Can a Quantum Computer Fit on a Chip?
Jason Lynch uses paracetamol to show what classical computers still cannot simulate. But there isn't a scientist alive that is able to say exactly the mechanisms.
In this episode · 13 chapters
We’re all educated classically, and then we’re trying to get our minds around this stuff. And maybe actually we need people who are quantum first and thinking about that, and then we’ll have quantum algorithms coming out.
Jason Lynch, CEO of Equal1, builds quantum chips on standard silicon, with the qubits, control electronics and classical processors on one chip. Because its spin qubits run at about one Kelvin, he says, the system fits in a rack in an ordinary data centre.
He explains why Equal1 trades some qubit quality for a better overall system, why every project starts with a customer's use case, and why he thinks quantum should narrow the gap between rich and poor countries.
Jason Lynch, CEO of Equal1, argues on the Thinking On Paper podcast that quantum computing will scale the way classical computing did: on silicon chips. Equal1 builds rack-mounted quantum computers that put spin qubits and their control electronics on the same chip and slot into an ordinary data centre. Lynch says chip economics can bring quantum to the masses, not only the richest companies.
Can a quantum computer fit on a chip?
Jason Lynch says it can. Equal1 puts spin qubits and their control electronics on the same silicon chip, in a rack-mounted machine that slots into an ordinary data centre.
Why do Equal1's spin qubits run warmer than superconducting qubits?
Equal1 trades some qubit quality for a better overall system. Lynch says its spin qubits target about one kelvin, much warmer than the 40 to 50 millikelvin superconducting qubits need, which leaves enough thermal budget to put the control electronics right next to the qubits.
Can Equal1's quantum computers run in an ordinary data centre?
Equal1's first machines are built to slot into an ordinary data centre. Lynch describes rack-mounted systems about a metre high that use the type of closed-cycle cryocooler found in MRI machines, and says four have run in Equal1's own data centres for the past two years.
Who should quantum computing benefit, according to Jason Lynch?
Quantum should narrow the gap between rich and poor, not widen it. Lynch calls the view that quantum will serve only the richest companies and countries “a dangerous philosophy”, and says chip economics can bring it to the masses.
Why does simulating molecules need quantum computers?
Simulating molecules is the prize, and classical computers cannot do it. Lynch says no scientist alive can explain exactly how paracetamol works because we cannot simulate it, and that the job needs machines in the thousand-qubit range, which he expects Equal1 to reach in two-plus years.
What does Jason Lynch think AI has got wrong on power?
Lynch believes AI has taken a wrong turn on power. He contrasts the megawatts used to train models with the 20 or so watts the brain uses, calls quantum AI a bet at this stage, and says Equal1 puts a neural network on its chip to detect and correct qubit errors in real time.
Where Mark and Lynch disagree
Mark suggests Equal1's approach makes IBM's pursuit of near-perfect qubits look “almost like redundant”. Lynch rejects it: “I wouldn't say redundant.” He expects the market to split by application, with encryption work such as Shor's algorithm needing higher-quality qubits, and room for several winners.
Links from this episode
Questions
Can a quantum computer fit on a chip?
Jason Lynch says it can. Equal1 puts spin qubits and their control electronics on the same silicon chip, in a rack-mounted machine that slots into an ordinary data centre.
Who is Jason Lynch?
Jason Lynch, CEO of Equal1, builds quantum chips on standard silicon, with the qubits, control electronics and classical processors on one chip.
What does Equal1 build?
Equal1 builds rack-mounted quantum computers that put spin qubits and their control electronics on the same chip and slot into an ordinary data centre.
How does Jason Lynch think quantum computing will scale?
Jason Lynch, CEO of Equal1, argues on the Thinking On Paper podcast that quantum computing will scale the way classical computing did: on silicon chips.
Transcript
Jeremy Gilbertson Disruptors and curious minds. Welcome to another episode of thinking on paper. My name's Jeremy. This is Mark.
Mark Fielding We get to talk to the people that are designing the future of the world. Things are going to change guys.
Jeremy Gilbertson Newsflash. A lot of things are changing. Technology drives changing and we explore and unpack what that means to us. What that means to culture, families, all of that fun stuff. Um, so today I want to, I want to just kick off a few ideas, Mark, for, for this conversation. That's quantum season.
Mark Fielding I'm, I'm so excited about quantum season. We had our first episode last week. We're, we're continuing the conversation this week.
Jeremy Gilbertson And as a reminder, you know, quantum mechanics, quantum physics is counter to how we experience the world. Just a reminder for those listening in case you missed yesterday, we're going to kind of tease this in. And, and it's not really supposed to make sense. Like Einstein called it spooky, spooky action. It freaked him out. Like it, it was like, he didn't know what to do with it. And, you know, there were things about talking about things moving faster than the speed of light that he was like, no, this can't be possible. They ended up figuring out quantum mechanics is real equations. Prove it all of that fun stuff. But it's really interesting to think about quantum, like the small pieces of the, uh, that operate the world.
