Dr. Peter Leek, OQC | theCUBE + NYSE Wired: Mixture of Experts

[00:00:00]>>... Palo Alto studio connecting Silicon Valley and Wall Street.

[00:00:03]I'm John Furrier <inaudible> Dave Vellante, my co-host.

[00:00:17]>> Welcome back to theCUBE studio here at the New York Stock Exchange.

[00:00:19]I'm Gemma Allen with NYSE Wired: Mixture of Experts.

[00:00:23]And joining me now from Oxford in the UK is Dr. Peter Leek, co-founder and chief scientific officer at OQC.

[00:00:30]Welcome, Peter.

[00:00:32]>> Thank you. - So OQC, Oxford Quantum Circuits, maybe just start by explaining to the audience what is it exactly that you do?

[00:00:43]>> Sure, yeah. So we build quantum computers that we've known for a long time that it's possible scientifically.

[00:00:50]It's incredibly difficult to build actual real machines that do this, but the technology is getting better and better.

[00:00:58]And when it gets to a certain level of advancement, it's going to be able to do computing that you can't currently do with the technology we currently have.

[00:01:12]What it really is, is that you're trying to do computing with systems which perfectly obey quantum mechanics at all times during the calculation.

[00:01:23]And that's actually a very, very difficult thing to engineer because you need to engineer perfection in the technology as it runs a calculation.

[00:01:34]So it's very difficult to do that.

[00:01:36]It's all about controlling errors and correcting errors in a way that quantum mechanics is obeyed during a calculation.

[00:01:49]The getting cold part, actually, so this is something which I think is... I think people hear this a lot, that certain types of quantum computing technology require low temperatures.

[00:02:00]That is true, but actually the technology we have to reach these low temperatures has been around for the best part of a century.

[00:02:06]And actually that isn't actually the difficult bit.

[00:02:09]So we've been able to get to these very low temperatures for a very long time, and the technology needed for that actually is very well-developed.

[00:02:16]The difficult bit is rather getting these very complex devices to behave precisely how you need them to behave to get to the end of a difficult enough or useful enough calculation to get the answer out.

[00:02:32]>> So in terms of OQC, you have created a business model that you've brought to the market around this, which again, correct me if I'm wrong here, but as I understand and read the brief, it's almost like quantum as a service, right?

[00:02:45]>> Mm-hmm.

[00:02:47]>> Maybe talk me through how that works in practice.

[00:02:50]You partner with different cloud hosting providers.

[00:02:57]How is it commercialized from you to them to customers?

[00:03:01]And I'm also very curious to understand what sort of use cases are you seeing?

[00:03:08]>> Sure, yeah. So maybe to start with, quantum computing I think is best thought of as a type of supercomputing.

[00:03:16]So it's a type of computing that you would use for particular applications that are very difficult things to calculate.

[00:03:24]And the other aspect is that some of the use or a number of the use cases are to do with computing using data, which can often be sensitive data.

[00:03:36]So the business model that we have is that we build machines and we deploy them to data centers where such data exists so that a company that has that kind of data and wants to do one of these difficult types of calculation on that sensitive data can then do so with our quantum computer that is deployed there in the data center.

[00:03:57]And they would then buy time on the machine to run the calculation they need to run, and they'd be able to do so with lowest latency and ease of access to the technology because it's there deployed in the data center that they're already used to using for other types of computing.

[00:04:13]>> Can you talk me through the partnership strategy around that?

[00:04:17]I presume this is your technology co-located within an existing footprint.

[00:04:23]Are we talking about companies like Equinix?

[00:04:25]Who would we be seeing you guys partnering with to bring that to market in the space?

[00:04:31]>> Yeah, that's exactly right. So partnering with data center providers.

[00:04:35]So we would deploy our systems into the data center.

[00:04:38]So we would own the system.

[00:04:40]It's deployed there into the data center, and the customers of the data center then have access to that as part of the compute infrastructure in the data center.

[00:04:50]>> Okay. Again, it's very interesting. How is it priced?

[00:04:52]Is it priced per minute, per output?

[00:04:56]How do customers avail of this in a, again, I guess, cost- efficient way in a market and part of the technology stack that's always seen as so unachievable?

