[00:00:03] Speaker A: Welcome to Fernway Insights, where prominent leaders and influencers shaping the industrial and industrial tech sector discuss topics that are critical for executives, boards and investors. Fernway Insights is brought to you by Fernway Group, a firm focused on working with industrial companies to make them unrivaled. Segment of one leaders to learn more about Fernway Group, please visit our
[email protected]. dot.
[00:00:38] Speaker B: Hi, this is Nick Santanam, CEO and president of Fernway Group. Welcome to our podcast series on disruption 2.0. Today we are delighted to host Professor John Martinez. Professor John Martinez is a professor at the University of California, Santa Barbara, where he moved in 2004. He now holds a Worcester chair in experimental physics at the university. In 2014, Professor Martinez and his team at UCSB went on to work at the Google Quantum AI Lab to build a quantum computer using superconducting qubit. In 2019, Professor Martinez and his team published a paper on nature with a title, Quantum supremacy, using a programmable superconducting processor, where they presented how they achieved quantum supremacy for the first time using a 53 qubit quantum computer. In 2020, Professor Martinez moved to Australia to join Silicon Quantum Computing SQC, a startup, and in 2021 he received the John Stuart Bell Prize for research on fundamental issues in the quantum mechanics and their applications. Professor Martinez received his P's B's in physics in 1980 and his PhD in physics from the University of California, Berkeley. Professor Martinez, welcome. I'm sure our audience will be delighted to hear from someone like you who's a pioneer in the field of quantum computing. So let's start with a basic question for the benefit of listeners that have perhaps heard about quantum computing but haven't been able to keep up with the latest developments. What's the current state of play? What is the reason for so much buzz and excitement about this topic in the past few months past year?
[00:02:22] Speaker C: Well, thank you for the invitation. And yeah, indeed, this is a very exciting time for the field of quantum computing. The ideas around this have been around for several decades now, and physicists and others have been working very hard to build up these systems. And then what's happened in the last few years is people have made the systems that work well enough and they're working big enough that they can start doing interesting computations. And in particular, the quantum supremacy experiment that you talked about was the first experiment with really good data showing that you could do a quantum calculation that far surpassed what could be done with a classical supercomputer. Now, there was a little bit of a mathematical problem, maybe not that useful yet, but given that people have now shown that it's powerful. Then everyone's excited about moving forward, and the next step is starting to do something useful with the quantum computer. And that's what all the buzz and excitement is about, is when will we get to a useful quantum computer?
[00:03:30] Speaker B: So, John, you got us to the next question. What kind of applications benefit from quantum versus classical, and what really value does quantum unlock?
[00:03:40] Speaker C: Yeah, so, first of all, a quantum computer can very well map to quantum physics problems. It's a matter of recompiling the quantum physics and translating into quantum mechanics in quantum computer. And obviously, that's a natural application for the quantum computer. And this was something that was actually brought up by Richard Feynman in the mid 1980s, that one should be able to do that. But it took many years to figure out a. So that's a natural application. You can imagine a lot of quantum materials and chemistry developments along this line. Just as an example, maybe you could figure out how to make batteries better by being able to compute the chemistry of it, instead of having to do many, many experiments or maybe coming up some new ideas with that. That's kind of a natural application. Actually, there's probably a pretty big market for that in developing and developing that in a wide variety of ways. And there's a second thread that people are working hard, is that to solve optimization problems or help with artificial intelligence problems? And this is more of a strict computational kind of problem. And people think that the quantum computer, with its power, might be able to explore this. Now, this is a little bit trickier, and it's not quite clear that it's going to work yet. Okay. They think it would work. Maybe it's going to work in a heuristic way, in the sense that you can't really prove that it's going to be better, but it might be able to be proved to be more practically better. And I would say there's lots of applications for that. This is a big business driver. Optimizing processes can save companies billions of dollars, of course. So there's a lot of effort working on that, too. And of course, this is a new field, and people are going to be thinking of new ideas and algorithms along the way.