Jeremy Gilbertson That operates more like nature, like rather than this controlled thing, like when humans get control of stuff, we like to organize it. Right. So like traditional computing is organized into ones and zeros, right? Quantum computing does things a little bit different, which we'll explore, but it's also important to have that bridge to how we do things, which Jason's going to get into today. How we're kind of merging classical compute and quantum compute, which is super exciting.
Mark Fielding Um, well, that's one of the things we've kind of learned already in quantum season that there is no one quantum computing, there's different facets to this. And I think today, and I might be wrong, but I think we're going to be exploring that kind of zone where classical computing and quantum computing merge, and we're talking about the scale, scalability, scaling, quantum computing, that useful quantum computing that everybody can access. And I, yeah, I think that involves maybe that, that point where the two technologies merge. And, uh, PS, if you want to dive in a little bit deeper into quantum with us, a new book club book announcement, it's out Michio Kaku's new book, quantum supremacy.
Jeremy Gilbertson I'm so looking forward to reading this book. This he's been one of my favorite people to listen to. I saw him at a conference years ago and he talked about the, we'll have the ability in the future. We don't need spaceships. We need the ability to beam our consciousness between stations across the universe. And he had me at that. So I can't wait to jump in Mark. Uh, let's intro our guests and let's dive into this stuff.
Mark Fielding Yeah. Our guest today, Jason Lynch, he is the CEO of Equal1, and he is on a mission to bring quantum computing to the world. So welcome to the show, Jason. Pleasure. Thank you for being here. Gentleman delighted to be here. Thanks.
Jeremy Gilbertson Thank you. Yeah, we're, we're excited. So, um, Jason, as you know, we, we like to string these episodes together. We, we talked to, uh, we had another quantum guest on last week that, uh, that left a question for you, Mark, do you want to kick off with that question and we'll, we'll dive right into that and then go into some other things.
Mark Fielding Yeah. Shout out to Joe Fitz Simons, who I think maybe your and his past have crossed. His question for you was, um, what are the applications of quantum computing we haven't thought of? Or what, what is going to be, um, the surprising outcome of this that nobody's speaking about?
A Feynman quote
Jason Lynch Yeah. So I think what comes to mind here is maybe, um, a quote by Richard Feynman. Right. So Richard Feynman, um, one of the legends of physics in general, but also quantum physics. And he did a lecture in 1959 that was titled there's plenty of room at the bottom, right? And the whole lecture was around nano electronics. And at that stage, you know, transistors were just coming out and I, you know, one, uh, hunch I have maybe that, uh, we could use quantum computers for that we haven't talked a whole lot about is the next level down, right? The quarks and these other things that, you know, we're, we're, uh, trying to put experiments in, in, uh, CERN and other places today to figure out, are they real and how does that world look like?
Jason Lynch And so if I think about the most powerful computers today, some of the things we're trying to do on them is figure out that next level down. And I think quantum computers can help us learn more about the world. That's below electrons and, and, and go to quarks and these other subatomic particles.
Jeremy Gilbertson Jason, you had me at Richard Feynman. He's one of my favorites. Like I, you know, surely you must be joking. Richard Feynman was like the rabbit hole book for me a long time ago. And, and that particular talk is, is brilliant. And the root of nanotech, uh, largely, um, I love that.
Jason Lynch And I think it applies now more than ever, right? I might, I think we have a tendency as a species to think we've most of this, uh, solved, right? I think we have only a tiny, tiny percentage of this solved and there are so many, uh, challenges, both, you know, uh, lower in scale and, uh, and so, yeah, I think this is a, this is a quantum computing is going to be a massive tool to help us understand more about our world. Um, and I think that's one of the reasons why it's so exciting. Yeah.
Mark Fielding Um, as if I have to tell the audience watching this, that, um, we haven't spoken about Richard Feynman before the show, so Jason has brought up Richard Feynman and me and Jeremy were just talking about Richard Feynman before the show, and we were actually planning to try something new on thinking on paper today, like a new segment, a new section called explain it like you're Richard Feynman, um, unbelievably, because going through the Equal1 website and doing the quantum season, there's a lot of terminology. There's a lot of vocabulary, which if you're not in this can seem daunting, scary, or just completely incomprehensible. Um, so I've taken a few of those words from the website and could you explain them like you're Richard Feynman, Jason, if I try my best, it's amazing.
Jeremy Gilbertson Mark, before we do that, let's, let's just, so listeners check this out. Like, so traditional like computing, right. You have one or zero, you know, that's it. It could be on or off one or zero. And, you know, the, the way kind of like the qubit thing is, or at least how I get my head around it is there's an infinite spectrum of identity between one and zero that quantum allows us to do. So this whole like infinite possibility of compute paths is at least, at least how I get my head around this thing, um, is, uh, is it just trying to set the stage for that, but let's go in the terminology. Okay.
