mad or not? You tell me. You tell me what I- I love this idea, David. You know. Yeah, this is great. Okay, AI lady, you tell me what I meant. Yeah. Am I actually okay? But how far from, that's a silly example, but how far do you think we are from that? I mean, what are you seeing out there in terms of things where you're kind of like, whoa, they did this with technology I'm responsible for. That was impressive. Or have you heard of things that are on the horizon that, again, they're the big issues. But anything kind of interesting and little? I think we're seeing it perfected and tweaked, right? You know, I think, going back to my daughter, it goes from her screaming at Alexa, because she didn't hear her right the first time, to now, oh, she understands and modifies, right? Because we're constantly tweaking that technology to have a better experience with it. And it's a continuum, right? The voice-to-text capabilities, right? You know, I'd say early on, it got most of those words right, right? Now it's getting pretty dialed in, right? That's a great example. So, you know, little things, little things. Yeah, I think, I love this thought of your daughter as the example of training AI. What sort of, you get to look into the future quite a bit, I'm sure, with your role. Absolutely. Where, what is she going to be controlling next? The world. The world. Like her. No, I mean, if you think about it, just from a generational front, you know, technology when I was her age, versus what she's experiencing, she lives and breathes it. I mean, that's the generational change. So as these are coming out, you have new folks growing with it, that it's so natural, that they are so open to adopting it in their common everyday behaviors, right? Yeah. But they never, over time, they learn, oh, how it got there is, because of everything we're doing now, right? Yeah. You know, one fun example, you know, as my dad was like, machines are taking over the world, is not quite, right? Even if when you look at manufacturing, there's a difference in using AI to go build a digital simulation of a factory to be able to optimize it and design it right before you're laying the foundation that saves cost, time, and money. That's not taking people's jobs in that extreme event, right? It's really optimizing for faster outcomes and helping our customers get there, which is better for everyone. Yeah, and safer too. I mean, using the factory example, you're able to model out what a workplace injury might be or what could happen or even the ergonomics of how people are using it. Yeah, should it be higher so they don't have to bend over, right? Exactly, there's so many fantastic positive ways. Yeah, so for your dad, you know, I mean, it's going to help us. It's going to take away, well, I'm curious what you think, David. When I think about AI, I think it's going to take out a lot of the boring things in life that we don't like doing, the monotony, the repetitive, and let us optimize our creative selves. Maybe. However, some of the boring things are people's jobs. So it will push a transition in our economy, in the global economy, in my opinion, that would be painful for some period of time, but overall beneficial. But definitely there will be people who will be disrupted. Tech's always kind of done that. No, but I think we need to make sure that the digital divide doesn't get so wide that people might not be negatively affected. But I know that organizations like Dell, I believe what you actually see is, no, it's elevating people. It's actually taking away, yeah, it's allowing people to spend their focus on things that are higher level, more interesting tasks. Absolutely. So a net good, but definitely some people disrupted. Yes. I feel disrupted sometimes. I was going to say, are we speaking for a friend or for ourselves here today on stage? I'm tired of software updates. Yeah. Maybe if you could. Okay, that's a fair, that's a fair, yeah. Just standardized. So AI and ML. People talk about machine learning and artificial intelligence. How would you differentiate the two? Good question. It's just the different applications and the different workloads of it, right? Because you actually have artificial intelligence, you have machine learning in which it's learning from itself and then you have the deep learning in which it's diving deeper in its execution and modeling and it really depends on the workload applications as long as well how large the data set is that's feeding into it for those applications, right? And that really leads into the versus, we have to make sure we have the versatility in our offerings to be able to meet every dimension of that, right? You know, our XE products that we announced are really targeted for those extreme AI HPC workloads, right? Versus we also have our entire portfolio products that we make sure we have GPU diversity throughout for the other applications that may be more edge-centric or telco-centric, right? Because AI isn't just these extreme situations, it's also at the edge, it's in the clouds, in the data center, right? So we want to make sure we have versatility in our offerings and we're really meeting customers where they're at in regards to the implementation and the AI workloads that they have. Let's dig in a little bit there. So what should customers expect with the next generation acceleration trends that Dell's addressing in your team? You had three exciting product announcements here. This week, which is very exciting. So you can talk about that a little bit and give us a little peek. Sure, so, you know, for the most extreme applications, we have the XE portfolio that we built upon, right? We already had the XE 8545 and we've expanded that out in a couple ways. The first of which is our