[00:05:07]>> Yeah, to be honest, I would say it's a nascent market right now.

[00:05:10]So as you mentioned earlier, people have been talking about it being five years away, maybe it's getting closer.

[00:05:17]We haven't yet lost this threshold where running computations on this technology gives a clear advantage versus a classical alternative method.

[00:05:28]So all the uses right now are exploratory, looking towards that point in time where the technology gets to a level of advancement where you can see that kind of clear advantage of running the application on the quantum computer versus a classical supercomputer, say.

[00:05:44]And I think when we get to that point, then the pricing is going to become more clear, but just basically it's time on the machine.

[00:05:55]So you pay for an amount of time that you run a calculation on the machine.

[00:06:01]>> Wow. Okay. Interesting.

[00:06:02]So this really has been, I guess somewhat, you've grown up in the world of academia, right?

[00:06:07]You're a doctor, you were at Oxford.

[00:06:09]I think you were at Cambridge as well. Am I correct?

[00:06:12]>> Yeah, I was at Cambridge for my studies. Yeah, that's right.

[00:06:14]>> Okay. Two competing voices there.

[00:06:16]You're at Oxford now, but again, there's a discoverability element to this, like a proof of concept element.

[00:06:23]In terms of the ongoing developments though at quantum broadly, and especially as it relates to Europe, this is a company that's very much headquartered in the UK.

[00:06:34]I know you've also raised a very large round.

[00:06:37]We've had another quantum entrepreneur on this show as well recently from Europe.

[00:06:42]So it seems though there's certainly a market and enthusiasm and a level of genius in Europe that hopefully can continue to have global impact.

[00:06:50]Talk about that, about being a British-based company and raising money and also the role of academia and science in technology when it's all speed, speed, speed, right?

[00:07:02]>> Oh, yeah. So that's a lot of things.

[00:07:04]So maybe just to say I am the scientist, so you're pushing me outside my comfort zone to answer questions about commercialization and pricing, but yeah, so I'm absolutely the science side of things.

[00:07:22]I mean, it's really incredible to be at this position of development of the company, having started out in science, founded the company nine years ago to be at the point where we're at now to be able to raise that $350 million Series C fundraise.

[00:07:41]It's really incredible how well- backed this is now as a developing technology.

[00:07:48]I think I definitely couldn't have foreseen that 10 years ago looking at starting out a company based on the science I was working on, but it is extremely exciting to be at this point where it's really, really moving towards really commercially useful technology.

[00:08:06]>> Peter, when we think about quantum broadly and we think about a world whereby it could become more mass available and scalable, what sort of use cases come to mind for you, I guess, immediately, right?

[00:08:19]I don't want to call them low-hanging fruit per se, but what areas of society right now do you think could avail of this tomorrow in a very positive way if it was there to be availed of?

[00:08:35]>> Yeah. So I think for me, this is one of the reasons why I've been excited about this my whole career is actually more the longer-term applications of this.

[00:08:42]So the fact that what the machines we're building are able to calculate things according fully to the laws of quantum mechanics, what that means is that you can now do calculations to simulate the behavior of quantum mechanics in the real world.

[00:09:04]And this is one of the things which is extremely difficult for us to do with conventional computing technology.

[00:09:09]So for example, it's extremely difficult for us to calculate the behavior of chemicals or calculate the behaviors of materials or predict the behaviors of chemicals when they interact with each other or the behaviors of materials that we might think might be a useful design for some application.

[00:09:30]So when we have this technology advanced to a certain level, we'll have technology to hand, which will enable us to confidently simulate the behavior of real-world materials, chemicals, and their interactions to a degree that it's precisely matching the real quantum mechanics that occurs in those chemicals and materials.

[00:09:54]And so at the moment we can only make... I mean, it's a course thing to say in a way, but we can only do this very roughly with classical computing.

[00:10:07]And the reason I say that's perhaps a bit of a course thing to say is, of course, there's extremely well- developed techniques, extreme levels of expertise, very impressive methods to calculate behaviors of materials and chemicals, but they have to make assumptions.

[00:10:25]They have to use computers which don't have quantum mechanics at their core.