Jason Lynch Am I still here? Cause I think my internet dropped out.
Quantum system on chip
Mark Fielding You're there. Okay, great. So quantum system on chip, QSOC.
Jason Lynch Yes. So, uh, basically our, our approach, I mean, Equal1 was formed, uh, basically with the thesis that, uh, all of computing has advanced with silicon and with CMOS and with the semiconductor industry that we have built. Um, and our belief is scalable quantum will take the same path. So, you know, time and time again, people questioned Moore's law, people questioned, could we do certain things in semiconductors? And we've come up with the answer every time that we can, if it's, you know, whether it's from 4g 5g now 6g, whether it's autonomous driving, all of this has advanced along with computing, um, along with the advancement of silicon. And so our sort of starting thesis is that, and our thesis is that we should build on the shoulders of, of what we have today and not start this, uh, start again.
Jason Lynch And so what quantum system on chip means, and anybody in the classical technology semiconductor industry will know what system on chip means. It means that you put as much on a chip as possible and you integrate, you know, as much as possible onto the chip to enable massive functionality. And we, again, think this same approach, uh, should be taken in quantum and is what is going to lead us to scalable quantum. So when we talk about quantum system on chip, we're referring to a chip that has the qubits, it has, uh, all of the control electronics that you need to talk to the qubits, but it also has, you know, the mo the latest and greatest cutting edge, uh, arm processors, um, you know, all a tensor processors so that, uh, right next to the qubits, we can do some of these algorithms that, that, that people are talking about. So, so that's quantum system on chip. Okay.
Jeremy Gilbertson Got it. Got to follow there. So, so you have, you have quantum and classical like living in the same space and they're two very different technologies. Does the classical side of that chip act as this translator and allow you to use existing algorithms in a more efficient way, rather than completing completely having to reinvent the algorithm side?
Jason Lynch Um, in, in a way, I mean, one way to think about it is if you look at these quantum algorithms and the ones we're particularly focused on, and we can get into sort of more on the applications, but they're chemistry and optimization. And typically what you need to do in those is you have your quantum, your qubit array, and you make some, um, Controls on that qubit array, and then you read it out and then you have to do quite a bit of classical processing. So whether it's error correction or whether you're processing to see, you know, have you gotten the, the optimum result, there is a significant amount of classical processing done in a quantum algorithm.
Jason Lynch And so our belief is that sort of not really discussed today, but when we go to massive, you know, if you look at these chemistry systems that we want to simulate, we need, you know, thousands and ultimately hundreds of thousands of qubits, there's going to be massive control needed for that. And so it's really about, you know, if you look in under the hood in these, uh, in these quantum algorithms, there is in a lot of them, quite a bit of classical that goes hand in hand. That makes sense.
Mark Fielding If you want to learn more about quantum software, watch last week's show.
Jeremy Gilbertson That's that's right. And, and Hey, curious minds out there. Listen, uh, so the idea, you've probably seen what these quantum computers look like, it's a giant chandelier, it's gold, it's huge and big, but like the compute part of that is the tiny little black box sitting at the very bottom of it. Everything else is to control the environment that makes the qubits operate and work. So let's go, let's go to other terms. Mark, what else did you have on the, on the Richard Feynman hot list?
Mark Fielding Well, some of these, um, might not be the right terminology, or you might have already said this, but, um, UnityQ qubits. And I guess, is that different to a, a qubit?
UnityQ
Jason Lynch Uh, well, so, um, UnityQ is really our, um, our brand around unifying classical and quantum, and it's really to try to drive that message home that, you know, qubits on their own. Uh, you need, we need to be able to talk to them. We need to be able to read them out. We need to be able to do, you know, classical processing around them. So UnityQ is really the brand, uh, of, um, of the system on chip. Uh, in terms of the qubits, I mean, there's maybe something to say about those and, and how we're approaching it. So, uh, you mentioned the chandelier. I mean, our, um, and the reason you have a chandelier like that is because the particular qubits are superconducting. So there's, let's say, you know, a handful of, um, technologies that are vying for, um, you know, the, the, a scalable qubit platform.
Jason Lynch Superconducting are extremely cold, 40 millikelvin. So point, you know, 0.04 degrees above absolute zero. They need to be cooled to, to operate, to operate. And the reason being again, you know, for some context, these qubits are extremely sensitive, right? So we don't see quantum effects at room temperature because of the thermal noise and all the other things that are going on. So when we, when we get these qubits into a quantum state, we need to make the environment extremely pure and we need to reduce the noise as much as possible. Superconducting qubits need to be cooled down to 40 or 50 millikelvin. Our qubits, which we call spin qubits actually can operate at about a thousand times warmer than that. So we're targeting ultimately about one Kelvin. Now that's still extremely cold.
