 Howdy y'all and welcome back to theCUBE where we're live from Dallas at Supercomputing 2022. My name is Savannah Peterson, joined with Elle Davide today as well as some very exciting guests talking about one of my favorite and most complex topics out there talking about Quantum a bit today. Please welcome Ken and Matthew. Thank you so much for being here. Matthew, everyone's going to be able to see your shirt. What's going on with Hybrid Quantum? I have to ask. Wait, what is Hybrid Quantum? Yeah, let's not pretend that everybody knows. Everyone already knows what Quantum Computing is if we go straight to Hybrid. Okay, so with a brief detour to regular Quantum Computing. Yeah, no, let's start with Quantum. Okay, so you know like regular computers made of transistors gives us ones and zeros, right? Binary like you were talking about just like half of the Cheerios. It turns out there's some problems that even if we could build a computer as big as the whole universe, which would be pretty expensive. That might not be a bad thing, but yeah. Yeah, good for Dell. Cut the mill. Yeah, yeah. We wouldn't be able to solve them because they scale exponentially. And it turns out some of those problems have efficient solutions in Quantum Computing where we take any two state Quantum system which we'll explain in a sec and turn it into what we call Quantum bit or qubit. And those qubits can actually solve some problems that are just infeasible on even these world's largest computers by offering exponential advantage. And it turns out that today's Quantum computers are a little too small and a little too noisy to do that alone. So by pairing a Quantum computer with a classical computer and hence the partnership between IQ and Dell, you allow each kind of compute to do what it's best at and thereby get answers you can't get with either one alone. Okay, so the concept of introducing hybridity, I love that word. I don't if I made it up, but it seems to fit. I'll give you credit for it. There you go. Hybridity, ding, so does this include simulating the Quantum world within the, what was the opposite of the Quantum world? Classical? Classical computer, yeah. So does it include the concept of simulating Quantum in classical compute? Absolutely. Okay, how do you do that? So there's simulators and emulators that effectively are programmed in exactly the same way that a physical Quantum machine is through circuits translated into chasm or Quantum assembly language. And those are the exact same ways that you would program either a physical QPU or a simulated QPU. So access to Quantum computing today is scarce, right? I mean, it's limited. So having the ability to have the world at large or a greater segment of society be able to access this through simulation is probably a good idea. It's absolutely a wonderful one. And so I often talk to customers and I tell them about the journey, which is hands-on keyboard learning, experimentation, building proof of concepts, and then finally productization. And you could do much of that first two steps anyway, very robustly with simulation. It's much like classical computing where if you imagine back in the 50s if the cube was at some conference in 1955, we wouldn't have possibly been able to predict what we'd be doing with computing 70 years later, right? That teenagers would be making apps on their phones that change the world, right? And so by democratizing access this way, suddenly we can open up all sorts of new use cases. We sort of like to joke, there's only a couple hundred people in the world who really know how to program Quantum computers today. And so how are we going to make thousands, tens of thousands, millions of Quantum programmers? The answer is access. And simulators are an amazingly accessible way for everyone to start playing around with the field. Very powerful tool. Wow, yeah, I'm just thinking about how many, there's, are there really only hundreds of people who can program Quantum computing? I kind of generally throw it out there and I say, you know, if you looked at a matrix of a thousand operations with hundreds of qubits, there's probably, I don't know, 2,000 people worldwide that could program that type of a circuit. I mean, it's a fairly complex circuit at that point. Yeah, I mean, it's pretty phenomenal when you think about how early we are in adoption and the rollout of this technology as a whole. Can you see quite a bit, as you look across your customer portfolio, what are some of the other trends you're seeing? Well, non-Quantum related trends, or just any time. I think it was both. Yeah, so. We're a thought leader. This is your moment. Yeah. So we do quite a bit, we see quite a bit actually, there's a lot of work happening at the edge, which you're probably well aware of. And we see a lot of autonomous MOBA robots actually lead the research office. So I get to see all the cool stuff that's really kind of emerging before it really takes over. If it's also secrets, what's coming next? Let's see, oh, I can't tell you what's coming next. But we see edge applications, we see a lot of AI applications and artificial intelligence is morphing dramatically through the number of frameworks and through the types and places you would place AI. Even places I personally never thought we would go like manufacturing environments. Some places that were traditionally not very early adopters, we're seeing AI move very quickly in some of those areas. One of the areas that I'm really excited about is digital twins. And the ability to eventually do acceleration with quantum technologies on things like computational fluid dynamics. And I think it's going to be a wonderful, wonderful area for us moving forward. So I can hear the people screaming at the screen right now. Wait a minute, you said it was hybrid. You're only talking the front half that's cat. What about the back half that's dog? What about the quantum part of it? So, I apologize, Ion Q? Ion Q. Yeah, Ion Q, because you never know. You never know. Where does the actual quantum come in? That's a great question. So you guys have one of these things? Yeah, we've built, we currently have the world's best quantum computer by some measures. Casual drop there. Yeah, yeah. No big deal. Just give me some snaps for that. Yeah, Kendall's had a pick of, yeah. So our