In The Arena by TechArena - From Nobel Physics to Modern Quantum with John Martinis & Nilesh Shah

Episode Date: June 4, 2026

In this episode of Data Insights, co-hosts Allyson Klein and Jeniece Wnorowski sit down with Nobel Prize winner John Martinis and ZeroPoint’s Nilesh Shah to unpack where quantum computing stands today and where it is headed next.

Welcome to Tech Arena, featuring authentic discussions between tech's leading innovators and our host, Alison Klein. Now, let's step into the arena. Welcome in the arena. My name is Allison Klein. And I'm Janice Naurowski. And we have a tremendous episode for you coming to you from Santa Barbara. We're outside in this beautiful weather, and we're joined by two very, very special guests. So the first one I'd like to introduce you all to. is Mr. John Martinez.

And you recently, John, just acquired the Nobel Prize for your work in Quantum. So really a pleasure to have you here. And next to you, we have Nilesh, Shaw, who many have seen traveling all over. Nalesh, you are also esteemed into the data center space. You recently said everything from dirt to tokens. So welcome. Thank you. Yeah, thanks.

It's a great pleasure to be here with you guys, Alice and Janice. And of course, John. And yeah, so I drive business development for zero point, but I do advise across the data center space and recently also trying to figure out if and how quantum computing fits into our AI data center stack. I'm so excited for this interview. We have been exploring quantum on tech arena in 2026

with a lot of focus. And what a better person to learn from than John. So you teach at the UC Santa Barbara. And your win. in the Nobel was about your macroscopic quantum tunneling. I want to hear about that. And that really was a key enabler of modern quantum computing that we're all talking about today. Can you talk a little bit about what that early research was and how it relates to where we are in 2026? Yeah. So the research that was cited here was actually my thesis experiment in the 1980s, beginning to

mid-1980s. So it's kind of a strange event to get a Nobel Prize for your thesis experiment. Of course, I was working with John Clark, who was doing some research on quantum effects and devices, and then Michelle Devereux was also working there as a postdoc. And it was very much a collaborative enterprise between the three of us trying to figure out what's going on. So we basically had a proposal by Professor Anthony Leggett, who passed away, unfortunately recently, on stating the fact that no one had ever actually shown that a macroscopic variable, let's say currents and voltages in a circuit this big, good data-bate quantum mechanics. And that was kind of a forefront of a scientific demonstration. So we were motivated by very basic questions about whether we would see that. And then, of course, when we observe this, This started off a field, first with these quantum devices, and then shortly afterwards trying to build a quantum computer out of this. But those ideas kind of came later. We were just doing a fundamental experiment at the time.

That's amazing. And now you and two others were awarded the Nobel Prize back in 2025 as we first started out here. And for your specific discovery in the 1980s around macroscopic quantum mechanical tunneling and energy. Can you tell us a little bit more about this promise and what this kind of holds for the future of computing? The basic idea is people are trying to build a quantum computer right now, and there's a bunch of systems you can do. People traditionally think of quantum mechanics as how you describe atoms, small things. And people are working on those systems.

But it's very natural to think about building an electrical quantum computing, because our computers that we use today are based on electrical circuits. And what has happened over the years is ourselves and many other people have just developed the physics and technology to understand how to put together these electrical circuits in a way that you can see the quantum effect and then to have them coupled together and do all the complex things you need for a quantum computer. Now, you use something called the Josephin Junction for your work. For those of us who are really imbued in electronic circuits, but not into quantum,

how would you describe this and, you know, how would you describe the difference between classical circuitry and what you've created? Well, the Joseph Junction is nothing but a tunnel junction. And when people build semiconductor devices, you use tunneling fairly often to build specialized devices. So it's not something that people are unfamiliar with, but it's basically In this particular case, a Jocin junction is made from superconductors, so superconducting wires. So you deposit a thin film superconducting wire, and then you introduce oxygen, which forms a metal oxide, which is an insulator, but very thin, just a few nanometers.