[00:05:38] Speaker B: Sunjan, I can't resist myself comparing technology to the nineties. When the Internet came on, people said, why do you need the Internet? What do you do with it? And if you look 30 years later or 20 years later, the applications in the Internet today are things you and I would have probably, or at least in my case, would have never thought about. I agree with you, it's going to be really hard. But is there one application you would sort of say, boy, if I have quantum computing, it's just going to unlock value? Is it like to your point, is it truly in the process optimization side? Is it finding new material? Is it to break the Internet and create the web 3.0? Where would you place your chips?
[00:06:17] Speaker C: You mentioned encryption, and one of the applications is to factor large numbers into their primes. This is the basis of present day encryption. And people, of course, worry about whether a quantum computer can do that. The nice thing is that there's other ways to do Internet security other than this, and people are working on it. So my guess is that these other things, they have already been invented. They will be fortified, made better, tested faster than it will be possible to build up quantum computers. So I'm not too worried about that thing. But in terms of what's the first application, I think we just going to have to see with that. I think if someone can figure out a heuristic algorithm to do these optimization problems faster because it's heuristic, then it just may appear suddenly someone has a good idea and show that it works better. And in that sense, that might be one of the first applications. But we don't know. Maybe that's hard, and maybe we just have to work hard for a decade or so to get an error corrected quantum computer to work. And then the chemistry applications and materials applications will come. So we really don't know yet. But people are working very hard on it. Everyone understands this. Everyone wants to find that first good application. And any day I can open a journal or look online for an article, and someone might have done this. And that's what's very exciting about the field.
[00:07:54] Speaker B: So, John, then that brings me to the next question of what are all the building blocks to have a workable solution.
[00:08:00] Speaker C: Right?
[00:08:01] Speaker B: And again, I don't want to keep going back to the Internet analogy, but I keep thinking about it, because when the Internet came, people said, what do I do with it? And people realize there's nothing you can do till all the pieces came together to create the ecosystem. So maybe I'll start with part a of that question. To have a workable quantum solution. Right? What are the big building blocks? And what are the current status of where people are or where companies are progressing along those key technologies?
[00:08:27] Speaker C: I'm an experimentalist, a physicist, experimentalist. So I'm obviously thinking about this. And right now, there's a variety of technologies people are thinking about in terms of building block. I'll use a simple analogy in that we have building blocks of atoms, and the atoms can be put together and made into molecules. We already have complex quantum systems, but those quantum systems are basically just exist, and you can't necessarily manipulate them however you want at will. And that's what's, of course, interesting about a classical computer, is you can manipulate the program however you want and be extremely inventive. And that's what we want to do with a quantum computer. It's very hard to build these new quantum systems where you can piece them together and build it in a special way to make a computer. Now, one of the interesting things that I've talked about often is that the problems of atoms and molecules, that they're too small. You have to get in and you have to control each atom. How are you going to do that? That's very hard. So in some sense, what we're trying to do is build quantum systems that are big, so big enough that we can get in light or get in signals with a wire to control the particular quantum systems. And of course, it's very different than what's happened with classical computing, because there you started big, and then over time, it's getting smaller and smaller and smaller, for obvious reasons here, it's like these systems have gotten bigger, and then that's where you can be able to put them together and control them. And the people are working on that, and they're thinking about, it's not quite, it's also hard. It's not quite what's the best system, because it has to be big enough to control, but you want to have it dense. And I consider this now, right now, to be a big systems engineering problem. So for people who've been working on this for many decades, they have very different approaches, which is good. It's very diverse team. And then we just have to see what's the best way to put it together to build a system. And I'm just going to say we don't know that yet. The system engineering is still being worked out. I was always interested in superconducting qubits, and they're big, so it's easy to microfabricate them. You can make big arrays of them using processing, and then it's big enough that you can get wires in to control the signals. And basically you get everything to work at once, which is key. Other systems such as ions and electrical traps and semiconductor qubits and photons and other systems are getting worked out, and we just have to wait and see what happens as people build systems.
[00:11:24] Speaker B: So, John, but those are all on the hardware side. If I go back to the old classical computer, then you needed the operating system, then you needed the application, then you needed the service providers. Is there a similar analogy in quantum where you would say, hey, those pieces are there or not there, they're coming up. Where would you put that?
