Loading summary
Transcript95 lines
- [00:02]
A
Constanza M. Vidal Bustamante has done a dramatic work and impressive work of public service writing one of truly the best think tank reports I have ever come across. Quantum's Industrial Moment. Strengthening US Quantum Supply Chains for Scalable Advantage, co authored with John Burke, which just goes so deep. And I learned so much about all the things that it takes to make a quantum computer. It really reminds me, Chris, of reading those like 2018, 2019, 2020 Washington, wrapping its head around the semiconductor supply chain that ended up kind of delivering what ended up turning into the CHIPS act, the chips office program office and what have you. So there is, there's an enormous amount of kind of detail and knowledge every few, every few sentences. I was like, oh man, I need ChatGPT to like tell me like the 10 page version of this like 3 sentence thing that Constanza alluded to. But we are excited to give you a taste. You should all read and dive into it. The link is in the show Notes, co hosting today Chris Miller as, as well as Zachary Yerushali who is with us for part one, of course, Quantum Week.
- [01:24]
B
Constanza, welcome to ChinaTalk. Thank you so much, Jordan. I think I just found my biggest hype man. So I'm excited for that anytime.
- [01:33]
A
Chris, where do, where do you want to start this?
- [01:35]
C
Okay, well, Constanza, congrats on the report. I think maybe to start, tell us, what does it take to build a quantum computer?
- [01:43]
B
Yeah, well, that's a big question in the report. I think we try to lay out this kind of an answer to this complex question because I think it seems like it would be a simple answer at the start, but it's actually quite complicated starting from the fact that there isn't just one kind of quantum computer. Right. So different companies are pursuing different modalities. We have the superconducting computers, the atomic computers that could be neutral atoms or trapped ions. We have the photonic computing. There are many other modalities that have been cropping up as well that all of them have a different bill of materials. Right. Each of them will be pulling up from the quantum supply chain, from the various layers of the supply chain in different ways. Some of these are overlapping, so they're partially overlapping. They're not fully distinct, but they're distinct enough that it gives rise to the idea of there isn't just one supply chain. We have multiple supply chains that we should be taking care of. Now in terms of kind of the commonalities, I mean they're all drawing from kind of similar layers of the so called quantum stack. Right. So if you think about, you know, they're drawing from specific materials or maybe they're using distinct atomic sources, isotopes and so on. They put these perhaps elements within an environment like a cryogenic environment, very ultra low temperatures or maybe an ultra high vacuum environment. They're using different components to interface with these atomic sources or these materials that generate the quantum state. So maybe lasers or different control electronics. There's a software layer, right? There's like networking. If you think about kind of a little bit later in the future as we kind of start putting together different, maybe chips or so on for these different modalities. So kind of all of these different modalities kind of pull together from these various layers, but the specific element that goes in that layer or the multiple ones that go in that layer vary quite a bit. So that's why it's very important, I think, to be pretty. And why the report was on the longer side. It's because there isn't an easy answer, a very, you know, just one list of elements that you can point to and we're done. I think also these things are changing over time, so it gets complicated.
- [03:59]
D
Quantum is at a super early stage as a technology package. We are pre transistor and because of that you have to deal with the inherent uncertainty of supporting all these different supply chains in their current state. But it's also a wildly fast moving industry as well. And the next phase will require a step change and a reinvention of that supply chain, even if a lot of these existing modalities are successful. And the example there would be something like photonic integrated circuits, which are the photon equivalent of the kind of integrated circuit of the electron era. Right now most of the amo, which is one of the big approaches in quantum computing, it's a big cluster of them. They're using approaches that don't scale just based on the manufacturing methods. And to scale them you have to move to pix. And to move to pix yet again, you need to reinvent the supply chain and do that continuously. And it's a fascinating one to kind of grapple with, with a level of uncertainty that I actually really don't think we see in any other technology package at this scale.
- [05:13]
C
Yeah, and so the analogy is sort of like, you know, it's 1945 and we're two years out from the transistor having been invented and we're trying to think through, you know, what's the computing supply chain look like in 1955. And we don't know what the transistor is going to look like. Exactly. And so we're going to go through the cabinets at Bell Labs and figure out like on average, what are the scientists using as they run their experiments. Is that a good analogy?
- [05:37]
D
Yeah. Almost to the point we're almost pre vacuum tube from Stanza's perspective.
- [05:43]
B
No, I was. When we think about the heterogeneity of supply chains, it's not just across these modalities kind of horizontally, but also along the time dimension as we think about the prototypes that are being built right now, where if you map those supply chains and we have kind of a good sense of what those are right now, those are very globally distributed right now. And we can point to, you know, sources of dependencies and so on, and maybe some things that we're importing from China, where maybe it's the only source. Right. Or some other things where the best in market is from Europe or from Japan and so on. But as we think ahead to when we will have the quantum computers that will be capable of breaking encryption, for instance, so those kind of the version of these machines that we think will truly be revolutionary, the supply chain for those is actually probably going to look pretty different to the one that we have right now. And as we are in the report with John is when you think about geopolitical stakes and international competition, that's the place where the United States can still dominate because nobody has control over that supply chain yet. That doesn't exist yet. And so if we think carefully, and I don't think we're like fully without any idea what that will look like. As Dag said, we have some models where we need to move to photonic integration if we want to be manufacturing this at scale and at competitive costs and so on. And for us to actually build these machines at volume, we have an idea and we know kind of roughly the path to get there. It's just a matter of breaking this kind of chicken and egg cycle of waiting for enough market demand before you make these major investments in the supply chain. But because you don't have those investments, you never get to a point where the product becomes very attractive to the market. And so there's a path there. It's just a matter of gathering enough momentum and political will and capital. Honestly, at the end of the day, it's capital that we need to unlock that, the next gen supply chain for these machines. And it's something where the United States definitely is still in time to dominate. If we move quickly a couple of
- [07:58]
D
examples on the stakes of locking in your role as a country in that supply chain and why you get so much leverage when you do it. Again, think of the two dominant approaches. One solid state or superconducting, which requires cryogenic systems of a wild scale, the other on Amos. What's fascinating that I think take the cryogenic side innovation there. Right now you need a dilution refrigerator to operate these systems. It takes you 40 hours to go from room temperature down to the level of cold that you need to operate a superconducting circuit. What the 40 hours means is that you can only run one test a week. If China invents an ability to take that from 40 hours to 12 hours, you go from running one test a week to one test a day. And so your iteration cycle and then they'll lock that down and grab that supply chain on the other side. Picks for the AMO approaches. Nobody has really made a scalable pick like the architecture transistor part for it for that computer. It's really hard. The country that does that has a total lock on the ability to scale whole approaches in quantum computing. And actually that reads across to quantum sensing as well because AMO and quantum sensing approaches are pretty similar.