And then you deposit a metal on top of that. And that thin insulating layer is thin enough so the electrons, the Cooper pairs for the superconductivity can tunnel. through it. And then you form what is essentially a superconducting weak link. Okay. And then that has particular properties that allow you to build classical devices with the Joseph injunction. People have tried to make classical computers with it for years. And then you can also build the quantum computers based on that. For the technically people here, the Joseph's injunction looks like an inductor. It has similar behaviors to that, only it's a nonlinear inductor, even with one microwave photon of energy. Small energy, it can have nonlinear effects, and then you play off those nonlinear effects to do

something useful with it. Got it. Amazing. Okay. So with that, your team has also delivered the Sycamore processor, right, back in 2019? That's right. We published that then, yeah. Okay.

So, you know, delivering the first proof of quantum supremacy, why was it? Is this such a critical milestone in advancement? So I'll tell you a little story. You know, when we moved the research group, which we're doing at Sanibuio, you see Sannevar really well, we decided to bring the group to Google. And we felt we needed at the time corporate support in order to hire people and retain them and build up the knowledge and have the funds available to build up all the hardware and processing you how to do that.

But when we moved there, I think there was some skepticism from the executives, rightfully so, that you build these big data centers, and then how possibly could you have this little physics experiment do something so powerful? So after we were developing the technology, the theory team, I was part of that too, came up with the idea of the quantum supremacy experiment, where for a mathematical problem, you could show that the data, you're getting from the quantum computer would take much, much longer with a classical computer. That's why we were able to show that. For me, as a physicist, what it also showed,

that as you build up the 53 qubits and you had this powerful processing, that the quantum computer still worked. It wasn't perfect. But the errors in the quantum computer were exactly predicted by looking at one and two-cubid systems at using high school probability to predict with the whole system, which is what quantum mechanics tells you. And for me, that wasn't obvious you would experimentally see that. And that proof kind of gave people the idea that you, yeah, you could build it and things would go right, and it would be possible to do more complicated things, which, of course, people have done in the last seven years. That's been great. Yeah, actually, I was just going to say, Alison, John, to your point about quantum supremacy,

we were at the accelerated compute show last week, Alison, and in New York, and we had a whole, several quantum computing panels. I think one question that came up is, when you say quantum supremacy, we're not talking 100x or even 10,000 X better than a classical computer. I think when you say supremacy, is it fair to say it's more like an exponential amount of differentiation. Right, that's right. And that's what was important. Some people don't like the words supremacy, okay, and they want to say quantum advantage. But I've always said supremacy because of this fact. It's not just a little bit better, but it's kind of exponentially better. And that as you make the

systems bigger, it just becomes hopeless to try to simulate that with a regular computer because of that. In fact, I think at the time in 2019, when you published this, it was a performance of 200 seconds to get your calculation. And the comparison for conventional computing at the time is that it would have taken 10,000 years. Right. And that's been debated of, you know. Well, actually, people have come up with better classical algorithms and they can now compute that. But what we did is we took data for harder systems. Yeah.

And then as they figured out the algorithms, they verified that our data was still good. But then, you know, all you have to do is increased by a modest number of cubits and then it becomes uncalculable again. So we actually put that in the paper and it was clear, you know, that's the game we were playing. Sure, of course. I think that when I look at that and I think about the performance characteristics, I think that one question that I have is, we know that quantum doesn't necessarily apply to every single workload. And so where is this performance going to be best utilized and aimed when you think about the physical sciences or even beyond.

There have been a couple of papers this week about breaking cryptography systems, and that's actually kind of the easiest and smallest systems you can do. And the projections, people have figured out a lot, and the projections are to build something that people are thinking we should be able to build. I give an estimate of five or ten years. Might be shorter. I think it's going to be harder.

But something you can really imagine. And I think the cryptography world has to really come to grips that these, you have to be prepared for the quantum computers, people are working on it for years. I'm particularly interested in using a quantum computing, not to break the internet, okay, but to do something useful. In my particular use case is quantum chemistry, quantum materials, figuring out how to optimize materials better. maybe you can replace rare earths, we're not rare so earth, make it more ecologically sound, make it cheaper, so we can electrify transportation easier, things like that. There's other applications too in optimization that people are quite interested in. Yeah, I think it's interesting. You brought up cryptography, and I think the research I think you're talking about is the Bitcoin is based on elliptic curve cryptography, and I think the recent paper showed it requires, maybe only 10,000 cubits to about 500,000. It's still a big range.

