Podcast with Wenmiao Yu — Quantum Dice
My guest today is Wenmiao Yu, co-founder and Director of Business Development at Quantum Dice. We spoke about quantum random number generators: why they are needed, how they work, and much more.
THE FULL TRANSCRIPT IS BELOW
Yuval: Hello Wenmiao, and thanks for joining me today.
Wenmiao: Thanks for having me, Yuval.
Yuval: So who are you, and what do you do?
Wenmiao: Yeah, so I’m one of the co-founders and the Director of Business Development for Quantum Dice. We are a relatively new spinout from the University of Oxford, and we are developing a quantum random number generator for cybersecurity.
Yuval: How is a quantum random number generator deployed? Is it a chip? Is it a service? Is it something else?
Wenmiao: That’s a really good question. Right now there are multiple ways for quantum random numbers to be deployed. You can have random numbers as a service, or you can have a physical device implemented inside, for example, your computer that generates random numbers. But what we are focusing on at Quantum Dice is making a hardware device that’s a quantum random number generator. And we are making several different form factors of it in order to service some of our client demands.
So for example, right now we have a shoe-box sized quantum random number generator. And this is what we are giving to, for example, telecommunication operators to use as an external source of entropy within their security systems. But at the same time, we’re also for example working on a miniaturized and on-chip version of our quantum random number generator. This will be embedded inside IoT devices, ranging from smartphones all the way to small satellites, to act as the on-device key generator for cybersecurity protocols.
Yuval: So why is a random number generator required? And I’m sure there are random number generators today, so what’s different about quantum random number generators?
Wenmiao: Oh, that’s a really good question. So if I start from the beginning, right now there are, I’d say, three main types of random number generators. So historically, people have used pseudo random number generators, and these are based from computer algorithms. And as the name would suggest, they are only pseudo random, which means that, for example, in high security applications, they are vulnerable to bias, and brute force attacks.
And then after that, people have looked into using classical processes to generate random numbers. So for example you can also have physical random number generated devices which are based from, for example, radioactive decay. And it’s from that random process that you collect the random numbers from. However, the limitations of the classical hardware based random number generators is that they are first, limited by the generation rate of true random numbers. And also because they are physical devices, they are vulnerable to direct attacks on the hardware itself, or for example, to physical degradation of the device as time goes on.
And hence for the past, I’d say about 10 or 20 years, academics have been looking into using quantum mechanics, which is itself an inherently random process, to generate true random numbers. And I’d say about 20 years ago, that slowly began to move its way from university labs, and slowly into more commercial applications.
And how most of the current quantum random number generator works, is that they are based from a quantum process. So for example, a laser which then hits a beam splitter, and when a laser hits a beam splitter, the photon could either pass straight through, or be reflected upwards. And that’s a 50:50 chance, and that chance is where the true random numbers come from.
And this is also what Quantum Dice’s QRNG is based from. But what we have is a step forwards to that, which is a patent that’s called Source Device Independent Self-Certification, and this was developed at the University of Oxford. And what it means is that we have a protocol that we call DISC TM, and it continuously monitors the physical device, and is able to both detect in real time any changes that’s occurred on the physical device, but also to adjust in real time to the changes that has occurred to the physical device. This means that our users can always be guaranteed of the minimum level of true random numbers that’s only coming from the quantum process itself.
Yuval: You mentioned the generation rate of the quantum numbers. How fast do you need to generate them, and why? I mean, isn’t it just, oh, I’ll generate a random number once a second, or once a minute?
Wenmiao: That would really depend on what the user wants to use the random number generator for. So for example, as you said, if I was at home on a Friday evening playing a game of chance with my friends, then we would only need to generate one random number, I’d say, per 10 or 20 seconds to the game that we’re playing. But for example, in higher security applications, for example within telecommunication networks where there are huge amounts of data that needs to be encrypted at any given time, the rate of the generation of the encryption keys that comes from the random number generator needs to be able to meet that demand. And right now in our labs, we can generate up to 8.05 Gbps of post-processed and truly quantum random numbers from our device.
And in one of the applications that we are also looking at, which is to use our quantum random number generator within satellite quantum key distribution, which is a form of quantum secure communications, they need at least 400 Mbps of random numbers to come from the QRNG.
Yuval: Does the random number need to be known on the other side as well? Or is it just one sided?
Wenmiao: That depends on the encryption protocol which you use. But where we stand at Quantum Dice is we are focused on becoming a leading supplier of hardware components for optical computing, and our first product is this quantum random number generator which can be used in both classical encryption protocols, such as the RSA protocol, but it can also be used in quantum key distribution. The exact protocol will then determine whether or not the key is known.