- [09:28]
C
So we've got these different kind of qubit modalities which are sort of like different transistor structures or if that's our analogy. But we, we, we know the supply chain underneath them has some similar parts. Maybe should we start since we had cold temperatures mentioned. Constanza, what does the cryogenic supply chain look like today? And maybe also just give us a, a bit of a glimpse as to how cold we're actually talking about here.
- [09:53]
B
Yeah, very, very cold. But I guess it gets, if you look even within the cryogenic section or cryogenics kind of part of the supply chain. It gets more complicated too because you have some modalities that require millikelvin temperatures and others that still are cogenic temperatures but are over one Kelvin. And that's a very, it doesn't sound like a huge difference, but it's actually quite substantial in the energy requirements and even in kind of the specific components that or subsystems that give rise to those temperatures. So those are almost like again a fraction fractioning of the supply chain yet again. So for instance, the superconducting modalities that we were talking about, so maybe the computers that companies like IBM and Google and I guess not Microsoft, but or to some degree maybe that they are building, they require these, you know, super shiny chandelier like dilution refrigerators that we usually see portrayed in the media whenever there's a quantum piece published. And those, you know, then you can get, like, farther into what it takes to make those refrigerators. And then for the photonic computing modalities, some of their subcomponents also require cryogenic temperatures, but not as slow as to require the dilution refrigerator. And so, but. But that also leads to other complications. So maybe we can talk about those. In the kind of dilution refrigerator camp, there are a few different issues. So starting from the fact that they use, and this perhaps is more well known, they use Helium3 as part of their. Their cooling approach. And so that is an extremely rare and highly regulated isotope that you can just kind of build or supply at will on demand. It's very highly regulated. It comes kind of a subproduct of nuclear processes and so on. And that, again, that seems to be an area where you can, if you start developing machines that scale and you need to access a supply of Helium 3, where you could find a choke point.
- [12:22]
A
But there is.
- [12:24]
B
Sorry. Oh, well, that is. Actually, I'll invite Zach to chime in here, because we discussed that as kind of like a. So, yes, it's very likely that there, or I don't think it's questionable that There is helium 3 on the moon. The question is whether it's ever going to be feasible to extract it. And so there.
- [12:44]
A
We're going to put data centers up there.
- [12:47]
B
I know, right? I guess I don't already see everything. But maybe, Zach, you've looked more deeply into this, into the lunar sourcing option.
- [12:58]
C
Should we step back and say, what is a dilution refrigerator? Like, how do they actually work before we get to the moon? Okay, okay, let's do it.
- [13:05]
A
Let's do it.
- [13:06]
E
Okay.
- [13:06]
C
Sorry, sorry, sorry.
- [13:07]
A
You guys better learn about refrigerator. Bottom line, we're going to the moon.
- [13:10]
D
Chris. What a buzz kill. I'm kidding.
- [13:13]
C
So before we get to the moon, so walk us through, how do these machines work as cold as outer space? What does it take to make a machine that makes things as cold?
- [13:24]
B
So I guess at a high level, they take several different stages to get to. You don't go from, like, you know, ambient temperature directly to extreme, colder than outer space temperatures. So the process. So, you know, you have these chandeliers, and they kind of, if you look inside of them, they have, like, different cooling stages that they go through. And so you go down to. I think it goes down to maybe 4 to 10 Kelvin first, and then you keep kind of going down all the way to the bottom is the coolest stage. And that's where you place the quantum chips for the superconducting or semiconducting spin modalities of computers. And that's the actual, the coolest part. And that's the part that's colder than outer space. So to get there, you require these combinations of helium 3 and helium 4. Helium 4 is not really a source of concern. That's like the most common helium isotope. So that's not so much a supply bottleneck that I'm aware of, but it's the Helium 3 part of the mixture that is the absolutely necessary element to get you to the millikelvin temperatures that these systems require.
- [14:40]
D
The whole point, the whole point of building a quantum computer and why it's hard is because these quantum states are incredibly fragile. They just get like messed with everything. Heat messes with them like the wider environment, cosmic rays mess with them. Looking at them messes with them. That's like the whole point of quantum mechanics. And so what you have to do is you have to isolate these things, these quantum states from absolutely everything. And the most effective way to do that is to get them wildly small and wildly cold. And also when you get it wildly small, it also a different type of physics takes over that enables you to manipulate these systems in such a way that you can do these useful calculations. And then on the Helium 3 side, I think the only this is one of those things that I really wish that helium A quantum computer made going to the moon economically viable. The sad thing is, and particularly for America, the major supply of helium 3 is tritium decay from the nuclear stockpile. And so as long as we don't go nuclear free in the US from most of the calculations that I've seen, we should be okay. That said, access to these systems, not just ones today, but ones that actually enable that scale, is critical. The last piece to this is there are three credible suppliers in the West. Not only that can supply these, it's Blue Fours, it's Oxford Instruments, which they just got bought by another company, and then maybell over in the U.S. it's that hard. There are only three companies, really, only two of them are credibly there at scale. China went from having none to just. In the last couple years, they have created more companies building these systems than the rest of the world combined and went from not publishing in this at all, and it takes decades to get good at this, to now dominating over 50% of the publications in New Innov in this space. And so folks are coming up the chain very quickly in places where the Finns will go and share their dilution refrigerator IP with us. China's not going to do that.
- [16:56]
B
To Tag's point about China announcing at least all these different manufacturers of dilution refrigerators, some people point to the export controls that the United States, along with several other international partners put in place back in 2020. Well, starting 2020, early 2024 and within a year or so to kind of evidence that we that this essentially backfired. And you had a situation where you had foot controls and the delusion of reservators importantly were part of those controls, accelerated that ecosystem where they rapidly just mobilized to be able to procure their own systems to continue innovating in the computing front.