[00:11:42] Speaker C: Yeah, it's interesting to put it that way because, yeah, classically you have to develop all those things, and that's what's happening right now. We have companies who are building a quantum computers. We have companies who are specializing in the control system. For example, there are companies who are thinking about the final algorithms, the software. There are also people who are working in the lower parts of the software stack where you have to compile these algorithms and control it and doing the calibration. People are working on all the above, and it's just a matter of getting everything to work in the end. And of course, one of the problems is that we have to compete against classical computers and supercomputers, which are really, really good. And that means we have to make an amazing system and engineer it really well to get it to function properly. But fortunately, as we saw with the quantum supremacy experiment, you can do that at a modest scale that physicists can imagine. Of course, we have to make that bigger and make it better. But it's not like we have to go to some totally insane engineering to make something useful. This point.
[00:12:59] Speaker B: I mean, you mentioned this comment, John, there are a lot of companies working on these various technologies or various pieces maybe sort of stepping back and taking a commercial angle. I mean, this has been super hot in the last year. There have been a few companies which have gone public. There are a few companies talking about going public. There are a few companies which are being well funded. If you were an investor, where would you place your chips? What gets you excited? Which part of this value chain gets you excited? Where you'll say, I think there's going to be a lot of money to be made or a lot of money to be lost.
[00:13:28] Speaker C: Yeah, well, I don't want to give investment advice, okay.
[00:13:34] Speaker B: Not about the company, but about the space, right? Like within.
[00:13:36] Speaker C: Yeah, yeah, I understand. And I would say from my personal point of view, the most important thing is people have to get something to work and get it to work well and do something useful. The problem is with, you know, these startup companies and the like, there's a lot of high people talk about the promise and it's all true. And that's really great, but in the end, you really have to get something to work. And I think getting a quantum computer to work is way harder than people think. Now, you use the example of the Internet, and all these things had to get invented, but it was developing things that people already knew how to do. We're developing something really new here, and in my mind, it's not even 100% certain that we're going to be able to build this, never mind what's the best approach. So I would say it's hard for investors because you have to have some knowledge about the technology and maybe a little bit of the physics and, like, what would be a good system engineering? These are all very technical things, and of course you want to look at the founders and what experience do they have? And I'm going to say also, is the company pushing a lot of hype or are they pushing, like, real results? That's really important. And for one example, I think what would be really good for the field is to have better metrics, benchmark metrics on performance. Not just have one, but have a suite and have them really well thought out so that people know if these systems are making progress.
Sorry, but it's mostly like we need to understand the technology really well and what are the risks and whether people can really build these systems well enough to do what we want to do.
[00:15:27] Speaker B: Ill take a slant towards a little bit of a macro issue. When you presented at our conference in September, where we had amazing CEO's from semiconductor companies and high tech companies, you made the point like us has the risk of losing the edge in quantum computing, quantum supremacy. What would be your thoughts? Or what should the government do? Or what is the government policies needed for us to have the lead, maintain the lead, accelerate the lead?
[00:15:57] Speaker C: I think, for example, it's very good that government is investing in basic R and D, say at the universities, because we have to do workforce development and we have to get people trained in quantum and in talking, going to universities and talking to people. I think it's great that it's not just physics departments, but engineering departments are thinking about that and training students both in engineering and the basic physics of a quantum computer because you need a lot of talents to build this. I think that's going well. I think there's good government funding and the universities are doing the right thing. The bigger issue, or the second issue is I talked about the metrology and doing benchmarks, and I think people are beginning to work on that, and that's good too. But the big thing is to build a quantum computer is very complex. It takes a big team. And where is that going to be done in the United States? Now, fortunately, in the United States we have a very good startup culture where there's companies being put together to do this. We have relatively rich, let's say Internet or other electronics companies who can invest in this and build these teams. And certainly when I was at Google, I was very appreciative of that, and that's really good. But China, for example, is investing in a couple of technologies that look really good and they're putting big teams on it, and that's fine. In the end, it's going to come down to who has the money to do it. And did the government choose the right places to put the money in? And was the management of the groups good and did they really focus and work on the same problems and the important problems? And these are all kind of the things that's hard to, let's say, throw money at. Okay. It's good that there's a lot of different things that people are working on and different physics implementations, but we have to wait and see.