Yuval: Where do you see the initial users for your technology?
Wenmiao: The initial users of our technology, which is the quantum random number generator, is the entropy source for the encryption keys that are used within any sort of encryption protocols. And almost all encryption protocols used nowadays, that’s both classical and quantum, need a source of random numbers to generate these encryption keys. So that’s one of the earliest users.
And in terms of the sectors that need this ultra-high level of security offered by a QRNG within the encryption system, we have seen demand from the financial, as well as the government sectors. But what’s even more interesting is that people have also been looking at using the higher quality of random numbers from a quantum random number generator within their simulations. So for example, there have been several research groups and several papers where they have used the random numbers from a quantum random number generator within their Monte Carlo simulations. And they have actually been able to show that, due to a higher quality of the randomness, they were able to reduce the convergence time for these Monte Carlo simulations. So simulations is one of the areas that we’re seeing a lot of interest in QRNGs at the moment.
Yuval: But for simulations, I don’t need that many random numbers. I mean, I could imagine buying one of your chips, generating eight giga numbers, and then there’ll be enough for 15 years, right, for simulations?
Wenmiao: Mm-hmm (affirmative), that’s true. That’s true in terms of the rate of random numbers that’s needed. But the key benefit of our QRNG within simulations is this higher quality of randomness, which they then use to seed the simulation protocols. And it is purely due to the fact that quantum mechanics is inherently a truly random, and a truly statistical process, that we can leverage this higher quality of random numbers. Which then is able to, in some cases, speed up the simulations.
Yuval: I’m not a random number generation expert, as you could have realized by now. But I think there are other companies that do quantum random number generation. Is your uniqueness in the ability to detect changes in the source, or is it something else?
Wenmiao: Yeah, for sure. So I’d say there’s quite a number of companies that’s developing quantum random number generators. But where our unique selling point lies is in our source device independent self-certification protocol. And this is our patent, and this self-certification process means that we are able to guarantee that the random numbers coming from our device is only coming from a quantum process.
And if I just take a step back and explain a bit about why this is important, both for our users, and also for the end users of the encryption protocols. So any physical implementation of a quantum process will introduce unwanted classical noise into the system. And what I mean by this is, for example, we use a laser as our quantum source, but within our QRNG, this laser has to sit within other electronic components. And it’s these extra components making up the QRNG device that adds unwanted classical noise. And what, so far, no one has been able to do, but what we are able to do due to our self-certification protocol, is to effectively filter out the unwanted electronic noise, or the unwanted classical noise, before generating the random numbers, which means that the random numbers coming out of our quantum random number generator’s only coming from a quantum process. And hence we can have a certain level of guarantee of the quality of the random numbers which have been generated.
Yuval: So this is inherent to the random number generation process. It’s not a post-processing algorithm that could be used with other sources of QRNG, is that correct?
Wenmiao: Yeah. So that’s something we’re looking into. But for us, it’s both in terms of how our physical QRNG architecture is set up, as well as the post-processing that happens afterwards.
Yuval: How soon before it becomes available?
Wenmiao: It’s available right now, actually. So right now we have the shoe-box sized quantum random number generator, and this is what people have been using for technology trials or for proof of concept style development projects. And we envisage having a smaller kind of USB-sized quantum random number generator to become available towards the end of this year.
Yuval: Tell me a little bit about the company. How large are you? When did you get started? How did the company get started? How are you funded? Anything that you’re willing to share.
Wenmiao: Oh, gosh, more than happy to. So I met my four other co-founders back in July, 2019. And we all met through Oxford University’s Student Entrepreneurs program, and this was the first time that our university ran this program. This was a program that was hosted by Oxford Sciences Enterprises which is our local venture capital firm, as well as Oxford University Innovation which is the technology transfer department in Oxford.
And their idea was to, over a period of four weeks, give groups of students access to patents that was owned by the university, to slowly introduce them to how to do market research, how to fundraise, how to make a business plan, and how to do a pitch. And at the end of the four weeks we had the opportunity to take part in a demo day, where we pitched our business plan for commercializing the self-certifying quantum random number generator, to both directors from Oxford University Innovation, as well as some of the venture capitalists from Oxford Sciences Enterprises.