- [17:44]
C
I'd love to dig into that as well, but can I go helium first? Jordan, is that okay? All right, helium. So it's on the moon and we're going to mine it on the moon. Maybe. But it also comes from the nuclear stockpile. Zach, is this from civilian energy production or nuclear weapons or a mix of both? Nuclear weapons, nuclear weapons. The number of weapons you have is the amount of is correlated roughly. Yeah.
- [18:09]
D
The majority of the source. You can get it from a couple other sources. Evidently Canada has loads random layer. There's storage for lots of helium 3 over in Canada. The reason why there's lots of helium 3 on the Moon is because cosmic radiation strips away and takes it from helium 4 to helium 3. But unfortunately, you need to launch a rocket there, you need to harvest it, and then you need to bring it back. The only economically viable use of that helium free from what I understand, is if you need to go to Mars and build quantum computers. So you launch off from Cape Canaveral, you get to the moon, and if you're still going with Elon Musk right on board, then you have an awesome business. But if we're focused on the Helium 3 supply for the US and keep it tertiary, then I've heard a bit of skepticism around that and I'm bummed.
- [19:12]
B
Right.
- [19:14]
D
But sadly, we are where we are.
- [19:17]
C
So we know that things like dilution refrigerators are hard to make, which is why there's a small number of companies and you need to mine helium on the moon or something comparably hard. On the other hand, there's an argument that the export controls that the US And Europe put in place on dilution refrigerators in China a couple of years ago spurred this brand new industry industry in China, which suggests maybe it wasn't that hard or at least it happened pretty quickly. So help us understand how we should think about this case study and maybe does it tell us anything broader about the relevance or not of export controls in the quantum computing space?
- [19:53]
B
Yeah, I mean, I think it really depends on the specific inputs that we're talking about and the timelines that are related to kind of the volumes at which you need them. So I think, you know, I think maybe what Zach was trying to say is that perhaps we shouldn't worry. So the helium three piece of it might not be so much a concern in the near term like we have. At the rate that we're building the refrigerators and that those are being purchased and acquired, I think maybe we'll be fine for the next few years. And luckily the US has a big source for that and kind of like we're in a privileged position in that we are the provider for much of the world as well. But what I'm thinking about too is as we, and you know, going back to what we were saying earlier about the kind of next gen supply chain, so as we start scaling these systems and you need to not just have, or you don't have one chip to cool, but then you start building these machines where maybe you'll need this kind of like a side problem of cryogenic or. Yeah, cryogenics in general, the dilution refrigerators, how much they can scale to support much bigger qubit counts. Once you start getting those rates of maybe you have a lot of demand for them for many, for many systems that are pretty large. Then I started worrying about whether we'll be able to keep pace with the, whether the sourcing of the helium 3 or the refrigerators themselves will be able to keep pace with that demand. And so going back to like the China example, I think, I don't know, I don't think that they're building these machines at volumes yet and certainly it doesn't seem to be the case at least yet that they're selling these machines beyond just being able to procure it for their own experimentation within their, their top level academic research labs that are kind of like at different tier of hardware development for quantum computing at the moment. So I'm not sure they were able to, I don't know, to develop these machines for maybe one or two systems or maybe more, but definitely not in the hundreds yet. Right. So it's not like they made enough, it seems like, to continue to make progress in prototyping and iterating over their machines. But I wouldn't Say that they're at the level where they're making as many as Blue Force in Finland or Maybelline in the US So that points to like, okay, well, you kind of accelerated their start of the development of these machines in house. Or maybe they weren't thinking of building that domestic capacity quite as quickly as they were pushed to because of the controls. But I don't think they've yet reached the stage where they've now equaled what Western companies are able to make. But they seem to be in that trajectory. So I guess you can still think about whether it was the right time to be putting controls on those.
- [22:52]
D
The China anecdote ultimately boils down to the stakes of this industrial competition, both how high they are and how different they are to other technology packages. Because we are so early in that race. Semiconductors, we covered this a little bit the last time. The US actually has an incredible moat around semiconductors. That doesn't mean we can go and sleep on it. But we've been doing that for decades. We have friends and partners and allies all across the world in China, by definition, because we haven't built a fault tolerant quantum system, a commercially useful quantum system, we don't have the same moat. And that means that China gets that ability to leapfrog and be pretty close to us in certain parity for manufacturing capacity, certainly with the US and then as you add in friends and allies, it's getting closer. But then if you look forward, the stakes of that are big. And this also hints at something that can stand. Panza spoke to, spoke to in the report, which is because we need to rethink how we do that supply chain to get to real scale. If China is the organization that comes up with the intellectual property on the core method to get to that scale, right. They invent the kind of transistor of the scaled cryo that you need, then they will have an unfair manufacturing advantage, which China's typically quite good at. They'll also have an unfair IP and understanding advantage on the key path that you need. And to reinforce this on, we have to think of it differently. It actually gets to. If we just project forward the existing engineering design of these subscale systems, we will not have enough helium 3, which is why we have to reinvent the systems that make These computers, the QPUs, the quantum processing units really like actually get to, to that temperature. We have to rethink that process. And that means that the country, the country that innovates and then locks that down and has the manufacturing intellectual property about it, they will have an unfair advantage to win. And we just, we just don't have the decades that we rest on as advantage and semiconductors here.
- [25:01]
B
Yeah, totally. I think I was just going to add quickly in the report. Exactly. We elaborate on this point where it's like okay, kind of the near term is we're thinking now the most advanced solution refrigerators that are available in the market, they can kind of host around 1000 qubits within them. So if you want to get to these machines where you need a million qubits or more, obviously that's not the only metric to look at. But if you're roughly. If we're thinking about millions of qubits, then you either like the path that we would have right now is to just put together dozens of these dilution refrigerators. But the, the scaling doesn't quite work that way either because as you have more qubits. Well, at least for these superconducting modalities that would require them. There's cables that connect the qubits together and that adds to the heat load and then that makes the cooling less efficient. So it doesn't quite. It's not as simple as just like multiply the refrigerator by X number. You need to be thinking kind of what Zach was saying. You need to innovate so that the cooling approach that you're taking is a lot more efficient at scale. And that's where we're kind of seeing. I think actually Mabel just put out a new system just at APS this week. Right. I haven't looked into the details yet, but that's where we need to put a lot of attention like Zach said, so that we don't get out innovated in this space. And then it becomes a lot easier for countries like China or others to get to that scale before we do. So we need. But that's the challenge. It's like we need to focus both on the near term supply chain to continue iterating and innovating while also keeping a very strong eye on what comes next because that's where we will actually reap most reward for in terms of economic benefits and security benefits from the utility of these large scale machines.