[00:18:04] Speaker B: But you don't think there's a specific need, like a national policy, like how China has done on quantum computing? Is there something from a policy point of view, the us government, or the US, or for that matter, any country can do to enable this industry to take off within that country?
[00:18:21] Speaker C: I don't understand. Well, okay, let me. I'll just tell you my own personal story. I know how to build a quantum computer. It's in my head.
I know all the components, I know what we have to do. And I'm sitting here essentially retired in my home thinking about it. Now I'm trying to work on things and I think the private sector may come to something, but I wish there was a little bit more funding for the people who really had good ideas on what to do. I mean, there's funding for big projects out there and that's great. But yeah, maybe funding for a few people who really have good ideas would be good in the US.
[00:19:05] Speaker B: John, now I'm going to switch over to your journey. I mean, you're very well known in this field. You've done a lot of work. Talk to us about your journey. I mean, how did you get started on this quantum supremacy project when you joined from UC Santa Barbara working with Google AI Labs? Walk us through the journey. What got you excited, and probably more important is if you could wave a magic wand, what would you be doing differently so that the audience can learn from your experience.
[00:19:34] Speaker C: Yeah. So I was a professor at UC Santa Barbara and I was being funding very well by the us government and really thank them for their funding and pushing forward the field, not just our group, but the whole field. And we learned from everyone. And then at the last few years at UC Santa Barbara, we had more or less started to begin to understand how to put together a quantum computer with superconductors. And we started that with the five qubit device and then moved on to a nine qubit device. What was quite amazing about that is that the devices work better than I thought it would. It's kind of the first time in my career, you know, experiments, you know, works better than you thought it would be. And, you know, maybe I was too pessimistic, but I also think we had been working hard for quite a few years to really understand. And then at that point, I was very excited about starting to scale up. We had the people in the group talking about that. The problem is our funding in the us government at the time, it wasn't nearly as big as what it was now, and that didn't seem like it would work out. For example, they really didn't want us to make more than ten qubits in a device. And for me, the whole exciting thing is scaling up and building a big system and figuring out what's wrong and fixing it. And I wasn't quite sure that was going to work out. And Hartman, Nevin and Google came along and suggested that we could transfer our project over to them and that there was funding at Google to do this, in that they wanted to build a useful quantum computer and I wanted to build a use of quantum computer. And in the end that worked out great. I brought myself and my team at UCSB into Google. And what was really great about Google is that we could really focus on all the engineering and technical challenges to bring this up. We basically followed the plan that we had come up with at UC Santa Barbara and started building it out. And then along the way, the theorists at Google understood the quantum supremacy experiment theoretically. And it sounded like a great milestone, very challenging milestone, but a great milestone. And we kind of knew that if we could show the power of a quantum computer, then more funding would be available and it would show there was something. So it was a very natural development into there. I was very happy that we were able to get to that milestone. Looking forward, I left Google and learned about semiconductor quantum bits. And that was really enjoyable working with Michelle Simmons and doing that. The personal reasons I came home to Santa Barbara. What I've been focusing on, what I think is the most important thing for the field is that's making qubits better. We understand from classical computing that making more bits is good, making a computer as larger is good, or faster, things like that. But my particular view right now is what's key is making the qubits better. And that's something that I've written essentially two papers on since leaving Google. I thought about that and then we're also thinking about how to do the fabrication. When I think all the qubits need to get better. This is one of the key things that's hindering the experiments right now. And that's what I'm focusing on. What I think is the key issue.
[00:23:04] Speaker B: One question I wanted to ask, I mean, you've talked about quantum supremacy. Obviously it was a great achievement. If you sort of look at that. How did that come along? Was that mainly because of better funding, better team, better resources? Because that was truly an amazing achievement in such a short period of time. And you obviously led it. It's been talked about, you have talked about it for the audience. Would be great to talk a little bit about how you drove that success.