very first XE 9688 way offering in which we have NVIDIA's H100 as well as A100 because we want choice, right? A choice between performance, power, what really are your needs? Is that the first time you've combined? It's the first time we've had an eight way offering. But we did so mindful that the technology is emerging so much from a thermal perspective as well as a price and other influencers that we wanted that choice baked into our next generation of product as we entered the space. Yeah, right. The other two products we have were both in the four way SXM and OAM implementation and we really focus on diversifying and not only from vendor partnerships, right? The XE 9640 is based off Intel's Datacenter Max. We have the XE 8640 that is going to be, or NVIDIA's NB Link, their latest H100. But the key differentiator is we have air-cooled and we have liquid-cooled, right? So depending on where you are from that data center journey, I mean, I think one of the common themes you've heard is thermals are going up, performance is going up, TDPs are going up, right? So how do we kind of meet in the middle to be able to accommodate for that? I think it's incredible how many different types of customers you're able to accommodate. I mean, it's really impressive. I feel lucky we've gotten to see these products you're describing. They're here on the show floor. There's millions of dollars of hardware literally sitting in your booth. Oh, yes. Yeah. Which is casual. Oh, great parts for you. Yeah, we were chatting over there yesterday. And oh, which one of these is more expensive? And the response was they're both expensive. It was like, okay, perfect. But assume the big one is more. Yeah. You mentioned thermals. One of the things I've been fascinated by walking around is all of the different liquid cooling solutions. Yeah. And it's almost hysterical. You look inside, it looks like something from, it's like, what is this a radiator system for a 19th century building? Super industrial. Because it looks like, yeah, yeah, exactly. It's exactly the way to describe it. But just the idea that you're pumping all of this liquid over this very, very valuable circuitry, a lot of the pitches have to do with, this is how we prevent disasters from happening based on the cooling methods. But... Quite literally. I mean, you look at the power requirements of a single rack in a data center. And it's staggering. We've talked about this a lot. People who aren't kind of EV, electric vehicle nerds don't appreciate just how much power 90 kilowatts of power is for an individual rack and how much heat that can generate. Absolutely. So Dell's view on this is air-cooled, water-cooled, figure it out, fit for function, yeah? Optionality, right? Because our customers are a complete diverse set, right? You have those in which they're in a data center, 10 to 15 kilowatt racks, right? You're not going to plumb a liquid-cooled power-hungry, or air-cooled power-hungry thing in there, right? You might get one of these systems in that kind of rack architecture. But then you have the middle ground, the 50 to 60 is a little bit of choice, and then the super extreme, that's where liquid cooling makes sense to really get optimized and have the best density and the most servers in that solution. So that's why it really depends, and that's why we're taking that approach of diversity of not only vendors and choice, but also implementation and ways to be able to address it. So I think, again, I'm an electric vehicle nerd. It's hysterical when you mention a 15 kilowatt rack that kind of flippantly. People don't realize that's way more power than the average house is consuming. So if your entire house is likely more like five kilowatts on a given day, air-conditioning. I don't know if you have still the paneling. Maybe, yeah, yeah, yeah. It's not where you're living, the weather. Yeah, in Austin, I'm sorry. Yeah, yeah, yeah, yeah. We'll go to California, Austin, we're going on here. But yeah, it's staggering amounts of power, staggering amounts of heat. There are very real problems that you guys are solving for to drive all of these top-line value propositions. It's super interesting. It is super interesting. All right, Andrea, last question. Yes, yes. Dell has been lucky to have you for the last decade. What is the most exciting part about you for the next decade of your Dell career given the exciting stuff that you get to work on? I think really working on what's coming our way and working with my team on that is just amazing. I can't say it enough. From a Dell perspective, I have the best team. I work with the smartest people, which creates such a fun environment. So then when we're looking at all this optionality and the different technologies and partners we work with, it's that coming together and figuring out what's that best solution? And then bringing our customers along that journey that kind of makes it fun and dynamic that over the next 10 years, I think you're going to see fantastic things. Before we close, I have to say that's awesome because this event is also a recruiting event where some of these really, really smart students that are surrounding us, there were some sirens going off there, there were competitions back here. Yeah, yeah, yeah. So when they hear that. Where you want to be. That's exactly right. That's exactly right. Well played. That's exactly right. Well played. Well fun, come on over. Well, you've certainly proven that to us. Andrea, thank you so much for being with us. This was such a treat. David Nicholson, thank you for being here with me and thank you for tuning in to theCUBE. A lot from Dallas, Texas. We are all things HPC and super computing this week. My name is Savannah Peterson and we'll see you soon. Awesome.