Jason Lynch But what it does allow us to do is it enables the system on chip. So one of the challenges and one of the reasons you need the chandeliers are essentially the control electronics for the qubit because we can do them at one Kelvin, we have enough thermal budget on the chip to put them next to the qubit. And that again is sort of our, um, differentiator and, and, uh, and our disruption where, um, because we can operate these qubits somewhat warmer, we can, we have more thermal budget available to do more things to control them.
The silly card
Jeremy Gilbertson I'm going to have to play my, my silly card here and ask, help, help me, help me understand more how you can be a thousand, um, percentiles warmer, like how can you do this in such a warm, much more warmer environment than traditional quantum computing leads me to believe. Sure.
Jason Lynch Uh, so for that, I'm going to go a little bit to the team and our, and our experience and our background. So we're, we're a bit of a different team than the standard quantum teams. I mean, typically a quantum computing companies up to now are typically academic teams that come out from university and we have that, we have the academic prowess and we have, you know, some of the best scientists in the world, but we also have some of the best technology commercialization. Um, people in the world, people who have done the first smartphone chips, um, you know, people who have done, uh, AI accelerators that are on, um, you know, um, processors today for Intel. And what we've learned is that it's not really about perfecting one thing. It's about a system optimization.
Jason Lynch So from the very beginning, we're, we're about five and a half years old. And from the very beginning, we took an approach that said, okay, ultimately the game here is not to make the best qubit and, uh, you know, the best single qubit and, uh, you know, try to improve on that. The game here is, is to make the best system and optimize the system such that we can actually get a quantum computer built. And so to answer your question, these will never be the best qubits, but we believe it's the sweet spot between qubit performance and being able to maintain. The qubit in a quantum state for long enough, and also the other things that a system needs to be able to deliver scalable quantum. So when we talk about scalable quantum, that's really what we talk about.
Jason Lynch It's, it's about saying, okay, we may need to give up a little bit on the qubit performance, and this is not going to be the ultimate, you know, ultimately the best raw qubit, but when you put it together with everything else, it ends up being the best system. And that's really a philosophy that we've, uh, you know, we've learned the hard way through, um, these, you know, these products that, that the team has built over the last number of decades, that that's the way to, to bring technology to commercialization that you can't just focus on one piece of this. This has to be a system optimization.
Accessible and applicable
Mark Fielding So accessible quantum, accessible and applicable quantum, then. Useful.
Jason Lynch Yes. And, and again, I think one of the things I feel strongly about and the company feels strongly about is, um, you know, if you do look at these, uh, at, at, um, typical quantum presentations, it is these massive machines, it is sort of big infrastructure, it is the impression that this is going to cost a lot and it's only going to be available for the richest of companies and the richest of countries. And I think that's a dangerous philosophy, right? I think quantum needs to be used to actually level the playing field rather than to, to, you know, increase the gap between the haves and the have nots. And I think there is a sort of a, um, a view that this is extremely costly in the, you know, hundreds of millions or billions of dollars.
Jason Lynch Our company was formed on the basis that if we can get quantum working on these processes that have been, that have given us AI and all of these other things, then we can leverage the economics of that. And we can bring quantum to the masses. And again, not have it just for the richest companies and the richest countries.
Jeremy Gilbertson So that's, that's awesome. That's just, that's disruption. That's, that's, you know, we always talk about the calcified systems that are operating in the world. And some of those that need to be shaken loose a little bit and accessibility to technology is one of the bigger things. And that's really cool that you guys are, you guys are thinking about that. So, all right, so let's say, let's say Mark and I run like a mid-sized, uh, I don't know, like, no, no, no, no. Let's think about it. Some kind of healthcare technology that could potentially benefit from the use of quantum computing. So we've got a data center sitting in Atlanta, Georgia. It's got racks of computers, you know, rack mount computers. It's all operating.
The compute side
Jeremy Gilbertson We're doing some interesting things on the compute side, but how, how do you approach like a CTO, say Mark's our CTO? How do you approach someone like them to kind of say, Hey, you know, here's how you can transition what you're doing today into the quantum world?
Jason Lynch Absolutely. And actually we're, we've been out, um, talking to companies like that, uh, for the last six months and there's a massive amount of excitement. I mean, again, coming back to our philosophy, I think what we would say is we believe quantum should look, uh, you know, a lot like compute today. Um, we believe that we should be building on everything that we have. So the most powerful GPUs, the most powerful processors, we should leverage that and add the qubits on. And so, you know, our first machines are fully contained, uh, rack mounted systems that are, you know, a meter high and we'll plug into your standard data center. You don't need any special environment. You don't need any special vibration. These things are like your system, the systems that you use today.
