Rene – Hi! Welcome to QuBites, your bite-sized pieces of quantum computing. My name is Rene from Valorem reply and today we're going to talk about Quantum random numbers and for this I'm honored to have a special expert guest today, not a random guest of course, Wenmiao Yu. Welcome to the show, how are you today?
Wenmiao - I'm alright, thank you. Thank you for having me today.
Rene – Awesome, well can you tell us a little bit about yourself and your background as it relates to a quantum and physics and also of course related to topics, business, and so on.
Wenmiao - Oh well interestingly my path into physics and Quantum was quite non-linear. I studied chemistry for my undergraduate, my Master's at Oxford and actually only came into Quantum through a student startup program that I did in my last term. And that's where I met the rest of my Quantum Dice co-founders and that's where I first began to learn about Quantum random number generators and then afterward of course more about the wider Quantum technologies industry here in the UK and now globally as well. A bit about myself. I'm a co-founder and director of Business Development at Quantum Dice, I focus on how we commercialize our Quantum random number generator and in terms of the quantum ecosystem here in the UK I'm very familiar with programs and there's incubators such as V Quantum technology enterprise in Bristol where a lot of the UK-based quantum startups have come out of.
Rene - Got it, got it. Well, pretty impressive. So from chemistry to quantum chemistry and you know chemistry of course has a lot of relation to Quantum these days and but let's dive into our today's topic which is about random numbers but before we talk about Quantum random numbers can you explain for our audience why we actually need random numbers in computation at all and what is the gold standard these days?
Wenmiao - Of course. So right now, there are two main application areas where random numbers are needed. One of course is in security so any sort of encryption protocols or encryption algorithms that secure our data while our communication systems depend on our source of random numbers for the encryption keys and of course for other application area of quantum numbers is using them as seeds into Monte Carlo simulations for various sorts of simulation. So for example drug modeling or biochemistry modeling where we see the most frequent use of random numbers of a gold standard as you mentioned earlier you know. It really depends on which application area. So for example in security there are tests such as the NIST suite of tests developed in the US or with BSI AIS 31 tests herein Europe and these are really designed to test for the quality of the randomness being generated from any sort of random number generator and these can either be pseudo random number generators from computer algorithms or they can also be the hardware-based classical random number generators. So for example I'm sure a lot of people saw on Twitter or else online Cloudflare lava lamp random number generators.
Rene – Yeah, that's a nice one. Right, they have a whole wall of lava lens or at least one I don't know and then the random movement of nature basically, right, or like not 100 random but those thermodynamics involved with the lava lamp right but of course, there's like random stuff because we you know our world with quantum physics we have random things which lead us to the next question what is actually so special then about Quantum random numbers and why they are better than pseudo random numbers and what they cannot fulfill basically to the random numbers why would you choose Quantum random numbers? So I think I'm probably going to unpick that question a bit because the types of random numbers that you choose, would really depend on where you want to use them. So, for example for me personally if I think about security etc. I know that my requirements for long-term data security wouldn't be as stringent as say for example a government or a metric so the types of random numbers that would that I would personally feel quite happy with is using, you know, classical based random number generators and kind of going back even a bit further they are really two main classes of random number generators. They are these pseudo-random number generators which we talked about briefly and these are based on computer algorithms but of course, they as the name will suggest they are pseudorandom and hence they are vulnerable to buyers and brute force attacks and the other class of random numbers are hardware-based random numbers and these for example could be based on radioactive decay or very curiosity based on thermal situations but the main drawback of the classical hardware-based random number generators is that they are based on deterministic processes and also because they are physical devices the hardware is actually vulnerable to degradation over time or to attack summer hardware which would lower the quality of the randomness being generated and hence being used as encryption keys I think. One example that has been created is actually the 2013 smart card version Taiwan where a floor in the hardware random number generated that was used actually compromised around 10 000 of Citizen smart cars that actually managed to somehow pass government certifications. So this is a really interesting aspect because a lot of the times attacks on the hardware-based random number generator stores would go unnoticed and hence fair is a silent failure problem that then would feed into whatever encryption protocol it is being used in.
Rene - Well that's a good point like really I mean if you can kind of guess the random number of how random is it right and so Quantum random numbers or the real deal yeah I mean quantum physics based. So we have random events happening non-deterministic and well we can we can use this and so your company produces one of these random Quantum random number generators and so how does it work and how can it be integrated actually in an existing it and cryptography landscape because we also got a, you know, integrate the system into existing infrastructure.