[00:10:30]Once you build that in, you can really fully simulate what's going on in the real world, which is obeying the laws of quantum mechanics.

[00:10:36]>> Wow. - So maybe that sounds a bit sciencey, but the calculation of the behaviors of chemicals and materials underlies a lot of advancement of society technology.

[00:10:51]So I think that really will open things up a lot.

[00:10:53]>> When we think about the risks, we sometimes ask folks, especially in the AI research space, what keeps you up at night?

[00:10:59]And recently I had one interviewee telling me he worries a lot about quantum encryption, and he sent me a link to a video about the possibilities of a world where bad actors have some control over quantum encryption before good actors are able to keep a pace, right?

[00:11:17]What are your thoughts on that, on that larger looming risk that we, I guess, are potentially facing?

[00:11:24]Do you feel that that's real, that that's grounded in truth and potential, or do you think it's potentially, again, there's a lot of dystopian conspiracy theories circling right now?

[00:11:37]>> Yeah. So there's a kernel of truth to this in that it's been very well known for many decades that one of the things you'll be able to do with a quantum computer is run mathematical calculations which map to the breaking of encryption.

[00:11:54]So one of the reasons we use the methods that we currently use for encryption is precisely because it's very difficult to calculate the reverse, reverse the calculation so that you break the encryption.

[00:12:08]And basically it's about the fact that it's easy to multiply two numbers and get the answer.

[00:12:15]It's very difficult to take the answer and work out which two numbers were multiplied.

[00:12:19]It just turns out that a quantum computer can do that reverse calculation much more efficiently, but it will only be able to do that on numbers of the scale which are used for encryption some way into the future.

[00:12:32]So it's still a little way off to have a machine that will be able to run that kind of decryption algorithm.

[00:12:42]Nevertheless, it comes forward in time because those who are working on the compilation of that algorithm onto one of these quantum computers are continually working out ways to more efficiently run that calculation.

[00:12:57]So as people work out how to more efficiently run that calculation and those of us developing the hardware improve the performance of the hardware, at some point those two things are going to meet and we're going to then be able to break encryption that we broadly currently use.

[00:13:16]So I think there's a reality to it, but I personally don't consider this a big risk to the world because I think it's not going to happen overnight.

[00:13:26]And as we progress in time towards that point when we can run that decryption algorithm, the world has plenty of time to adapt to that new reality.

[00:13:38]So still a couple of years away, perhaps it's five years away.

[00:13:41]Some might argue it's 10 years away still.

[00:13:43]I think it will be surprising if it's beyond 10 years away.

[00:13:46]>> Wow. - But the jury's out a little bit exactly how many years it's going to be before we can do that.

[00:13:52]>> Well, luckily, I guess we have great minds like yourself and other academics and experts in the field trying to stay ahead of these risks.

[00:14:00]So Peter, last question, what's ahead for you and the team?

[00:14:05]I mean, you obviously just landed a very big round.

[00:14:07]I'm sure there's a bit of runway to spend that money, but I also imagine it gets spent quite fast in the world of quantum computing.

[00:14:14]So what's the focus now for the next year out, and what are the, I guess, next big goals?

[00:14:21]>> Yeah. So yeah, it's really exciting to have these resources, and I think it's actually very clear what we now do with the money we've raised, and it's really putting it to work on R&D to be able to build that machine, which crosses the threshold of being beyond the capabilities of classical computing.

[00:14:40]So it's a really exciting period to come now for the next few years.

[00:14:44]So we now have the resources to be able to do the R&D.

[00:14:49]So there's going to be some expansion of the team, the R&D capabilities, and some real focus on the aspects of R&D that we need to do work on in order to build that quantum computer, which in about three years time should be beyond the capabilities of classical computing and really begin to address problems that are of commercial value.

[00:15:11]>> Wow. Well, Peter, we'll certainly be watching and paying close attention to all of the developments in the space.

[00:15:17]Dr. Peter Leek, thank you so much for joining us on theCUBE NYSE Wired.

[00:15:21]>> My pleasure. - I'm Gemma Allen here at theCUBE studio at the NYSE.

[00:15:25]This is NYSE Wired's Mixture of Experts.

[00:15:28]Thanks for watching.