Jason Lynch Now they do contain a closed cycle cryo cooler, and these are used. Uh, in MRI machines today, right? Uh, and, and I'm not sure, um, it's widely known, but an MRI machine contains a system that cools the components to three Kelvin. So we leverage that. And then we need to, we do need to get from three Kelvin to one Kelvin. So that's the bit that we add, but this is in an MRI machine today, a very contained, very reliable it we've had, you know, four of these machines running for the last two years in our own data centers, which is what gives us the confidence to go out. So what I would say to, you know, a CTO of a company that's interested is we're, we're, we have a roadmap that brings companies along the journey with us over the next few years.
Jason Lynch We have the benefit of, you know, having a very, um, economic, uh, upgrade cycle. Again, you don't have to build a new building. You don't have to spend tens or hundreds of millions more. We can very, because it's a chip, we can very easily upgrade you along that path. And you can have your engineers start to dabble, start to look at algorithms. And if it's sensitive, we can have it even in your own data center.
Jeremy Gilbertson That's, that's awesome. I'm going to, I'm going to stay down this track a little bit. I'm a, I'm a data center nerd. I worked in data centers for a long time doing transformation, doing migration, DR ops, all that kind of stuff. So like, so you have these rack mounted systems that can, that can fit into existing footprints. That's all great. How do you, cause I think one of the things that is confusing to some, and I'm not saying all CTOs are confused, but like, you know, how, what, what sort, what do they need to bring to the table to you guys to say, well, where should I put quantum, what should I, what systems should I direct to quantum compute processing to, to help out? Do you. How do you coax that out of people?
Jason Lynch Sure. So again, I think the way to think about this is starting with the use case. So I think the meeting point of a company like ourselves and a domain expert who has a problem that they believe could be relevant to quantum is that's the starting point. So with a lot of companies that we're working with today, we start with the use case and say, okay, what is the use case that you can see? And what are the challenges that you have today? So again, to give some examples of the applications that we're talking about, if you look at drug discovery or chemistry, I mean, the reality of our compute systems today is we can't even simulate the simplest of molecules. And one of the things that struck me on my own journey in quantum is, well, that's the reason why we do human trials, right?
Simulating drugs
Jason Lynch We can't actually simulate these drugs and figure out when I put this in a human body, is it going to be safe and effective? And that's why we have to trial them in humans. What if we could do that in a quantum computer? So drug discovery, I mean, we're working on another use case on optimization. So any sort of multivariate optimization problem with today's computers very quickly runs out of steam. So if you look at financial companies on portfolio optimization, if you look at network providers, they're looking at how can we optimize the network. So if we fast forward and think about when we have autonomous taxis or autonomous planes on our network, how can we, you know, the network optimization becomes a much bigger problem.
Jason Lynch So I think you start with the use case and then you look at, okay, how does quantum intersect that, right? So I think we'll get on to sort of quantum advantage and that discussion. But you look at, you say, okay, how many, what resources do we think you need to actually tackle that problem? And that may be, you know, maybe two years away. And so I think the journey then becomes, you know, putting some resources on, okay, what's available today so that I can start to dip my toe in, I can start to understand what these systems do. Because the other challenge here is, I think, as you guys mentioned at the start, quantum is a very different paradigm. It's a very different way of thinking. It's a completely different software stack.
Jason Lynch And so, you know, you're not going to be able to say, okay, I'll wait for two years until this quantum resources is ready, and then I'll figure it out, right? This is a brand new way of thinking. And so I think the value proposition or the meeting place is, what is the use case? When do we think quantum is going to intersect that? And let's build a roadmap to get you there so that when it arrives, you're ready for it and you're ready to take advantage. Awesome.
Jeremy Gilbertson Mark. Yeah, real quick. Sorry to commandeer this. I just got to get these two things out of my head real quick, and then we can juggle. That's cool. This is a follow-up to what you said, John. So, or Jason, I'm sorry. So, Jason, the idea of quantum being like they're atoms, it's a chemical, biological thing, right? So what you're saying is potentially quantum will be better at modeling biological and chemical things more so than traditional compute. Is that safe to say?
Jason Lynch Richard Feynman, again, that's exactly what he said. He said you're going to need a quantum computer to model quantum mechanics, to model nature, because it's quantum mechanical. So that's absolutely chemistry, biology. We're going to need quantum to model that or to understand how that works.
Optimisation
Jeremy Gilbertson Awesome. And then number two, the idea of optimization, right? So I think about route optimization, right? In the idea of qubits being infinite states in between one and a zero, just from a layman's perspective, right? So say you have, I think I ran across, it was Michio Kaku, actually, I think that had this analogy. You put a mouse in a maze, right? And traditional compute will look at each step at a time where quantum can look at the whole realm of possibilities and model it quicker and faster. Is that right?
Jason Lynch Exactly. So essentially what happens in, if you think of the optimization problem, it typically is modeled as a plateau and you have these areas of optimization. And the challenge we have in classical compute is we may find one of these optimization areas and our classical computer thinks that that's the optimum. And it may or may not be, we can't actually tell. What quantum systems do well is sort of analyzing the whole plateau and making sure that you're not in a local minimum. And that's really the promise, I suppose, of quantum and optimization is that because of the quantum mechanical characteristics, we can model the entire plateau rather than just very small parts of it. Makes sense. Thank you. Wow.