[00:23:27] Speaker C: Yeah. What's interesting is when you look back to the UC Santa Barbara results, just when we were leaving, we were making good qubits and we were able to put it together. And in fact, in the supplement of the paper, we talked about how to do a 2d array. So back then we kind of, because the qubits were working right, we had a good idea that we could scale it up and we knew enough about what was going on that we thought it would work. When we built a 2d array, we just had to develop the technology up, bomb bonding and many other things, of course, but we felt we were on that road and that we could make progress. And then when the quantum supremacy idea came along, it was more or less doing what we were already wanting to do. We had to make the performance just a little bit better, not too much. It was something we were thinking about for, I don't know, five years at that point. So it was very natural to do that. I think people were surprised that we were able to get the whole system to work well. And in fact, that was the present surprise. Nothing horrible went wrong with that. But on the other hand, if we looked at the data that was coming out of the UC Santa Barbara devices, that was pretty good data too. And we knew we had to fix things and we fixed it. But it was seeming like it would work. I mean, that's what was exciting about it, is we had a good idea at Santa Barbara. The team had a good idea. They knew what to do. And then going to Google, we just had the support and kind of the security of Google funding us for many years just to go ahead and try it and do it. It just kind of naturally came together, although it was very hard. And everyone in the lab work was very inventive to get this to work, but it was just kind of the natural evolution of what we were doing.
[00:25:27] Speaker B: So outside of your work at Google and SQC, John, you talked briefly. What are the quantum efforts? Are you involved in? What are you working on?
[00:25:36] Speaker C: Well, I'm working on how to make the superconducting qubits better, like I talked to about before. And we have a specific plan for how to do, how to make them better and how to do the fabrication and testing in a way to do that. And, you know, I'm collaborating with someone at a university who's doing some of the initial tests, and he's going to talk to me about some data today that they took yesterday on and that we'll see if some of these ideas are working. And then in my spare time, I'm also thinking about how to form a startup company. And along these ideas, of course, the hard part is figuring out the business plan and to make it attractive to investors. And you still working on that. But we have some nice ideas there and we'll see if it happens. It doesn't work out. I have a lot of things still left to do, kind of in the academic realm and a lot of ideas, but we'll see if that comes out.
[00:26:42] Speaker B: You're keeping yourselves busy for sure. John, I know you have enough on your plate. Maybe I'll do a little bit of a switching of topic. John, at Fernway, we continue to see a big divide between technology providers and end customers in which companies don't know how to adopt and scale quantum technology. That's something we are very involved in. We talk to these companies. You got a bunch of great companies trying to offer quantum offerings, and there are companies which want to use it, but there seems to be a divide. And so there is a little bit of a wait and watch approach. There's a little bit of I have the money, but I don't know how to spend, I have a product I don't know how to sell. What's the right approach? If you are advising a company to adopt quantum technologies, whether it is a pharmaceutical company trying to develop a new material, whether it is a logistics company trying to develop a new optimization, what's the right approach? What can companies do today to get going?
[00:27:34] Speaker C: This has been a question that we I, and when we are at Google, we've been thinking about this for quite a few years. And I think what I'm going to say here is what people have a little bit adopting. And the problem here is for the companies, this is a very new technology. Quantum mechanics is a little bit mysterious. It's unclear what the algorithms are, but it potentially could make a big impact and be a real competitive advantage if you know about it and disadvantages, you don't. What has been recommended and people are doing to some degree is the various companies see that it's a future technology and they're investing by at least starting with a small, let's say, software or algorithm team so that they understand the applications and the like. And then one of the things that these teams do is they will then partner with one of the software companies and maybe some of the hardware companies to start learning how to run and think about these quantum algorithms. And mostly what you're trying to do is identify whether the areas of impact within your company and also have some expertise in the company. So if there's a big advance, then you kind of know what that means and then you can fairly quickly respond to that. But it does take some time to build up that expertise. And I think that's good that companies are doing that. And I think it's a good ecosystem where there are these quantum algorithm companies that are trying to partner in finding these clients and working with them. I think it's good, for example, that people are spending money on hardware, the hardware. Now, you can easily simulate that with the classical computer, but it's always good to know what the real situation is with the quantum computers. People are just learning right now, and that's appropriate thing to do.