Yeah, it depends how you build it and the like. But these are numbers that people think we could actually build, and people are planning and proposing that. So that's why it's interesting. Like I say, that's maybe five or ten, or maybe it's harder than we think. But it's something we can really imagine building. Yeah. Question for both of you guys. So in this very very moment, what do you guys think these computers are genuinely capable of in this moment? Right now? Yeah. Right now, people are still doing kind of demonstration experiments, academic experiments, testing out the algorithms, see what kind of ideas work and not work. So the algorithms are getting smaller over time, but at the same time, that's getting more

realistic, and especially these last, since this week. And then we're trying to be. build bigger and bigger quantum systems. And it's unclear when these are going to meet. It could be this week. Someone will figure something out. It could be three years, five years, ten years. We don't know. But everyone is very optimistic that this gap is closing. And we all feel they can be done. Yeah, I think at least from what I've been seeing in the application space, the financial world seems to have jumped on these quantum processes. For testing the potential, they may not be able to do something, let's say, optimizing a portfolio like Janice to your point today.

But they want to be prepared, as John said, it could be next week when this inflection point arrives. I think people want to be prepared and integrate these flows because how do you actually tie it to a practical application? So you don't want to be caught unprepared, kind of like what happened with the AI space, a lot of people were not prepared for this. I think that's what the finance space, the enjoyment. Optimization for supercomputing labs.

They seem to be building these out to really be prepared for that. Yeah, and I think that's really smart because if you have an experienced team that's been working on it for years, if something new comes up, a new idea, you can then jump on that very fast and take advantage of that. Sure. Sure. Now, we know where we are today. It seems like the zeitgeist is the error correction and reliability and sustainability of quantum computers is the hurdle that we're looking at. How are we going to get over that?

And what do you think is going to be happening in terms of the industry's work and all of research work to get over this hurdle and make a quantum computer viable for commercial applications? Yeah. So, you know, I've been thinking about this very carefully for the last five or so years. And with super inducting qubits, there's. There's a lot of advantages there. It works.

People are scaling them up. They're fast. All the things you want. But if you look at the quantum computers now, I call it the golden chandelier. Okay? Well, what you see is tons of wires and microwave components and the like. And this is very good for initial things as we're understanding the physics. But what we've been thinking about is how to manufacture that better and basically turn what

looks like a classical computers from the 50s and 60s with the mess of wires to the integrated circuit. So we're working with the semiconductor fab people to make the fab processes better and packaging on how to package it in a way so you can put, let's say, a few large chip wafers together to build a big quantum computer. So, you know, we have to kind of reinvent everything. Sure. But we think this is the modern way to do it. And we really think that the semiconductor industry has a lot to teach us there. Sure. Yeah, I think one interesting point is also how quantum computing scales. So when we go to classical computing, you know, we had Moore's law to guide us. And it's kind of a roadmap, you know, exactly what's going to happen, the doubling effect, et cetera. With quantum, I believe there's

a similar trajectory, which is projecting where we are today and based on where we were say five, 10 years ago, and maybe there's some projections out. We talk about qubits. I think everyone's familiar with that. But then I think when you double click down, as John was saying, the supply chain. So besides the qubits, you also have the actual gates. Then the other thing, like in our classical circuits, we care about the size of the chip, the size of the wafer, which drives the economics.

And then all the wiring actually on the chip, off the chip, and for cooling, etc. And then there are those metrics as well for being. able to scale, like, how deep of a circuit can I build, right? It's great if I have, say, 10,000 qubits, but they're only so good. If I can only do two steps of a compute, then there's not much I can squeeze out versus if I had, say, a circuit depth of a million gates, so I can do some meaningful logic. But again, there's all these, like you mentioned, error correction becomes crucial in that spot. But in terms of thinking about Morris law, Moore's Law was there even before the integrated circuit,

but it was the integrated circuit that really allowed that to go for a long time. And we think the same thing. You look at the increasing complexity. It is getting complexity over time. You're seeing that, but the slope is too slow. So we really want to go to the integrated circuit where we think we can rapidly increase that. Nice. But, of course, you know, you have to get everything to work.