Mark Fielding And I'm going to just shout out to a former book club, The Nexus by Julio Tino, where we spoke about complex systems in nature and then how tiny, tiny areas, maybe a meter squared of the ocean has this interaction of so many species where the classical computers can't even calculate those interactions between eight animals or something. So yeah, that natural aspect of it. I want to get into quantum advantage, what it means, where we are, how is that possible? But just because Jeremy spoke about optimization, and I want to just go back to this temperature. You don't need absolute zero to these qubits. So the way you explained it, it was almost, I understood it as kind of non-optimized qubits, but they're still qubits. So they're still doing what qubits do.
IBM's roadmap
Mark Fielding I spoke to Joe Brose at IBM, and I think IBM's roadmap is like a thousand qubits by 2030. But the way you're doing it, what they're doing seems almost like redundant because what you're doing doesn't need to be perfection. So therefore you can get the efficiency of quantum computing, you can get the speed of quantum computing without this outer space perfection.
Jason Lynch I wouldn't say redundant. I guess the way I would classify it, and I think, again, you guys alluded to this maybe as a bit of a point on Joe's discussion, is quantum today is really everything to everybody. Quantum computing today encompasses anything we can't do in compute really. And my view would be that it is going to segment and there isn't going to be one winner takes all here. So I think the applications that we're going after need certain characteristics, and they are a lot of qubits and maybe not the highest quality. The three that we've identified, chemistry, optimization, and AI, really are well suited to the way we're doing it and the way we're proposing.
Jason Lynch There are other applications such as encryption that would potentially need much higher quality on your single qubit because if you're doing Shor's algorithm, which is a very long calculation, you maybe need less number of qubits, but a longer coherence time. And so my answer there is that this isn't a winner take all. I think today it is maybe viewed from the outside as one thing. But I think as we progress here, as you see in classical computer or classical applications, there'll be certain areas that will be relevant to certain technologies, relevant to certain applications. And the market for any one of those I think is going to be massive. So I think there's plenty of space for multiple companies here.
Mark Fielding Important answer. Thank you.
Jeremy Gilbertson So those of you watching, Jason, we've got actually a large amount of people watching us right now. So this quantum discussion is definitely resonating. So those of you watching in real time, feel free to drop a question for Jason in the chat. It's okay to use your silly card. Mark used his silly card. No silly questions. I've got loads. He's got a bunch of them if you need more. I've got so many stupid questions. No silly questions. How can a qubit not be... I guess if it's superpositioning, it can be anywhere at any time. So it can be any level of quality at any time. So get it. Just like the film, Everything, Everywhere, all at once. Watch that and freak your mind out.
AI
Mark Fielding Jason, let's talk about AI. You mentioned AI. So AI is getting a lot of buzz, both good and bad. Great uses, bad uses, really scary uses, unimportant uses. What can quantum do? Mostly unimportant uses. Mostly unimportant uses. But what can quantum do? Specifically, what can Equal1s version of quantum do to support, fix, help AI? And vice versa, that relationship.
Jason Lynch Yes. Yeah, absolutely. So let me start with the first one. So here's my perspective on that. I think it's pretty amazing where we've got in AI and generative AI and chat GPT and all of that. It's absolutely amazing. But if you look at the power that it takes to train these models and you compare it to the power that our brain consumes, you're talking about megawatts versus our brain takes about 20 watts. And I don't think anybody would argue today that these models are anywhere a fraction of what we can do with our brains. And therefore, my view of this is that as fantastic as the progress has been, we've taken a wrong turn. We've gone in a direction that is not optimum and is not sustainable. And I think quantum can potentially help with that.
Jason Lynch I think if you look at what AI is doing, AI essentially is building neural networks that are, again, mimicking some of what's going on in the brain, but they're mimicking them with classical circuitry. So we're taking a quantum mechanical process, trying to make a classical version of that, and then scaling that to as far as we can go. I think because quantum is probabilistic itself in nature, and again, back to the commentary around, it's going to be much better at simulating atoms and electrons and nature, then the best, and at this stage, it is a bet, is that quantum will do AI much more efficiently. And I think what we really need to prove that, as we did in the classical AI space, is compute big enough for people to get on and experiment.
Jason Lynch And I think you're going to see that here in the next number of years, where we will make progress and we will be able to prove that quantum AI is going to have a major effect. The second piece of the question is also really interesting, because we have an angle on that where, you know, so you'll see people talk about quantum for AI and AI for quantum. And AI for quantum, again, because these qubits are extremely fragile, they're extremely challenging, we're actually using AI to control them and to detect errors. So we actually were putting a neural network on the chip to learn the errors, and then to correct those in real time. So yes, there's an element of this about, you know, AI for quantum and an element about quantum for AI.
