Infleqtion, Inc. (INFQ) September 2026 Conference Transcript

Okay, awesome. Well, thanks everyone for joining, staying with us. Delighted. I'm David O'Connor from the semiconductor team here at Piper, and I'm delighted to have CEO of Inflection, Matt Kinsla, with us here today. Thanks so much for coming. So, Quantum, tell us, you know, where exactly is that fitting in? What problems are we solving? And where is it fitting in with AI, basically? A lot of discussion around AI here. So a lot of folks asking, you know, does it displace it? Does it complement it?

It doesn't displace it. It definitely complements it. And it's kind of a virtuous cycle, I guess you could say, in which they each complement each other. But I will brought a little prop so this is at the heart of all of our quantum products and What we do is we take advantage of the quantum mechanical Nature of the atoms that we trap inside these chips, and then we turn those into a suite of products and those products range from quantum sensors to quantum computers and the reason why I lay it out like that is because is the interaction for and between quantum and AI cuts across both the quantum sensing and the quantum computing world. So where does quantum computing and AI intersect? Well, AI is helping us get to useful quantum computing faster. And so just like all of us are doing our jobs more efficiently, utilizing some kind of large language model tools, our quantum physicists are utilizing them and making progress faster than they were before they were utilizing those tools. And also AI ends up being very helpful for a critical part in the process of getting towards useful quantum computing, which is identifying the root causes of errors in the system. Identifying the errors is a quantum problem, but tracing them back to the root cause is an inference problem. And so utilizing large language models is quite useful. And so NVIDIA has equipped the industry with tools that are helping us get to useful quantum computing faster. When useful quantum computing gets here, and that will be, we believe, in the late 2028 time period, it will work very closely with GPUs to solve a swath of problems that GPUs or CPUs couldn't solve on their own. And those are traditionally problems that are quantum mechanical in nature, like the interaction of electrons coming together when you're smashing two atoms into each other. the way that those electrons interact and they come together to form a molecule is a very complex system and again quantum mechanical so it's inherently probabilistic and it would take classical computers like the life of the universe to model that or you would need a classical computer with like as many atoms as there are in the universe those are the types of things that a quantum computer because it's utilizing a fundamentally different architecture can simulate, and it will work hand-in-hand with AI when the time comes to solve those types of problems. And then quantum sensors actually can help utilize data centers run more efficiently, so we build clocks that are about a thousand times more precise than classical clocks that are used to synchronize workloads across distributed servers, and if you use quantum clocks, you can basically chop time up into smaller bits and therefore utilize more of the underlying infrastructure because you've cut the kind of computing capacity up into smaller bits that can be synchronized back together more tightly okay so a lot of overlap excellent excellent and you know quantum it's such a new new market a lot of investors you like to know the size of the market you guys are kind of on the computing and the sensing side of things so maybe you just help us

understand or what's your view on kind of the market size the opportunity and computing versus sensing part You know, it's hard to put your finger on, honestly.

I'll defer to McKinsey's view, which is, I believe, you know, something over $100 billion in the computing side of things and something in the $30 billion side of things for sensing. You know, that's probably off by an order of magnitude one way or the other, but they are sizable market opportunities. I believe quantum computing will be the larger opportunity because it's going to enable all sorts of use cases that we haven't even conceptualized yet. But even if you just zero in on some of the use cases that what I outlined could be made possible, like, let's say, utilizing quantum alongside AI to come up with better batteries. Imagine if your phone could last for a year on one charge versus a day, or better jet fuels or better materials to get Starship up into space and back down. and it's one of the things that SpaceX is dealing with is material science issues, and those are the areas where we'll see real benefit from quantum computing first. And so those will be the trillions of dollars of economic value is just you bring in that one subset of what quantum computers will be able to do into the world. So very, very large market opportunities. On the sensing side, I think you can think of it more as a replacement cycle, honestly. Like, it's a little bit easier to grok because you don't have to use your imagination to think of what are we going to do with these quantum sensors. What we build is we build quantum clocks, we build quantum antennas, and we build quantum gravimeters that can sense gravity with extreme precision. And there are existing classical analogs of each of those. The precision timing market is quite a large market. It's mostly dominated by two companies, Microchip and Saffron today. And, again, we're bringing a product that has orders of magnitude better performance into the market. and then the antenna market is hundreds of billions of dollars that our quantum antennas can address many of the use cases that modern antennas, classical antennas address. And then what we're working on in the gravimetry side of things, what we're doing with NASA is a good example, which is probably going to be already as cumulatively a $40 million contract, which will continue to get bigger where we're putting our sensors into space to sense changes in gravity on and below the Earth's surface. So all of the sensing applications are large market opportunities as well.

Okay, and you mentioned 2028 is kind of, that could be a key year where it kind of goes mainstream. Is there one particular application that's going to drive that hockey stick in 2028?

It's less about the application. There will be an application where I believe we'll point quantum computers at in late 2028, but it's more about when we hit that part of our roadmap where there's enough, what are called logical qubits to actually have enough power to do something useful. And so our roadmap calls for us getting to 100 logical qubits by the end of 2028. And then I think the industry generally believes that's the place at which we'll start to do something useful. It won't be useful across all use cases. It'll be useful for something in that material science world where you can actually model the electrons coming together in a way that you couldn't do without a quantum computer. And one of the specific use cases we have our eyes on is something called Indian tin oxide. It's a translucent conductor. You probably have some of it in your phone. And it's a relatively hard-to-come-by material, but it's the only material that we know of that can do the things that we need to do to have that type of conductivity in a translucent material. And we believe we can build a replacement based upon modeling it out with a quantum computer at those lower logical qubit levels. and so probably something that you could see happen in the 2028-2029 time frame.