If it makes the cube, it's better. it's a hard road that you have to master in the next few years. Yeah, and I think another metric is, do you think, like at least in the classical world, we think of scale up systems, scale out systems. And the quantum world, I don't think it's that clear. Like, you scale, you want to keep everything on a chip, or recently we've seen with some inference chip companies

actually taking the whole wafer and building out the circuit at the wafer scale. But then there are also these scale out. systems where you have interconnects, we have multiple, you have racks of computers. So when we think of classical scaling, we think of, okay, is it chip, is it wafer, is it a rack, an entire data center? But in the quantum world, it's maybe slightly different that you need to think. But for example, if you take the way people are doing superintendence, which I'm familiar with, you take the technology there, maybe improve it. You're talking many billions to tens of billions of dollars to build a million cubic quantum computer. Obviously too expensive. What we're trying to do

is do mass manufacturing of wafers and packaging so that, okay, initially it'll be in the 10 millions of dollars or more, but you can imagine that price coming down over time as you develop the process line, just like what's happened for chips. So that's kind of the way we're thinking about it, is how do we, I guess you call that scale out. Scale out. Yeah, so how do you scale out? So how do you scale out. And again, the semiconductor industry knows how to do this. And we want to use that. And that's how you get to commercial viability. Yeah. But you have to prove viability. You have to prove you can make them good enough initially. And I think you asked early, Alison, about, you know, the applications that I think John gave some great use cases like cryptography, quantum chemistry. But then ultimately,

it comes down to the economics. If the problem you're solving, if it costs more to build that system, the, let's say cracking Bitcoin, right? Bitcoin is worth a trillion. I'll just make it up. But so is it going to cost me more than a trillion to break something which could be fixed? And how does the economic, to scale out in terms of the physics and the engineering, but also scale in terms of economics is, I think. Yeah, a crucial point that I think actually that came out at the accelerated compute show where we had this roundtable of several quantum computing companies, hardware, software, and then integration parts.

So these are the things, I think, people are now figuring out. But I think, John, is it fair to say quantum has moved from being a science to now being an engineering problem? Or would you say it's... You know, when I moved to... I was at UC Santa Barbara that I went to Google. And it was great that we could do the engineering much better. Part of the reason I left Google is I didn't think they were focusing on the engineering

enough. So that's what we're doing, but I'm going to say we're doing something new, and we're really thinking about manufacturing and how to keep the cost down and how to manufacture it in a cost-effective way. So I've moved past engineering to manufacturing, and we're working with partners who understand that. Who knows how to manufacture these complex integrated circuits. So just to bring it back to real-world applications, what's kind of that one killer use case that you see quantum really driving? Well, right now, people are very excited about Bitcoin mining and other things. And I don't want to build a quantum computer to break the internet, okay?

So we can leave that aside. Anyway, I actually think that people will probably try to fix this before it becomes a problem. And that's why I think everyone's warning them that this is an important thing to think about. The one use cases I mentioned earlier, being a physicist, I like thinking, of that of simulating quantum phenomena. So, you know, we use classical computers all the time. I design electronics. I design my circuit boards.

When we remodeled our kitchen, we had a little program to design it, making sure it's okay. So the simulation of computers, of real physical systems, saves you a lot of money and time, and it's a big part of what they do. And I think if we could do that for quantum processes, too,

that would save a lot of money. And just going to where their value is, the big pharmaceutical companies spend billions of dollars to develop a drug. Sure. Okay. And there's a long sequence. But if you could somehow use a quantum computer to just save a percent, few percent, whatever in that long development cycle,

that's worth a lot of money for them. So I think that's an example, which is maybe in the millions, tens of millions of dollars, for a design, maybe more if it helps the drug become usable. But then I think there's more cases like that. So I actually think that's a profound use case that we should be looking at. So I want to go to where we are today, which is in the midst of the AI era. And I'm just going to ground us in conventional computing for a second.

Nalesh, he spent a tremendous amount of time looking at future technologies for the data center and build out of AI factories. and we know that there are challenges in terms of acceleration of technology to go faster and to meet this massive demand for AI. We also know that people are talking about how does quantum fit into this environment. Where do you think we are in terms of that integration of the quantum world and the data center world? I think so far if you go to the hypers that are renting out quantum computing systems, I think what you'll find today is quantum computing systems,

aren't really co-located with classical computing systems. So it's not like I have a data center and I have all my GPUs and CPUs. And right next to it is a quantum processor. In reality, those still sit somewhere else in some other facility because there are special requirement. But I think where we are with AI, right, and if you look at how is AI driven today, it's all about power. I have a megawatt of power.

Your input is power and your output is... That's the unit of measure of the data centers now. It is. Actually, not even megawatt. It's probably now gigawatts. I put in a megawatt and I need to get something out, a token. Now, what do I put in between? Honestly, I don't care.