And we were the winning team of the initial StEP program, which meant that we won some initial funding that allowed us to put into production our first batch of quantum random number generator prototypes. And, almost by luck, during that StEP program, we also met with some of the center managers of the Quantum Technology Enterprise center based in Bristol, which is the UK’s only early stage hardware quantum technology incubator.
And they told us about this fellowship program that they had, and myself and one of my other co-founders applied for and were awarded the Enterprise Fellowships, which meant that we spent just over a year in Bristol, further developing the business plan for Quantum Dice, doing market research, and at the same time, of course, working very closely with the rest of our co-founding team in developing our initial pitch.
Where we are right now: we closed our first round of venture capital fundraise in the middle of last year, and we moved into our offices and lab space at the Oxford Center for Innovation, which is based in central Oxford. We have been very busy onboarding new employees, so mainly photonics engineers, electronics engineers, and we’re currently actively looking for a cryptographer. And we have all moved in and started to work on developing our own product, but as well as some of the other Innovate UK funded projects that we have with some partners.
Yuval: And your background, what did you study at Oxford?
Wenmiao: Yeah, so my background was in chemistry. I did my undergraduate, and my integrated masters at Balliol College in Oxford. And for me, I always knew that I wanted to work with technology but I also knew that I wouldn’t necessarily want to spend my life in a research lab developing those technologies. And through my time at university, I was always on the lookout for opportunities where I would be able to experience entrepreneurship firsthand, which meant that as soon as I heard about the StEP program that our university was running, I knew that I had to take part.
Yuval: Did you put your degree on pause, or finish it? Or are you finishing it as we go?
Wenmiao: Oh, well, so the timing worked out really well, actually. I started the StEP program in the summer of my last year at university. And in fact, I remember that on the first day of the program, which was the 1st of July, 2019, I actually also did my viva for my master’s thesis. And I remember going to the first morning of that program, and then hastily cycling over to the chemistry department to do my viva. So I think the timing worked really well. I graduated in August, 2019, and then moved to Bristol in November, 2019, to take part in the Quantum Technology Enterprise Center Fellowship.
Yuval: As we get closer to the end of our conversation today, I’m wondering if there are particular geographies that have a greater interest in quantum random number generator. For instance, is it more in the US, is it more in the far east, is it more in the UK? Do you find any particular variance between continents, or geography, in the interest?
Wenmiao: That’s a really good question, Yuval. Okay, so if I maybe answer this in two ways. So first, if we look at interest in terms of developing, and improving upon, quantum random number generator technology, I think this is a worldwide effort. But having said that, one of the oldest quantum random number generator firms was actually a Swiss company, that moved to South Korea a few years ago. And since they moved to South Korea, we’ve seen, for example, Samsung smartphones called Samsung Galaxy A Quantum, and each of these smartphones have, in fact, a quantum random number generator chip to provide on device security.
So we’re seeing some of the movements in terms of QRNG applied into commercial IoT devices in the far east, but we’re also seeing applications here in the UK, in Europe, as well as in the US, where QRNGs are being applied to some of the more traditional security modules (HSMs), which are, for example, used to secure data centers or telco networks.
Yuval: And the chip itself, is it fairly easy to manufacture, or there are only a couple of places in the world that might be capable of manufacturing the device?
Wenmiao: In terms of the QRNG chip, I’d say that generally, it’s still very much in the early development stage. However, when comes to manufacturing that volume in a few years’ time, I think the foundries that we would need to go to will either be in the US, or in Asia.
Yuval: How could people get in touch with you to learn more about your work?
Wenmiao: Well, we’re always happy to hear from people interested in quantum random number generation. You can reach us at Quantum Dice on LinkedIn, or you can speak to me directly, if you email firstname.lastname@example.org.
Yuval: And last, any random number joke that you want to share with us?
Wenmiao: Oh, sadly not. But I have to tell you that the amount of Schrödinger’s jokes I’ve heard over the past two and a half years have left me with a very strong repertoire.
Yuval: Tell me about it. Well, Wenmiao, thank you so much for joining me today.
Wenmiao: Well, thank you very much, Yuval. It’s a pleasure to speak with you.
About “The Qubit Guy’s Podcast”
Hosted by The Qubit Guy (Yuval Boger, our Chief Marketing Officer), the podcast hosts thought leaders in quantum computing to discuss business and technical questions that impact the quantum computing ecosystem. Our guests provide interesting insights about quantum computer software and algorithm, quantum computer hardware, key applications for quantum computing, market studies of the quantum industry and more.
If you would like to suggest a guest for the podcast, please contact us.
Originally published at https://www.classiq.io.