approach, which is actually based on technology that's 50 years old. So it's quite, has a long history. The way we build atomic clocks is the basis for trapped ion quantum computing. And in fact, the first quantum logic gate ever made in 1995 was at NIST, where they modified their atomic clock experiment to do quantum gates. And that launched really the hardware experimentalist quantum computer revolution. And that was by Chris Monroe, our co-founder. So, you know, that history has flown directly into us. So to simplify, we start with an ion trap. Imagine a gold block with a bunch of electrodes that allow you to make precisely shaped electromagnetic fields, sort of like a rotating saddle. Then take a source of atoms. Now, obviously we're all sources of atoms. We have a highly purified source of metal, aterbium. We heat it up. We get a nice hot plume of atoms. We ionize those atoms with an ionizing later laser. Now they're hot and heavy and charged. So we can trap them in one of these fields. And now our electromagnetic field that's spinning rapidly holds the ions like balls in a bowl. If you can imagine them. They line up in a nice straight line and we hold them in place with these fields and with cooling laser beams. And up to now that's how an atomic clock works. Trap an atom and shine it with a laser beam. Count the oscillations. That's your clock. Now if you got 32 of those and you can manipulate their energy states. In our case, we use the hyperfine energy states of the atom, but you can basically think of your high school chemistry where you have like an unexcited electron and an excited electron. Taking your unexcited state as zero, your excited state as a one and it turns out with commercially available lasers you can drive anywhere between a zero, a one or a superposition of zero and one. And so that is our quantum bit. The hyperfine energy state of the aturbium atom and we just line up a bunch of them and through there access the magical powers of superposition entanglement. As we were talking about before, they don't really make sense to us here in the regular world, but they do exist. But what you just described is one qubit. That's right. And the way that you do it isn't exactly the same way that others who are doing quantum computing do it. That's right. And there's a lot of advantages to the trapped ion approach. So for example, you can also build a superconducting qubit where you basically cool a chip to 47 mil Kelvin and coerce millions of atoms to work together as a single system. The problem is that's not naturally quantum. So it's inherently noisy and it wants to deco here. Does not want to be a quantum bit. Whereas an atom is very happy to be by itself a qubit because we don't have to do anything to it. It's naturally quantum, if that makes sense. And so atomic qubits like we use feature a few things. One, the longest coherence times in the industry, meaning you can run very deep circuits, the most accurate operations, very low noise operations. And we don't have any wires. Our atoms are connected by laser light. That means you can connect any pair. So with some other technologies, the qubits are connected by wires. That means you can only run operations between physically connected qubits. It's like programming if you could only use, for example, bits that are adjacent with an ion-trapped approach, you can connect any pair. So that all-to-all connectivity means your compilation is much more efficient and you can do much wider and deeper circuits. So what is the closest thing to a practical application that we've been able to achieve at this point? And when I say practical, it doesn't have to be super practical. I mean, what is the sort of demonstration, the least esoteric demonstration of this at this point? To tie into what Ken was saying earlier, I think there's at least two areas that are very exciting. One is chemistry. So, for example, we have water in our cup and we understand water pretty well, but there's lots of molecules that in order to study them, we actually have to make them in a lab and do lots of experiments. And to give you a sense of the order of magnitude, if you wanted to understand the ground state of the caffeine molecule, which we all know, it has 200 electrons, you would need to build a computer bigger than the moon. So, which is, again, would be good profit for Dell, but probably not going to happen any time soon. That's kind of fun to think about though. That's a great analogy, that was, yeah. And in fact, it'd be like 10 moons of compute, okay? So build 10 moons of computer. I'm loving the sci-fi of it. Yeah, exactly. And now you can calculate caffeine or it just fits in a quantum computer the size of this table. And so we're using hybrid quantum computing now to start proving out these algorithms, not for molecules as complex as caffeine or what we want in the future, like biologics, new cancer medications, new materials and so forth. But we're able to show, for example, the ground state of smaller molecules and prove a path to where, you know, decision maker could see in a few years from now, oh, we'll be able to actually simulate not molecules we already understand, but molecules we've never been able to study in prior, if that makes sense. And then, yeah. I think there's a key point underneath that. I think it goes back to the question that you asked earlier about the why hybrid. Applications inherently run on the classical infrastructure and algorithms are accelerated through QPUs, the quantum processing units. And so are you sort of time-sharing in the sense that this environment that you set up starts with classical, with simulation, and then you get to a point where you say, okay, we're ready, you pick up the bat phone and you say, I want to... I would say it's more like a partnership, really, yeah. Yeah, and I think it's kind of, the way I normally describe it is, you know, we've taken a look at it from a really kind of a software development lifecycle type of perspective, where again, if you follow that learn, experiment, proof of concept, and then finally productize, we can cover and allow for a developer to start prototyping and proofing on simulators and when they're ready, all they do is flip a switch in a manifest and they could automatically engage a real quantum, physical quantum