Is it a GPU? It's a CPU. Now we're seeing hybrid systems where you have specialized inference pipeline. So like a few years ago, it was all about, oh, I need to train these foundations. models. Now we've moved to, well, we've mostly trained and figured how to do that. Now we need to extract value out of it. So I need inference. I need those tokens. Now the question is, when you have this kind of an economy where it's really about, I have so many megawatts, and that's the finite supply. If you look at the world, what's finite? It's the amount of energy that you have available,

to do whatever you want. So one megawatt in, tokens out. And by the way, not all tokens are. And by the way, are created equal. So it might take more to create, let's say, a video generation token versus say just a simple chat or maybe text to speech. Now the question is, where does quantum, and I should say if quantum actually fits in, it could do two things. One is it can help make that economics more efficient in terms of the one metric is power. The other metric we have is latency. So this is an impatient world, and I can't wait for an answer. I can't wait three seconds. I need it in 0.1 second. So just simple metrics is can it solve or address the power problem for this AI, this huge AI demand? And then the second is can it address the latency challenge?

And I would say yet another third problem is there are many data centers that have been built around the world. They're not ready for AI. They don't have gigawatts of capacity. But they have some megawatts of capacity and it's really expensive to retrofit them. Now, is there a use case where I could put in a few quantum processors and all of a sudden those spaces which data centers, which would have gone wasted otherwise, or I would have to build these gigafactories and somewhere in the middle of Texas, can I suddenly find a use? I think those are three ways that I look at, power, latency, and then can I actually reuse my legacy infrastructure that's already been built around the world and can quantum move the needle?

And yeah, talking about the later point, one of our collaborators, so co-lab stands for quantum collaboration, is you a Packer Enterprise who makes supercomputers and they bought Kray, for example. And that's more of a scientific computing endeavor than typical data centers. But given that there is a strong scientific computing component of this, and they know how to break up problems. They have a lot of software and experience and doing that. They're quite interested in the co-location of a quantum computer with their data centers. Sure.

So I just wanted to ask you a question. This is a little bit off topic. Okay. So what do you think about data centers in space? Oh. Oh. Well, again, it comes back to three things. It's power, latency, and then reuse.

Can I efficiently reuse the infrastructure? structure that I already have. So, yeah, I think there's a lot of promising use cases in terms of latency and availability. Availability is another key parameter. And I guess one of the ideas is the power that's available there. Exactly, yeah. So if you can use all of these efficiently, these metrics, I think that makes the case. Well, I mean, if you know that we are tapping on power and you can put a data center in space and tap unlimited solar, that could be something that would be very attractive. If you can solve the other parameters, it would be wonderful. Absolutely. So what I see right now, and this is what we heard from all these other companies that

showed up at the Accelerated Computing Show, was today it's actually quantum computing that is leveraging the power of AI and GPUs and CPUs to make these quantum computing elements more stable error correction. You brought up error correction. All of that error correction is being done on GPUs because only that has the parallel processing capability. And I think there are three things, at least that I observed in the quantum world. There's the quantum real-time error correction that you need to do because the quantum states are stable only for so long. Then you have a system-level error correction because you have this complex system of qubits and at the system level you need to do some error correction. And the third is algorithmic stability. So you have all these algorithms,

you need to be running stuff on the classical compute side. So you can make sense of what's happening in the quantum space, right? The quantum space is nebulous. So to make sense of it, you need to run almost like a digital twin on the classical side. So today it's actually the other way around is that the AI infrastructure that we have, GPUs and storage, because there's a lot of logging and things like that need to happen, you know, maintaining states and calibration. So that is actually driving a lot of this classical compute rather than the other way around.

At least today, that's what I see. But I think at some point it'll be interesting where quantum now turns around and is now assisting the growth and the scalability of AI systems. So when do you guys think that is? Like, will there be an inflection point in quantum similar to what we're experiencing in AI today? Yeah, the standard code I give is five to ten years. But with the caveat that people are being very optimistic now and maybe a little bit unrealistic about the system engineering challenges, but it's hard for me to know because a lot of these programs people are doing, they aren't talking about the details.