- [26:57]
D
If you could talk to policymakers and you could give them suggestions on what they can do, the policies, the tools that they can adopt in order to give the US the best shot here. What are some of the things that come to mind? What's the strategy to win on supply chain?
- [27:16]
B
Yeah, I mean so this maybe goes beyond the cryogenics is the subject we're discussing right now. But I think to this question in particular, you know, we have call outs there. And by the way, I mean the report tried to be comprehensive in its assessment of the problems, but the solutions that we provided, we know are very kind of preliminary and need a lot more fleshing out. And so maybe I'll have subsequent reports. I'm hoping to put a lot more detail into what the solution could look like. But broadly for this problem of cryogenics, I am calling for intentional and targeted multi year advanced R and D programs on cryogenics. Also similar dynamics apply for very highly precise laser systems and some other optical components and so on where the systems we have right now work for some of these prototype machines that we're building. But we know we need to start, you know, we need to continue innovating to get to the utility scale machines. So for those it seems like it is kind of an R and D tool, but it's not just fundamental R and D. It has to be given the race dynamic and kind of the time sensitive nature of this. It needs to be a very dedicated advanced R and D effort where another big point that kind of cuts across the report is bringing together the enabling technology manufacturers. So in this case the companies building the dilution refrigerators with the end users in the system integrators in the quantum world. So with the computing companies that would be using these machines to kind of have to the degree possible you can co design and really kind of get to very down to the specific requirements that these machines will have to kind of accelerate that process and not just build and then hope that they will be useful to their needs. I'm less worried about in the cryogenics sector that that's not happening already because I think the market for these machines beyond kind of physics research or quantum computing, I don't think it's that diversified. So I think they're definitely thinking about quantum as their primary sector and they're paying close attention to the requirements. But for others that have broader markets, you have to be, I think very deliberate. And from the government perspective, if you're setting up R and D programs to ensure that the enabling technology manufacturers are very closely aligned with the needs of the quantum end users.
- [29:56]
D
I look at three levers for it. There's supply, it's basically do I have the widget on the shelf when I need it? And that's the current widget. There's innovation which is am I skating? Do I have the support to skate where the Puck is going on the industry. And then the last is I think, of capabilities that exist in a market failure. And the canonical case there is high mix, low volume fabs like you see in the semiconductor era. And that's just because at the intermediate volumes that you need, there's a very explicit market failure on running those fabs given the cost structure of them. And the example that I give, just to make it specific, is the fab that we have at Elevate Roundabout, it costs about $40 million of capital equipment. And then our yearly basis, we hope and pray because we focus on a particular level of trlness, that we make about a million dollars a year on that. And that sucks because no investor is going to give you the money in order to, you know, give, give us $40 million. And we hope and pray that we make a million dollars a year. And so governments have to think about supporting that long term market failure in order to have that industrial capacity.
- [31:26]
B
I, I had focus on the R and D lever because I think that's the one that was most pertinent to what we were just discussing about kind of thinking about the next gen cryogenics. But certainly in the report, I mean we provide a menu of different policy actions that can be taken to support different elements of the supply chain. Again, kind of depending on the specificity of the issue at hand. So not all of the depending on the maturity, I guess, of the nature of the problem, the maturity of the specific components or systems. These will require different levels of support or things where the federal government might be more or less better suited to take action. So I think there's definitely not a one size fits all here, starting from Right. Like we started saying, the supply chain is so heterogeneous that it would be very shocking if one thing solved the entirety of the problem. And yeah, there are definitely some issues where I think there will be less need for federal activity than others. And others that will be actually like the Helium 3 is a good example of there's the highly regulated isotope. The private sector, it's not going to be the right actor here. It's like as simple as, or as simple, but as cleanly, I guess, structured as the isotope program under the Department of Energy can just make sure that they're taking very close look at their inventories and making sure that they still had parts of that inventory for quantum needs. Right. And ensuring they're doing the right calculations for, you know, repurposing some of the helium 3 that's already in use. You know, thinking through Potentially, you know, some kind of out there ideas for new sources of helium 3, but kind of being deliberate about that and that's a very specific example and there are others where you need you know, for if we get to kind of what you need to make some of our current foundries quantum ready or quantum grade, that's a completely scale of different scale of investment that's needed where and you know, you need to call on multiple actors to, to have a piece there. So there's a wide range of tools here that we can deploy depending on the specificity of the issue at hand.
- [33:49]
D
So my, my sense governments can either make markets or they can distort markets. Do you have any sense of a North Star on or heuristic on when government intervention is needed and when you should let the market do its work? Big question. It's so pertinent for this.
- [34:10]
B
Yeah, I think a big piece for me is not just that I don't think that the private sector could eventually do it. It's just that if you put all of that under a, a race, a geostrategic geopolitical race dynamic where it is time sensitive, you don't want to wait. I actually, I think yeah, if you press me, I think I would say sure, like we should just let the market take care of it and figure out, you know, which modality is actually best and if the maybe the one that has the most manufacturable supply chain and that relies the least on highly vulnerable items, that should be the one that wins, you know. But if we think that there's a higher priority objective here where we don't want to be second to anybody, especially China, then you are under very different set of circumstances where every day matters and we're not just where we want as many modalities for the US to nominate as possible. So I think that's another thing. It's not like will support whichever the most promising one is because let's say superconducting and I don't know that we have a good definition for what winning means. Right. But let's say a superconducting machine achieves is capable of breaking Shores algorithm first. Right. If we use that as an example or as a proxy for winning even at that point and it's a US based company, let's say at that point I wouldn't call victory. I would want to still get the other modalities to dominate in their respective categories because it is possible then that I don't know a different modality. Let's say I don't know. Photonic quantum computing in China will get to that, will win too. We'll clear that bar. And it's very likely to me to think too, that they might come up, they might break that barrier, and they might have figured out a supply chain that's a lot more nimble and cheaper and cost competitive that will then outshine the superconducting machine that the US Got to the finish line first. And so, I don't know, I mean, in a way, like, the finish line is moving and so it's kind of all hands on deck. And then when you start thinking about those circumstances, then there is a big role for the government to serve as an accelerator of that market. And so that's.
- [36:51]
D
I love this point, all of this
- [36:52]
B
in terms of kind of a broad innovation and industrial policy portfolio, because that's the scenario we're in.