Jeremy Gilbertson Awesome, makes sense. We're actually starting to get some questions rolling in from the viewers. So what is a real life application of quantum computing in drug discovery, right? So there's a lot of chemistry that happens in drug discovery, a lot of biology that happens in drug discovery. What would a real life application of this look like?
Mark Fielding Can I just add something to that? I just wanted to get upstream on drug discovery and maybe add something to that as well. It's actually accurate and precise disease, actually, because no disease is the same for each. If you have a disease, it's not the same as other person. Actually, defining the illness of somebody takes years and years and years at the moment. I think that quantum can also bring that down at the same time as they're using it to discover drugs.
Two angles
Jason Lynch Yeah, I mean, maybe two angles to this question. When I read it first, I thought it was about, you know, what is sort of today, what is the application of quantum computing in drug discovery? And I think that's a discussion around, again, quantum advantage. So the race today in the industry that ourselves and others are running as fast as we can is to get a qubit platform and a system where we can scale the qubits to get to enough qubits so that we can simulate these molecules. And that's in process. And I think we need something in the thousand qubit range, which we're not there yet, but I think we will be as an industry and an Equal1 will be there in the next two plus years.
Jason Lynch I mean, in terms of real life application, again, it is very much about being able to simulate these molecules, being able to simulate the energy state of them. That's something we can't do today, even for the most simple molecule. So if you take paracetamol, you know, we actually don't know the physical mechanisms of how paracetamol works. It's, you know, it's just been handed down generation to generation. But there isn't a scientist alive that is able to say exactly the mechanisms. And that's because we can't simulate it. And so it's really about figuring out, I guess, at a base level, the energy states of these molecules and how then they interact with other energy states.
Jason Lynch And, you know, you have protein folding and all of these extremely complex structures that, you know, the likes of DeepMind and these other companies have made massive progress on with an AI approach. Quantum is really about, and I think of it sort of like that plateau analogy, right? The energy system of a molecule is extremely detailed, extremely complex. And when they interact, it's extreme complexity as well. And qubits do that well, again, coming back to the analogy of they are, you know, that system is a quantum mechanical system. So our tools today are very, you know, high level in terms of trying to be able to predict that. Quantum mechanical is itself at that level. And so it will be able to predict those interactions. So hopefully that answers the question.
Jeremy Gilbertson So, yeah, so the innate structure of small molecules that are involved in creating medicine to make us better, that point to fixing something that is also small molecules in our body, we have to do that kind of compute on similar infrastructure. Is that safe to say?
Jason Lynch Exactly. That's exactly it. Yeah. Yeah. Interesting.
Jeremy Gilbertson This stuff is fascinating, man. The more we talk about quantum, the more I start, you know, I can't remember what maybe Feynman said this to or another quantum physicist, you know, if you don't understand quantum, you're on the right track kind of thing. And if you think you understand quantum, you're lying.
Jason Lynch Right. That's right. That's right.
Jeremy Gilbertson Mark, where else do you want to go with questions?
Mark Fielding Well, we always talk about. No, go Jason. Sorry for dropping.
Jason Lynch I mean, I was going to jump in on that. Right. I do think people shouldn't be scared of this. Right. I think if you look at. So, again, my analogy to kind of the classical, our classical semiconductor industry. Right. When I was in college, I did electronic engineering and, you know, we had a course where we did the energy band gaps of silicon and silicon germanium and gallium arsenide. And, you know, I haven't used that since that class. Right. And I think we will abstract up so that people don't need to have a Ph.D. or more in quantum physics to understand these things. Right. Our history of being able to progress technology, my view, is a history of abstracting away the complexities so that more and more people can become involved. And that will happen in quantum. It's happening now.
Jason Lynch We're at the qubit level today, but we're building those foundational blocks. And again, we did it in classical electronics when I was in college. Again, we talked about error correction and ECL and all these different types of error correction that were needed for transistors. That's all still in existence, but nobody worries about that because we've got it figured out. So I think one thing I would say is, you know, not for people to be afraid of this. You don't need to understand the, you know, the full quantum mechanics. It will get abstracted up. I think it's more about understanding frameworks and maybe understanding the software stack somewhat to understand, you know, what's coming down the line.
What needs to happen
Mark Fielding So I was going to ask, we talk about on the paper, it's all of these emerging technologies working together. And we often see kind of quantum as the puppet master pulling these or in the future, pulling these other emerging technologies onwards. But I was thinking quantum computing, for all the success that everyone's having, the progression from the outside looks quite slow. People talk about decades before you get perhaps useful quantum computing. What technologies need to improve or evolve outside of quantum computing to make quantum computing more efficient, to scale more, to get more qubits? Yeah. Is that, is that, yeah. I won't even try to say what are the technologies.