system. And so we've made it super simple and very accessible in a democratizing access for developers. Yeah, it makes such a big difference. Go ahead. A good analogy is to like GPUs, right? Where it's not really like, you know, you send it away, but rather the GPU accelerates certain operations. The QPU, because quantum mechanics, it turns out the universe runs on linear algebra, so one way to think about the QPU is the most efficient way of doing linear algebra that exists. So lots of problems that can be expressed in that form, combinatorial optimization problems in general, certain kinds of machine learning, et cetera, get an exponential speed up by running a section of the algorithm on the quantum computer. But of course, you wouldn't like port Microsoft Word. Yeah, exactly. You know, you're not going to do that in your quantum, it would be a waste of your quantum computer. Not just that. You want to know exactly how much money is in your bank account, not probabilistically how much might be in there. Ballpark, in the realm of. Ten moon ballpark, right? Yeah, yeah, yeah, yeah. Ten moon ballpark, I'm going to be using that for the rest of the show. Oh, I love that. Ken, tell me a little bit about how you identify companies like IonQ and end up working with Matthew. What's that like? What's it like, or how do we find out? Well, I mean, what's the process like? So, you know, let's say I've got the. Yeah, yeah, yeah, we're not going there, though. Yeah, yeah, yeah, we're not touching that today, but. It's very personal, really. Yeah. You can answer these questions however you want. You know, no, but what does that look like for Dell? And how do you curate and figure out who you're going to bring into this partnership, Ness? Yeah, you know, I think it was a very long, drawn out learning opportunity. We started actually our working quantum back in 2016. So we've been at it for a long time. Only in quantum would we say six years is a long time. Yeah, exactly. That was cute, by the way. That was like, we've been doing this for ages. For a long time, yeah, a very long time. It's before you were born, you know, like, yes. Feels like it actually, believe it or not, but. So we've been at it for a long time and, you know, we went down some very specific learning paths. We took a lot of different time to learn about different types of qubits available, different companies, what their approaches were, et cetera. And we ended up meeting up with INQ. And we also have other partners as well, like IBM, but INQ, you know, there is a nice symbiotic relationship. We're actually doing some really cool technologies that are even much, much further ahead than the, you know, strict classical does this, quantum does that, where there's a significant amount of interplay between the simulation systems and between the real physical QPUs. And so it's turning out to be a great relationship. They're very easy to work with and a lot of fun too, as you could probably tell. Yeah. So before we wrap, I've got it, okay, okay. So let's get, let's get, let's get deep. Let's get deep for a second, a little deeper than we've been. So our current understanding of all this of the universe is pretty limited. It's down to the point where we effectively have it assigned to witchcraft. It's all dark energy and dark matter, right? What does that mean exactly? Nobody knows. But if you're in the quantum computing space and you're living this every day, do you believe that it represents the key to us understanding things that currently we just can't understand? Classical models, including classical computing, our brains as they're constructed, aren't capable of understanding the real, real that's out there. If you're in the quantum computing space, do you possess that level of hubris? Do you think that you're going to deliver the answers? To be honest, I think the more you're in the space, the more mysterious and amazing it all seems. But there is a great quote by Richard Feynman that sort of kicked off the quantum exploration. So he gave a lecture in 1981. So long before any of this began, truly ages ago, right? And in this lecture, he said, you know, kind of wild at that time, right? We had to build these giant supercomputers to simulate just a couple atoms interacting, right? And it's kind of crazy that you need all this compute to simulate what nature does with just a handful of particles, right? And famously he said, nature just isn't classical, damn it. And so you need to build a computer that works with nature to understand nature. I think the quantum revolution has only just begun. There's so many new things to learn. And I'm sure the quantum computers of 40 years from now are not going to look like the computers of the day, just as the classical computers of 40 years ago look quite different to us now. And we're a bunch of apes, but you think we'll get there. Yeah, I mean, I think we've, I feel incredibly optimistic that this tool, quantum computing as a tool, represents a sea change in what's possible for humans to compute. Yeah, I think it's that possibility. You know, when I tell people, right now in the quantum era, we're in the eniac stage of the quantum era. And so we have a long way to go, but the potential is absolutely enormous. In fact, incomprehensibly enormous. I was just going to say, I don't even think we could grasp what- From the eniac age, they had no idea of computers inside of your hand, right? They're calculating trajectories, right? If you told them, like, we'd all be video chatting, you know? Right. Like, kids would be doing synchronized dances, you know? You'd be like, wow. Yeah. I love that. Well, on that note, Ken, Matthew, really great to have you both. Everyone now will be pondering the scale and scope of the universe with their 10 moon computer. 10 moons, that's right. And you've given me my new favorite bumper sticker since we've been on a roll here, David and I, which is just naturally quantum. That's one of my new favorite phrases from the show. Thank you both for being here, David. Thank you for hanging out. And thank all of you for tuning in to our Cube footage. Live here in Dallas, we are at Supercomputing. This is our last show for the day, but we look forward to seeing you tomorrow morning. My name is Savannah Peterson. Y'all have a lovely night.