- [37:00]
D
I love this point. The two things that occur, you know, first, if we got to vacuum tubes as a nation and we were like, I don't know, this works. This is good enough. DAG tools, you'd miss out on the transistor thing. That was actually pretty important to scaling these systems. So it's a repeat game. The thing that I do worry about is the stakes of getting policy wrong are wildly big. And the example actually that comes to mind is China itself. There's this technology area called quantum key distribution. We don't need to get into the technology of it. Folks can look it up online. It's really cool math, but unfortunately the math is so cool that if you do your postdoc in it, you just want to do that math all day and you forget that it is economically and cryptographically not all that secure. And because the head of China's quantum program, Penguin, is obsessed with this, he puts a wild amount of resources towards it. Even though you can literally just look up, look up NSA plus QKD and they go into like intricate detail on why this whole thing is dumb. And so it speaks to like literally, it's just like dumb. There's no easy way to categorize them. And so the upshot of that is we need industrial policy because the competition is so. Sigh. Is so intense, but it's very easy to get wrong. And we just hope that our competitor gets it wrong more than we do.
- [38:36]
C
Or at least I do on that point. Someone recently made the following analogy to me. It's that China has a Manhattan Project for quantum and that it's got one plan, one team, one system, and most ecosystem is kind of around that particular pathway. Whereas in the US the West there's these different qubit modalities and different companies and they're competing with each other and as a result you got these somewhat distinct supply chains. Is that analogy true or false? And if so, who's got the better strategy?
- [39:07]
B
I think that analogy used to be true, but it's changing rapidly. So I think the kind of comfortable narrative that we had of China for a while was that in quantum, was that, okay, China is leading undoubtedly in communications to the degree that they've deployed these large scale infrastructure, optic fiber, quantum key distribution that Zach was just referring to quantum key distribution systems to exchange keys supposedly in a tamper free way. And in addition to the fiber that they had deployed over, I Forget, it's like 10,000 kilometers or something in China you also have some quantum satellite link demonstration. Anyway, it sounds very impressive, but the assessment from the west was that even though they're kind of leading at least on the deployment of this technology and so on, this is not a technology that we care about or that we think about bringing a lot of value because it's a very narrow solution. It's not a full cybersecurity system in the sense that you still need to have a lot of classical encryption and systems for authentication for other parts of the cybersecurity piece. Even for the piece that they do cover, it's not fully secure. You can hack it in different ways. And so, okay, China can take that piece and kind of like we don't care about that one. In computing, the narrative used to be that they're kind of catching up quickly, but kind of like what you were saying, Chris, they're really only on superconducting, that they're putting a lot of their chips and they're moving quickly and they're impressive and we should watch them, but they don't have the diversity that we have. But I would say just in the last year or two, we're seeing a lot of actually startups appear in China. Usually many of these are led by pretty prominent academics from Pangen Wei's group or others that lead in quantum research in China across different modalities. So they announced I think maybe two different neutral atom computing companies last year. They have some photonic. Actually a photonic company has been kind of prominent for a while. They're growing the number of superconducting ones. So I think, and actually, I think just recently I was reading even topological qubits, there was a new development there. So all of those are new Companies. Right. They're not probably at the, with the information that we have, they're probably not very close to matching, you know, the capabilities of the various modality computing modalities that we have in the United States. But there's definitely a rapid movement there and it's not just again, kind of like, oh, but all of these are like state driven, they're not going to be effective and so on. Now these are kind of coming out as private startups from highly talented folks. And so I think we should worry about that and we should not just rest on our assumptions that yes, but you know, they're kind of limited on what they can do.
- [42:26]
D
So definitely the thing I would say, and this is one of the many reasons why I think Constanza's report reading that is literally a national security priority, is the reason that China is able to, to move up the chain so fast is because they are so thoughtful on their approach on supply chain. And what you can do is if you have the key components to manufacture all of these different modalities, these approaches to build quantum computers, then regardless of what you learn on which approach is better, you can react quickly and go and deliver against that. To bring this down to a very tangible example, we spoke before about photonic integrated circuits, critical tools in order to scale these systems. If you're in the U.S. even some of the biggest providers, because they don't have access to the fabs and the supply chain to actually manufacture those. It can take 12 to 18 months to go from an idea like, hey, I want this new pic to getting your pick in China because they've really invested in this area because you use it in lots of different areas in photonics, in certain material systems, you can go from hey, that's a cool idea to getting your pick in literally two weeks. And so the lock on the supply chain is a gift that keeps on giving because you can just be literally 10 times as reactive and adaptive as your adversary. It's like a supply chain OODA loop of sorts. And nobody's been attuned to this like this report that Constanza put together.
- [44:03]
B
I'll say two other things. Thank you, Zach, for that. I think two other things I came to know as we were speaking is two things like one for China and one for the US One for China. I think in addition to what I said earlier, that they have very prominent scientists that are now starting their own companies across different modalities for computing. I think what is true maybe and kind of, Chris, what you were alluding to, in the analogy with the Manhattan Project, is that they have been deploying those kind of moonshot programs to a much greater degree than the United States has. So they'll put out. And this kind of cuts across different levels of government. So a lot of the provinces or local governments are pretty frequently launching these kind of moonshot programs where they say, you know, submissions accepted for by 2026. Create a dilution refrigerator that can host 1000 qubits with these error rates. And those are usually just matching the Western, like the top performer of the West. And so in the timelines are usually pretty crazy. It's like within a year, you need to deliver this thing, right? So that might sound at first like it's not going to work, but I think they do it enough, frequently enough that like, eventually you kind of get there and maybe you get whatever, a thousand submissions, of which 999 suck, but maybe one does not suck, and when it's actually successful. And so I think there's that model that even if you have, you know, not talented folks that are running these programs, I think at least they have that coursing function of serving as a constant demand for these products. And there's some money attached to that. Maybe it's not super substantial, but it's enough to get enough submissions that maybe one of them will be good. So I think I worry about that, but I think the counterpart to that in the United States is that even if we don't have that kind of model incorporated so much, kind of the grand challenge or moonshot program style of programs so much in the US Although that seems to be changing with this admin. What we do have, I think, is highly talented government folks that are so deep on the specifics that they can craft programs. And I'm thinking here at darpa, doe, Nest, et cetera, programs that are really thoughtful that go for the right level of requirements, that have enough incentives attached to them that I think serves. You don't need to have a million programs like in China, you can have a few, but they're very thoughtful. They're driven by people that really understand the science and the technology. And I think that's an asset that we have compared to everyone in the world, I think so. China, I think, is a very easy counterpart to that. But even in Europe, I think you don't have the same level of technical sophistication as we have here. I worry a little bit about that changing maybe in the last few years, but I think in general, we still have incredibly talented folks in the government.