And you tend to hear only what's going well and what the problems are. But that's fine. But that's a good number. If you listen to Sundar Pichaya at Google, of course they have their AI program. and there. He says three to five years. Now, as a CEO, you have to be early and not late. Okay. So I would say our numbers are somewhat compatible. And I think that's a good time scale to get ready for this in terms of the encryption and crypto and whatever. So for me, that's a good timeline. We have a little bit of time, but you have to start working on it right now.

Yeah, and I think another viewpoint is, again, talking to David Moring, where he's at Cambium Capital and then also part of 55 North. I like to follow the money. Where's the money going? And in order to answer that question, that's another angle you can look at it. And actually, he put it nicely. What he said is how many people predicted the chat GPT moment in advance. So I think with Quantum, we are at that stage where, and I think, John, you said it earlier, it could be next week, right? There could be this. breakthrough of systems, hardware, software, and integration, which could unlock that value. So I think the smart money, I'll call it, approach is that, honestly, no one knows what is that, is there like a chat GPT moment for quantum computing? But at the same time, the belief is that you need to be prepared, it is coming, we see it. And that's why there's a lot of investment going on. And actually around that, the base quantum systems as well, right? So there's superconducting, neutral atom, there's a whole bunch of different modalities that people are investing in,

because they have slightly different characteristics and strengths and points of improvement, but it's the software layer around it, the control, the error correction, and then the integration with classical system that people are investing in, which will actually accelerate that quantum supremacy that John was talking about. That's where we are. Just going back to the chat GPT moment, which I think you know, my understanding is that Transformers paper was a big moment scientifically. This has happened at Google and that the authors of that wanted to, you know, scale it up and the like, but there wasn't a lot of support at Google for whatever reason. It's expensive. But then they went to Open AI where they had the support to do that and then they

could prove it out. And what I'm trying to say here is this is going to come potentially from something that you didn't expect, some new idea or somewhat pushing it in some direction based on things that we're already seen. It's nice because it's a little bit unpredictable. Yeah. Right. Exactly what's going to happen here. Does that make sense? I think it makes sense. And I think the question that people are asking is not should I invest in it, but the question is, can we afford not to invest. Right. And I think the other angle to think about is sovereign capabilities, right? So right now, we live in a different world. So everyone, countries are really concerned about having their own capabilities, technology leadership could make the difference between

having an unstable or a stable world. I mean, not to play on quantum computing. But I think, so that's another angle to look at it is there's a lot of sovereign investment going on into quantum computing technology has become a national mandate, not just in the United States. It's a geopolitical imperative. Yeah. When I was at Davos in January, I think there was a lot of concern in Europe that they had missed out a lot on the large language model development. And they want to make sure they're investing in quantum so that that won't happen again.

And it is true. That's going very well. Canada's good. China, of course. Australia, you have these national efforts that are very substantial so that they will be able to follow this in a good way. Yeah, and I agree with that. So there's all these other vectors. It's not just the technology and all, is it going to be profitable? There's all these other angles that people are looking at in order to have that advantage when the quantum scale problem is

solved. Yeah. Now, John, you founded Co-Lab in 2022. Yes. And good year to found a company. That's when we founded Tech Arena too. But you have a focus on manufacturing that you've talked about. When you look at this moment that we all know is coming, what do you want Colabs' contributions to that to be? The issue is there's a lot of companies like Google and IBM, any of Righgetti, other companies, Super League Space, other places that are doing quite well in building good systems. However, when I look at those systems, they have to be. more scalable, more reliable, and the like. So what we're trying to do is fix kind of the not

talked about problems. In the end, it goes to manufacturing, but also it goes to reliability and quality of the qubits and making them better. So we're trying to solve the really hard problem. Then I'm going to say most people don't want to solve because it's so hard. But, you know, that means there is a business opportunity, and we're just really working hard on doing that. And we find that by working with the semiconductor experts, they give us a lot of insight as to how to make it better and how to build it properly. So we combine our quantum expertise for many years with their manufacturing and fabrication expertise to figure out what to do. It's different than what everyone else is doing. It takes a long time to develop it so it's better than the current technology.

But I think it's going great. and I feel very confident on that. That's great. Yeah, and I think there's a lot of support from the classical world. So let's say if I'm doing AI, I'm making trillions of dollars, just setting up data centers and renting out GPUs or spitting out tokens. But the motivation is in the short term, quantum is driving a lot of demand for classical compute and storage.