Wenmiao – yeah, that's a great question. So of course we are a company that makes Quantum random number generators and we're not alone in that and in fact I think companies began as soon as early as 2001 to make random number generators based on Quantum processes and generally the difference between any Quantum random number generator and classical hardware-based random number generator is that is the core of the process from which the random numbers are being generated and of course within Quantum random number generators. It's based on the fundamentally probabilistic nature of quantum mechanics which means that whatever is generated should be theoretically true random numbers. But of course, as with any translation of theoretical concepts into a natural physical products that is used in day-to-day life there are less challenges you know it's very I think sometimes of course the theory takes a long time to develop but then actually turning that theory and engineering it into a product that can operate indifferent environments and work in actual operational systems but as a challenge and what we do at Quantum days is of course we use quantum mechanics to generate for random numbers but we also have a patent for a protocol that is called source device independent self-certification it's quite a mouthful but what it means for us is that it effectively we have a protocol that continuously monitors the physical device and is able to both detect for any changes to a physical device as well as adjust for output random numbers two phase changes in real time. Which means that for example if there was a change in temperature around the device or it diminishes third party decided to shoot microwaves added this Pro Tool will be able to detect those changes in the physical environment, adjust for it so that we can ensure that whatever leaves our random number generator over black box is actually verified and truly random numbers are only coming from a Quantum process and kind of linking that a bit more into day-to-day life. Of course our Quantum random number generators with many others are made using physical components, our Quantum process sits within a classical or an electronics architecture and all of these classical environments would introduce classical noise and what this protocol does is to effectively filter out the unwanted classical noise that can pollute and reduce the quality of the random numbers to ensure that whatever is generated from our box and being used by our customers is encryption keys are truly Quantum random numbers and in terms of integration we're working on a variety of different form factors and the reason for that is because really you know any sort of encryption system needs random numbers. So for example right now we have a discrete component device that we call Apex one and this is in the form factor of a 19-inch track mount server and this was specifically designed for users external source of entropy within classical systems .So for example within hardware security modules which are then used to secure government and data centers or networks and at the same time over the past few years we have also been developing a smaller more compact version of the quantum random number generated and this was designed for use in space. So it's been designed for use with being small sets of cubesats for two actors for sorts of encryption keys within satellite quantum key distribution systems.
Rene - Oh very nice. So now we're talking quantum key distribution that's actually interesting. I haven't thought about like this aspect of your solution but surely, of course, right you need random numbers there as well and so it's a more like space proof package, I would say.
Wenmiao - Right clearly I think classical one is a quantum security system developers is that they all really have their individual package of priorities that we need from the key generator. So for example in the classical world, it really is with self-satisfaction protocol that sets us apart because this allows a real-time health check of the source of encryption keys but it's also able to react to any changes so it's kind of kind of increases and also increases the assurance that their customers can have in their security system and then looking onto the space sector for example and especially within satellite-based quantum key distribution for them they need to have a quantum random number generator because we need that true source of random numbers but they also need the physical product to be small enough to be lightweight enough and to have a low enough power consumption so that it can fit inside small such as cubesat without having to compromise on the actual generation rate. So one example is within saturated ID for the bb-84 protocol is often used and for that they need at least 400 megabits per second of post-process random numbers and that currently is 100 times more than what the closest market offering of a qng with that form factor can provide. so there are ways to go around it of course you can buy multiple slower Quantum random number generators but of course these would take up more space on your satellite and the other option is to actually create a buffer so you could generate random numbers, store it and then deploy them once they're needed. But the introduction of this buffer actually also introduces an additional attack Vector which is something that you really want to minimize as much as possible.
Rene - And it makes a lot of sense and there's a big trend right like for quantity distribution also with satellites like China has I think in a satellite operational. I was just talking with Oscar Diaz and the previous episode of QuBites, he's from the European commission, he also mentioned of course that the European Commission the ASA, and so on are active show the NASA also and it all makes sense because you have less atmospheric scattering if you have a direct connection with the satellite instead of trying to transmit it wirelessly on Earth where you have a lot of you know interference from well the nature basically so yeah it makes a lot of sense and yeah you need two good random numbers you need them in a small form factor like you were saying low power consumption of course and yeah proved to you know the I wouldn't say like you know the vacuum out there basically so that it can sustain there, awesome. Well we're already at the end of the show, I'm sure we could talk much more about satellites and also the fantastic thing about Quantum random numbers in the nature behind it but let's keep it there, thank you so much for joining us today and sharing your insights this is very much appreciated.
Wenmiao – No, thank you for having me. It's been a pleasure to talk about Quantum random numbers with you.
Rene – Awesome, well and thanks everyone for joining us for yet another episode of QuBites, your bite-sized pieces of quantum computing. Watch our blog and follow our social media channels to hear all about the next episodes and of course subscribe to our YouTube channel and you can always visit the website to watch all episodes from season one to season six now until then take care and see you soon bye bye!