- [47:02]
D
The thing I would say to dovetail with this is that reactivity in the Chinese academic sector is incredibly powerful. I was chatting a couple weeks ago
- [47:17]
A
with
- [47:19]
D
a super nice prominent quantum physicists. They were talking to me about this paper that they read on a new type of pic where basically what happened was they were at cu, they had a certain type of material system that they were working on, and they had friends over at Columbia that were doing another type of material system. And these were basically adjacent publications and approaches. And what this Chinese group did was they looked at these two approaches and they asked the question, what would you do if you could just put them together? And turns out you get wildly better results. And what they were talking about was in America, hitherto you would never do that because the bureaucratic system around applying for grants is so intense that you couldn't just be like, oh, let's just like, put these two material systems together. Now what I would call out with the new admin, and again, I, for all sorts of reasons, don't want to get political, but I wouldn't say as a real positive is when you look at Deputy Secretary Debar, Undersecretary Gill, they kind of get that, and they're really driving to a totally new paradigm where you can say, like, oh, that material system cool, that material system cool. Let's drive this thing to see what new innovation you can do. And you don't have to spend new money. You just have to move fast and be creative. And they're going for it.
- [48:41]
C
Jordan, where do you want to go? We talk about lasers, right?
- [48:44]
D
Talk about lasers. Space lasers? No, not space lasers.
- [48:48]
B
Dark lasers.
- [48:49]
C
All right, Kansanzo lasers, critical for quantum computing. Tell us about where they are made today and how they should be scaled up.
- [49:00]
B
So in the report, we covered them as this big category of photonics and optics. So lasers are part of that and kind of, I would say, maybe the star of the show, but definitely not the only component to watch. And they, again, affect different modalities differently. So when we think about kind of the main subcategories that we cover, so the solid state, superconducting, semiconducting ones, the atomic ones, the photonic ones, it's mostly on the photonic or the. Yeah, the photonics and atomic modalities. And again, it's not that simple in that it's not just one laser either. It's multiple lasers that you need to do very different things. So if we take the atomic modalities, so the neutral atoms, the trapped ions, you need different kinds of lasers to cool the atoms. So instead of using the cryogenic systems, you use a laser system that brings the atoms when you beam, when you shine a laser beam on them, to bring them to ultra cold temperatures just to kind of be able to manipulate it. And then you have different lasers that then might elicit different energy transitions in them. And then you need other lasers to read out those effects. So it's a kind of a chain of lasers that you need. I guess what they share in common is that they're all highly specified in the specific wavelengths that you need to hit. And they need to be extremely stable in those wavelengths to maintain those frequencies that you need to elicit to, you know, make them usable for these computations. So anyway, I think just to say that it's not just one laser and it's not just lasers, it's like a lot of them. You need all these optical components to route the light and then kind of change different directions or maybe change the frequency of like double the frequency of the laser. Anyway, there's a lot of, of kind of subsystems that are relevant here, different lenses to focus the light and so on. The supply chains for those is also kind of complicated because they're also all different things. But the lasers that we cover in the report, I guess that we give the highest priority to, many of those are coming from companies in Japan, in Europe, in China too. I think here there's a very interesting case study that we cover in the report where a provider appeared in China that started manufacturing lasers that were essentially very close to identical to laser systems from a Danish company. And so there's been a lot, but at much cheaper price, surprise, surprise. And there's been a whole lot of kind of behind the scenes discussion over, you know, what was just reverse engineered. And this company is like well documented to get government subsidies and so on. So there's kind of like a pretty seemingly clear story of what happened there. But nevertheless, they have become a very important provider of lasers in the US ecosystem to this day. And I think what's especially kind of baffling is that even with the tariffs that we've had that have impacted everything, including the quantum industry, you can still call for an R and D exemption for those lasers. So that's still happening in our quantum industry. Those are still being purchased to this day by companies and even companies that can still claim an R and D exemption and universities. But that's kind of like, you know, they really have a good product going. It's very price competitive. They deliver apparently very reliably. And so it has become the preferred laser system for many of these organizations.
- [53:11]
D
This point on price is actually a really big deal. It's seen both in lasers and on the cryogenic side. There are key components like wiring trees which you need in order to operate dilution refrigerators. For most experimental setups you need a new wiring tree. The cost of a Chinese produced wiring tree is literally 1/10 even after the tariffs and all that different stuff of the US equivalent, I'd imagine. Similarly so on the photonics and then when you look at the photonics side, like cryogenics, there are only two credible laser providers for quantum systems in the us Vector, Atomic and Besant. Only two. They're still medium sized companies at that despite headcount. They're amazing teams. But the criticality is not just, well, if China underpins them, never allows them to get to real scale, undermines their ability to innovate, that's a real big deal. But even more near term, the laser systems in particular are used for quantum sensors. And quantum sensors, again, going back to a report that is literally seminal in the field written by Constanza vidalboostimonte, covers off how the thing that makes a quantum computer really hard to build, actually makes at a unit level these amazing sensors that can transform our world. One example of which is basically provide navigation without reliance on a GPS uplink, which as we find out in the war in Iran and we knew long before then is a really big deal. So these same laser systems play in.
- [54:48]
C
Okay, so, so, so question on, on this. So, so China's producing components for a tenth the cost of western firms. We've seen this far outside of quantum and many other spheres. And we're at this point where you guys are saying, and I think it's right, that we're going to have this dramatic scale up over the next half decade or decade of the number of all these components as we build bigger, more capable computers. So you know, option one would be to, you know, subsidize Western producers 10x so the price equalizes. That seems expensive. Option two, ban Chinese components from our quantum systems, but then you have higher prices. Option three, I'm not sure what option three is. What should we be doing here and should we be banning Chinese components from our quantum computers?