[00:29:35] Speaker B: So, John, you talked about that there's a lot of hardware company startups which are coming up with a quantum solution. They're putting it up on the cloud. It's a little bit like a toy, right? You can go play with it. You don't know what to do with it, but you can play with it. Do you think that's the right model? Both, if you are a provider, right? I mean, to put it in the cloud, because obviously you get access to the markets, but probably you're not going to get that much more customers because anytime new products come in, new technologies come in, you want the inefficiency of usage.
[00:30:06] Speaker C: Right.
[00:30:06] Speaker B: Because people buy a car, they drive it for 30 minutes, they keep it for the remaining 23 hours in the garage. And that's good because more people buy cars. In a similar analogy, do you think this whole point of putting hardware on the cloud, which enables people to try, but do you think it's going to slow down innovation and growth?
[00:30:24] Speaker C: Yeah, that's a really good question. I have talked about this for very few years now that putting it on a cloud is much better than buying your own quantum computer with the idea. With a quantum computer, you add one qubit, you've just doubled its computation power, and how are you going to keep up with that? And it's better to have a cloud next to the research lab where it's getting worked on all the time and improved. That model makes sense. And in fact, most of our computing these days with the cloud, when you have your cell phone, that is a good model. And I think it's good that people practice running their algorithms on the cloud. But in some sense, I agree with you. The computers aren't that powerful right now, and quantum computers are not that powerful. So if you're using one, it's just kind of testing your algorithm, but you wonder exactly how deep it is. And then if you build a cloud service, then that could be distracting your team to build a big quantum computer. And there's that point. But what I would say is I actually would rec. I think having these cloud services are really important, because whenever you're building a system and you think about systems engineering and whatever, and the customer, you want the customer using your product because they know how to break it better than you do, and they know that. And then it forces you to make it reliable. It forces you to understand your system better, to do good system engineering. So I think the cloud service is really functioning as a test vehicle to see how reliable you can build all your quantum components and your quantum computer. And that's what I think is really good. And you need to understand that, because right now, let's say we're at 50 qubits. In the end, people want to build a million qubits in ten years. And unless you're getting real world test feed, you won't be able to get the reliability up to that scale. Now, of course, if you worried about that and worried about reliability, you have less people worrying on developing the technology. But test and reliability testing is a big part of systems engineering. So I think that's the right model in my view. It's something people are working out right now, and it's good. There's different points of view, too.
[00:32:57] Speaker B: In closing, I want to make a personal question, or a personal ask. John, in your Forbes interview last year, you called yourself a definite optimist, meaning that you determine that one best thing to do and then you go do it by the way of benefit for the audience. This type of personality is a rarity, according to Peter Thiel's book of zero to one.
Tell us, what does that mean? How do you do it?
[00:33:23] Speaker C: Well, one thing, because it's very unusual and radical. Many people don't understand you, including management or people you work with or whatever, and it's kind of hard to find the right place to succeed that way. One of my heroes is Elon Musk, who had getting enough money early on in his career to do exactly what he wanted to do. And that's really great for me. The way I view this is I've been thinking about how to build a quantum computer, and I've been really focusing on superconducting qubits. Obviously, I know about that, and I've been identifying what are the key things we have to do. And like I said before, we have to make our qubits better. And there's a variety of things we have to specifically do. And because of that, I wrote very detailed papers in the last two years kind of describing that, which I think to a lot of people I might say, well, that's already known. But for me, it really got me to focus in on the real problems, understand it super well. And now I want to dive in and to fix those problems very carefully. So that's the way I've always done things. And because of that, I have some really interesting ideas on where we go and hopefully things will work out and I'll be able to talk about that in the foreseeable future. And I'm really excited. That kind of this mindset, there's maybe a way forward for me to keep contributing to this field in a meaningful way.
[00:35:00] Speaker B: John, that's a great way to wrap the podcast. You've done great things, and it looks like the best is yet to come. So first of all, all the best for your next chapter. I'm sure we'll be reading great things. And with that, thank you.
[00:35:13] Speaker C: And thank you. I enjoyed it.
[00:35:15] Speaker B: Thanks, John.
[00:35:22] Speaker A: Thanks for listening to Fernway insights. Please visit fernway.com for more podcasts, publications and events on developments shaping the industrial and industrial tech sector.