So I think there's this motivation to actually invest in supporting quantum. kind of like a circular economy almost where it benefits the classical computing. Well, I was talking to a hyperscaler that rents out quantum processors. And last year, they said they had a record year. And I was like, what does that mean? So they did several million dollars of rentals of quantum computers, which is great. And of course, it's not anything compared to where we see the dollar numbers for GPUs, for example. But what was interesting that they said is that actually led to actually a greater multiple and demand for classical compute in order for those users to consume the quantum processing unit. Many people might say, well, I don't care about quantum, I'll wait.

But now there's a real incentive where it's actually creating consumption of classical computing storage, compute memory in order to consume quantum processing capabilities. Amazing. We've covered a lot of ground we have. And I think in this space, we can continue to cover lots of ground. But is there anything, John, that we haven't hit on that you would want to... I think we've covered it pretty well. Let me think about it. Yeah. I have a question. Yeah. Oh, go for it. Allison. We're talking about a change that's coming in the next five years. If you were talking to students who are looking at scientific careers at this point, how would you advise them in terms of what directions that they should be going in for the post-quantum world?

Yeah. So, okay, that's a nice question. especially since I'm kind of an educator. It's really wonderful for me, having seen what this field was like in the 1980s, to now have thousands of quantum physicists who are well employed by companies and other places we do our part in that too. That's really great. And that we're really studying and solving very difficult problems in that. But what I would say is when you're doing something like building a quantum computer, it's a big system engineering problem where you have to know about electronics, microwaves, maybe optics and the like. And it's an opportunity for students to grow in their knowledge of technology. And I think that's something important to remember that you might be

interested in the subject. And of course, you have to take the physics courses. But you also need to take the engineering courses. You need to know how to program and you need to, you know, understand all these other practical engineering field. And I think that's good in the long term for people to have a little bit more practical basis for their knowledge, because that's going to help everyone in the long term to have that broader understanding. I love that. Yeah, I think you're absolutely right. I mean, I think back to when I started dabbling with quantum simulations and It was back at Intel when Intel had this open source quantum simulator called Q Hipster, and we're trying to simulate 50 qubits with one of the quantum chemistry initiatives,

one of the, with Harvard University. Just trying to understand the application, first of all, and that became the motivation to double-click down into, oh, this is interesting, let me learn more. And then actually trying to use real quantum computers. Maybe back in 2018, there was like a five-cub. computer available from IBM. I remember logging in and like building toy circuits.

I think just having that motivation for the students coming on because right now it's all about machine learning. I want to become a machine learning data scientist. I want to do LLMs. I want to build agents. So there's a lot of excitement. So the question is how do we generate that same excitement? And really there has to be some motivation, which is job opportunities, career, doing something

a breakthrough and really conveying that story to the broader audience and motivating them to actually take this on as something that you want to pursue. But yeah, it is a multidisciplinary type of a field. And I was talking about experiment and you're talking about theory. And in theory, it's the same way. It's very multidisciplinary. There's a lot to learn. And you can be useful. When I was at Google, one of the people who really made a big difference in bringing down the number a qubit you need for algorithms, had a very conventional software background, and he learned quantum on his own. And because he didn't come for the standard physics, he started breaking these limits that everyone put because they made wrong assumptions on it. And he was more broad

of that. So I really encourage people from other backgrounds to come in because your kind of knowledge and unique understanding of quantum and other things can really help the field. Yeah, and you could win the Nobel Prize like John. And I think that's a great place to wrap the episode. Denise, do you want to wrap? Yeah, this was a fascinating conversation, obviously not a quantum physics expert myself, but I've learned so much and you're a phenomenal educator.

So thank you, John for all the information today. Where can others go to learn more about your new company and just about the overall space. So our company website is CoLab QO-L-A-B. We're trying to invent a word for Scrabbler, so you'll need a you on it. So it's colab.a-I, and we have a lot of material posted there,

and you can see what's going on. And how about you, Nalesh? Yeah, so, of course, you can reach me at Nalesh. Dot Sha at ZPtCorp.com. Great. To both of you. Thank you so much for the time today. It was just a wonderful episode.

And I think that Janice and I learned a lot, but so have our listeners. So thank you. Thank you. Thanks. Thanks for having me. Thanks for joining Tech Arena. Subscribe and engage at our website, Techorina.a. All content is copyright by Tech Arena.