- [55:35]
B
Well, yeah, I kind of say I think this will be not well taken among the quantum quantum industry, but I do think we, we should not let this product continue entering the US market. But that aside, I think there are lots of things there And I'm so glad that Zaki brought up the sensors because that is, that is a much nearer term market that we will require these system or these laser systems and these various optical components much at scale, much more in the much near term than quantum computing. So it becomes a pretty real near term bottleneck. And I think in terms of the options, I think it's both we call there for some kind of support subsidies, but really kind of like strategic financing perhaps or tax breaks. Because if you think about also the supply chains of the lasers, some of those are also dependent from foreign suppliers, including for some of the tooling that they need to build them that might be from Europe or others. So I think there's some part of the solution, there is to provide some support for domestic suppliers while, you know, making it harder at least for the Chinese products to, to just take over the market, knowing that they were themselves potentially illegitimate ways of obtaining their IP or of accessing the IP that led to those products and also received very substantial subsidies from the ccp.
- [57:23]
C
Can I do a follow up question?
- [57:25]
B
Sure.
- [57:25]
C
So follow up on that. So suppose we were to go down the path of banning Chinese components from quantum computers. I think you kind of get in a similar set of questions of if you say to a lot of people in Washington, let's ban Chinese components from our AI data centers, people will often at first glance say, great idea. And then you say, well, wait a minute, what about the screws and what about the light bulbs? So where do you draw the line? Help us understand how to think about drawing lines in quantum computing?
- [57:54]
B
Yeah, no, that's an excellent question. And I think it's hard to have an easy solution to all. And even when, yeah, I don't know, I was thinking more along the lines also of even if you stop the import of some of these devices or some of these inputs overnight, that can also lead to a lot of problems in our own ecosystem in our ability to continue innovating in kind of the broader products. But to your point of like, okay, there might be all these other sub things that we weren't thinking about then should we also block those or is that worth blocking? I think, you know, so there, there, I think there are layers of complexity to some of these inputs and the ones that require the sophistication, I think, and they will have the value that we care enough to bring that in house. And I think in this case for the, you know, we have some domestic suppliers of these lasers, we have the talent and I think we have a path to get there and, and really good products that maybe are having difficulty because they're finding these anti competitive practices. I think that's an area where we know these lasers will be useful across different quantum technologies. So sensors, computers and to some degree the networking folks too, but also beyond quantum. So some of these lasers can to some degree serve telecom and to some degree various defense needs too. And so I think that's kind of a strategic enough enabling technology where I would want to preserve our domestic capabilities compared to much simpler inputs to the inputs. I think that's where I make a distinction. I don't have a super clear line in the sand, but that's kind of broadly how I think about it.
- [59:50]
D
So let's take the wire tree. If we say no Chinese wire trees, that means then that for some groups they can buy 10 times fewer wiring trees, which means they do 10 times fewer experimental runs. But it's not exactly linear, but it basically act immediately as a hindrance on our ability to innovate. And so they're real trade offs. What I am more clear on is, is the end state that we aim for, which is some mix of access, right. Is the widget on the shelf for security reasons? So for particularly end stage products, do we know the where that supply chain is so that China isn't, you know, putting in a little microchip into the thing so they're listening to our experiments or Stuxnet being something and then the last which is can we continue to out innovate? And now innovate is actually a lot more reliant, maybe even less on price than the speed that you can get the widget. And if for a lot of these systems it's not, you don't require new fundamental physics, you just require being able to run through ideas quickly. And there's a separate learning, I think it's associated with that which is understanding of how you get to scale. And so if you can balance between these different priorities, right, with that iteration speed as a proxy for how to stay innovative and how to do that at every stage of a technology cycle, I think we'll be in a good place. And there are lots of different ways to skin a cat. We just have to be mindful of those trade offs, at least from my perspective.
- [61:28]
B
Another key aspect to it's kind of related to what we were saying, kind of like how, how many layers down do you go? I mean a big point that we make also in the report is this category of specialized materials which are kind of like the ultimate substrate that you need to actually build a lot of the, for instance, the photonic integrated circuits that we were talking about. But several of the, just even like the bulkier lasers that we were talking about rely on these highly specialized photonic materials, many of which, you know, like these wafers that you need to then, you know, work process to make into these devices. Some of those are being sourced right now, so single source from China. And so that's another concern where it's like it's not the laser itself or not the optical or whatever photonic component itself, but it's the raw material that you need to build it. And that is highly problematic. So if you don't have access to that, then you can't go downstream in the innovation chain.
- [62:29]
D
I'd be curious for consent to your reaction. What is it Hackland's law of the second that you create the metric, it ceases to be useful? I mean, I'm forgetting what it's actually named after. But the thing that comes to mind is if you focus on whole product systems and you say how long does it take for you to go from like an initial design to that inception of the product and you try to reduce that as much as possible, then you identify the requisite bottlenecks that you need to prioritize for investment and you can both do that for non national security tech and just allow the componentry from anywhere. But then you have to apply a separate lens of national security stuff where you know, you probably don't want a certain chip coming from a certain place that's not the US and you look at the lead time for that. And if you just compare these two lead times and you try to ruthlessly bring that down in a general sense, but then also make sure you compare that with your adversaries, then you at least have a bit of a North Star on. How are we doing? Where do we prioritize? What do we do next? I hold that pretty lightly. And again, you guys have been thinking more on that, but that's where my silly bad supply chain brain goes to.
- [63:46]
A
So Constanza, you did your PhD thesis on managing people's stress over time. I'm curious, sort of as you're talking to all these people in the quantum supply ecosystem, like, what's their stress level? Are they like, you know, they got exams in a week, are they feeling good?
- [64:13]
B
Oh my gosh, what an honor. You went that back into my history. I guess it hasn't been that long. If it's an honest question. I mean, I'm happy to Answer it. I think in talking to the Quantum folks, I think there's a lot of excitement, but also a lot of uncertainty over. I think especially, you know, obviously I approach this from a policy perspective. You know, I think there's a lot of enthusiasm from the administration and Congress to do something big on Quantum and to build on the foundations of the National Quantum Initiative that came out during the first Trump administration and the National Quantum Initiative act that kind of solidified that and provided a lot of funding mechanisms for specific programs at different agencies to take that forward. So I think there's a lot of expectation, but also I think a little bit of fear on what will actually happen. So I think the stress levels are real and obviously all of these companies have a lot of pressure to deliver on these machines by the roadmaps that they swore by. Right. So they've all been claiming that they're going to have, they're going to start delivering some utility scale machines by the end of the decade and kind of the clock is ticking on that. Some have been more aggressive than others on what they will deliver. And so there's a lot of expectation there on whether they will deliver and if they don't, what will happen to the field, both to their own company, but more broadly, you know, even if my competitor firm fails, will that lead to a generalized, this lack of confidence in the field that will maybe bring down the investments, the private capital writ large? And I think there's a lot of fear like what will actually happen. So there's a lot of pressure that they're feeling right now for sure.