Jason Lynch Yeah, absolutely. I mean, look, I think, you know, quantum has been going on for some time, right? So that's the first thing to remember, even though it's, you know, fairly recently entered the public consciousness, you know, this is a 50 plus year journey. And even if you go back, you know, to Einstein and, you know, it's even further, right. So there's been consistent progress. I would say that the rate of progress has massively increased over the last decade and the last five years. I think you're now seeing major milestones from companies in the ecosystem almost on a weekly basis. And I think we're now on a, you know, on a cadence that to me says quantum is no longer decades away. You know, quantum is going to be here in the next, you know, five, five, you know, years, right.
Jason Lynch So what do we need to make that happen? So I come back to, I suppose, a scalable qubit platform. So what do you need for qubits? You need a material system that is very pure and enables the qubit not to have these decoherence paths to various errors. So we need, you know, we need material people, we need, you know, a lot of it is about the interfaces. And so there's a material piece to this. There is a cryogenic, you know, a cryogenic industry making more efficient cryogenic systems. And then, you know, I think, again, if I think about what it is we're doing, we're gone from controlling thousands of particles to controlling a single particle, right?
Jason Lynch So whether it's, you know, our blend of qubits uses electrons, you know, you have neutral atom companies who are using atoms in a vacuum, you have photonic companies who are using photons of light. But the one thing that unites us all is that we're no longer, you know, worrying about a flow of electrons or, you know, thousands of these things. We're now worried about one. And that means the control systems, you know, everything needs to be orders of magnitude, more accurate, lower noise, you know, more precise. And I think that's a massive opportunity for the electronics industry in general, right?
Jason Lynch If you look at the control of qubits, it's with classical electronics, but with classical electronics that are much more accurate and, you know, much lower noise and an order of magnitude, again, more difficult than what we've had up till now.
Jeremy Gilbertson I love that explanation. There's so many adjacencies that make technology possible. And like, I think a lot of folks, you know, maybe just picturing one guy in the corner trying to get one qubit to behave or one molecule to behave. And, you know, it's like there's so many other things that have to come together in different ways, different molecules, different things. Jason, I think what you're doing is really awesome. I love the accessibility. I love the pointing back to Julio Tino and one of our other books that we read, The Nexus, talked about rooting new things in the way old things are done, right? It's a bridge to kind of build to the future. And I think it's a really smart way that you're doing it. And, you know, I hope to hear more and more about all the great things you guys are doing.
Jeremy Gilbertson You're very well-spoken in making this complex technology, this complicated technology, a little bit more accessible to folks. So thank you for that. Let's do the carryover question. So next week, do we have the PillBot guys coming on next week? Is that who it is?
Mark Fielding Yes. I think that's their real name, but yes. Well, I think the product is called... So it's basically like a pill you can swallow that does a lot of GI diagnostics, which is really interesting. It doesn't have to be related to that. But what would you ask that group if you could ask any question? And again, no silly questions, like silly questions are permissible, like anything you want to ask.
Jason's question
Jason Lynch Well, you know, mine is, my question is maybe quite a serious one, I suppose. I mean, I spend a lot of time thinking about and discussing talent, right? And I think all of these new technologies, whether it be quantum or, you know, pills, digestible pills, they need talent and we need, you know, the world's best people excited about this and engaged. And I think we have a major challenge across the board in terms of STEM with young people, you know, in getting them in and getting them excited. And so my question would be, you know, what out of the box ideas would you have for getting more young people excited about these technologies?
Jeremy Gilbertson Brilliant question. Wonderful question. And I've got a quick answer. They could actually join our book club and read books about this stuff. What do you think, Mark?
Mark Fielding I'm not sure that... Maybe we need more young kids reading. So yeah, I'm all for it, book club, but I'll be interested in a more... I don't know what the answer to that is, I'll be interested to hear. Any ideas, Jeremy, what would you say?
Jeremy Gilbertson Jason, great question. I think it's about catalyzing curiosity in people and pointing that to the trend of emerging technologies and figure out what ride you want to get on, right? And what you could sustain your curiosity to power that ride, right?
Jason Lynch Absolutely. And I think it's, for me, it's about also talking, you know, I think, and it's, you know, it comes out a little bit when we talk about quantum, there is this kind of, whoa, this is, you know, this is something that is very different. And so I think trying to normalize it and trying to make it, I mean, one of the things I have heard speculated that, I think is maybe a relevant point is that maybe the challenge we have with quantum algorithms today is that nobody has been educated quantum natively, right? We're all educated classically, and then we're trying to get our minds around this stuff. And maybe actually we need people who are quantum first and thinking about that, and then we'll have quantum algorithms coming out. So I think it's, it's about, yes, there is challenges.
Jason Lynch And yes, this is difficult stuff. But also, it's not insurmountable, we're going to figure it out, you know, it is going to become normal. And so, you know, don't be scared about it, I think is the other message I would give. Wonderful. Quantum first is