- [66:00]
A
Can you compare that to the, you've also done some research interacting with the semiconductor community. I mean, I don't think they're worried. I don't think the chips folks are worried that chips aren't going to be a thing.
- [66:13]
D
No.
- [66:14]
A
But what other sort of like anthropological differences have you picked up on?
- [66:19]
B
Oh gosh, that's such a great question. I think it's just very different environment, but at the same time there is the incentive to present it as much as being close to the semiconductor industry as possible to kind of give this idea of we have a path to manufacturability or kind of build on top of the chip Chips act or the Chips and Science act energy and kind of come up with this big industrial moment, as I called it in the report. What's interesting, I think, is that compared to the semiconductor industry, you have all these different modalities that in like close to apparently 99, 90 now, companies that are Building quantum hardware across these very various different modalities that's just so different, I think, from the semiconductor industry that it leads to all sorts of competition among them over who has the best qubit and why the other one sucks. So I think that's always funny to hear where everyone will tell you endlessly why you should support their qubit modality, why they got the right one going.
- [67:26]
A
Yeah, that was really my big takeaway from my little quantum journey over the past few weeks in the semiconductor industry. It's kind of consolidated and like, okay, you got two EDA players, you got a handful of people who make photonic masks, you got one, one company that's making EUV machines, and it's kind of going to. It's been that way for a pretty long time. It's probably going to stay that way. Maybe you'll have like an entrant here or there on the design side that. That shows up. But, like, you really feel like, like this entire industry is sort of pulling in basically one direction and everyone's just like trying to make sure that they can maybe capture an extra 10 or 20% of where they are in the, in the supply chain. But when you sort of walk through the stack here, it's like, yeah, everyone's kind of using the same ingredients, you know, to more or less of an extent, but, like, the actual, like, what the computer is going to look like is totally up for grabs. It's not like Game of Thrones because as you said, there's like 90 different little empires, not just like six or seven royal houses all competing for the prize.
- [68:45]
B
That's right, yeah. And we'll see how many survive or how many what diversity survive. And even, like, we didn't even get to this. But even within a modality, there's different ways to build your architecture and so on. So there's. There's a lot there. Yeah.
- [69:00]
D
My mental model for Quantum is biotech. And so in biotech, when you're trying to cure cancer, you have small molecules and you have CAR T and you have antibodies and you have immunotherapies and you have all of these different approaches that are trying to go and address something out there which is kind of the unified target for it. And it's just to say that in other domains, we have figured out how to take hardcore fundamental science and maturate that ultimately to impact our life, even when there are lots of different approaches to it. So that's where it's a little bit different of a mental mindset to semiconductors but it doesn't mean it necessarily exists. The one thing I would say about this one of 50 reasons why I'm so excited that Jordan, you're covering this. Chris is super attuned to it. And then Constanza's serial seminal reports is folks have been spending decades from different perspectives saying how do we get biotech right? Right. Public policy. Folks like have a frame of reference around biotech. Public finance. Folks understand it, doctors understand it, physicists understand it. In quantum. That hasn't happened yet. We haven't had proper. Proper academic. I really don't think we'll get this right unless we bring that interdisciplinary best practice now at this stage. So I'm super soaked for more.
- [70:32]
A
All right, kids.
- [70:32]
D
Constanza, what's your next report?
- [70:35]
A
Constanza, I'm sure they'll find some. Find some work for you.
- [70:40]
D
Jordan, I. I have to interrupt because I have to thank you. The first episode, I was like wildly nervous to talk to you and Chris because I've been fanboying China Talk as a china. I'm not going to run the back.
- [70:55]
A
I'm going to stop.
- [70:55]
D
I know. I mean it. I mean it. I was fanboying.
- [71:01]
E
The council of cubics convenes.
- [71:04]
F
90 houses claim the quantum throne.
- [71:08]
E
Let them tear each other to the bone. 90 houses of the Quantum Throne.
- [71:20]
F
90 swear the crown is theirs alone. 90 roadmaps, 90 end of decade vows.
- [71:27]
E
Who will take the throne? Who will take it now? I am house superconducting the Millikovan king. IBM and Google bow before my ring. Trapped iron rusting gates and calls it all. Sit down ion. Your throne is small photonic, room temperature. Dreams come back when your wafer is more than it seems.
- [71:54]
F
I am house trapped ion. I do not lie. My fidelity touches sky superconducting 40 hours just to cool one test a week, you frostbitten full neutral atom you're. Your lattice is sweet, but your gate speed is a glacier in retreat.
- [72:15]
E
I am house neutral atom. I cool with light alone. No chandelier, no helium, no fossil throne.
- [72:21]
C
Helium 3. Helium 3.
- [72:23]
E
Go mind the moon, old man. I'll have 10,000 atoms before your fridge begin.
- [72:28]
F
I on your gates are clean, I'll grant you that. But I'll run the machine while you polish one quiver flat. I am house photonic eye Walt and room degree. No fridge, no vacuum, no cryogenic feet. Superconducting your chandelier is a two my own. Your vacuum chamber is a coffin in the room.
- [72:52]
E
Adam, your tweezers are a parlor trick at best.
- [72:55]
F
The pic is right. 90 corbits, 90 vowels. 90 knives and 90 vowels.
- [73:05]
E
The throne is empty. The throne is empty. Still, I'm not too busy stabbing to climb the hill.
- [73:14]
F
The hall falls silent. A stranger steps forth from the east. I am Origin Quantum. I am Spink. I am usdc. I am Jefe in the mall. Warning to you, Zhang. On the sea, you four spent an hour tearing out each other's throats. I've been building wiring, trees and lasers and boats. Your lasers, I clone them. Your wafers, I own them. Keep fighting, my lords. Keep your pride. I'll take the throne while you bleed each other dry. 90 houses, 90 crowns. But the substrate is the king, and the the one who owns the substrate owns everything.
- [73:57]
E
Read the report. Link is in the show Notes.