Okay, so to model out a new chemistry to replace that chemical chemical supremacy. Okay, very interesting. And you're at 30 logical qubits.

Yes, we will be at the end of the year.

End of the year, so you talked about 128. I think you've talked about 1,000 in 2030. You know, what are the problems you're solving on that roadmap? Like, to go from one to one, is it just engineering problems? Is there, maybe if you could talk around error correction, fidelity, what are the key metrics as you go from one product to the next?

Sure. Well, for the whole industry, a very big moment happened in 2023 when we got logical qubits for the first time. So no one had them pre-2023. The industry got them in 2023, and from that point forward, there's really been no looking back. And up until the logical qubit era, let's call it, it was still relatively unclear as to when, or not even when, if quantum computers would be useful, because no one had shown a logical qubit, and quantum computers would be really not useful without logical qubits. So post-2023, we get logical qubits, and now it's truly a matter of when, not if. Inflection showed its first logical qubits in 2024, and we have announced 12 logical qubits at the end of last year 30 by the end of this year and 100 by the end of 2028 so what are what problems are we solving to increase that logical qubit count at the end of the day what are we doing at least for our modality we are controlling atoms with lasers with extreme precision and then applying error correction code on top of those atoms that have been trapped in the in the in the little chip i showed you so everything we're doing is just learning and and honing in how to control those atoms with lasers with more precision. And that is entirely an engineering challenge at this point in time. The physics breakthrough happened before the logical key bits came about, which is really exciting. So now it's just, how do we do this with more precision? And that is truly in the realm of engineering and then software correction. So that's what we spend all of our time on the computing side of our business, figuring out how to do. Okay, okay. And you talked about the different modalities there.

Yours is, of course, the neutral atom. Why is that so much better than when we listen to superconducting, photonic, trapdine as well? Why is that modality approach a winning approach? Can you just talk about what the pluses and minuses on the modality?

Yeah, well, at the highest level, it's the only modality that can do both the quantum sensing and the quantum computing side of things. So that's exciting in and of itself that we have this whole other opportunity in the form of quantum sensors available to us. Specifically on quantum computing, the highlight of neutral atoms has always been their scalability. And so we can fit, basically we can produce many, many, many physical qubits. And that is much more challenging for some of the other modalities to do. And so in that little chip I showed you, we can scale to hundreds of thousands of physical qubits. With superconducting, you need to manufacture the actual synthetic atoms, which are little pieces of superconductive material, and put them into a large freezer. There are engineering issues that come with making those. We get our atoms for free from nature. They're ribidium atoms or cesium atoms. And so the scalability is really the primary thing to like about neutral atoms. We hold the record of 1,600 physical cubits. And then what had always been one of the knocks against neutral atoms, and before I became CEO, I was the first investor in the business. And when I invested in the business back in 2018, it was really just an idea coming out of CU Boulder and neutral items had only been explored in academia for quantum computing the other modalities have been invested in since the mid 90s and the mid 2000s respectively but there was a question as to whether neutral items could ever be high quality so we knew we could scale to a lot of quantity but the quality as measured by gate fidelities was very low it was call it like 50-60% when I invested in the business and and trapped ions were 99.9 percent say fast forward to today we've gone from 50 60 percent to 99.73 on our way to 99.9 and so we've really solved the quality side of things

um and that's why you've seen uh inflection and and other neutral atom companies really lead the lead the charge with um with their logical cubit counts is there any particular kind of applications more suited to um neutral atom versus others or you know task in other ways Does more than one modality win out in the end? Can there be a few?

I think there can be a few. I think it could be like the memory architecture, where you have different modalities playing a different role in a broader, complex quantum computing system. That said, I think in the near term, one or two of the modalities will actually get to that breakthrough of doing something useful. And at that point, you'll get an outsized amount of customer attention, capital, and talent rushing to that modality in those companies, and they'll get some, call it breakout speed, and likely will run away with the market for a bit, and I think others will catch up. There's not really necessarily applications that are specifically well-tailored for one modality versus another. It's really just getting these modalities powerful enough to do something commercially useful. And the ones who can do that first, I think, will get a lot of attention.

Okay, okay. And the fact with using less cryogenics, optics, cooling, all that, how much of a differentiator is that on the neutral-added approach?

We don't need any freezers, so everything's done at room temperature, and that is actually what allows us to have that flexibility to do the whole sensing side of the application, because sensors need to be field-deployed, and you can't really field-deploy massive freezers. And so the room temperature aspect has the benefit of allowing us to focus on sensing, but a lot of benefits within computing as well. It inherently lowers the bill of materials. Those freezers are expensive. We don't need to purchase them. It makes scaling easier because we don't need to purchase a bunch of them as we scale our qubit counts. And then finally, it has huge benefits from a power consumption perspective. We run our quantum computers on the equivalent of six to ten hair dryers or so. So when we were in discussions with the state of Illinois who were selling a quantum computer, they were asking us, like, oh, do you need to get hooked up with ComEd? We can get you, you know, special power supplies, and we just don't need that. We can plug it into the power outlet.

Okay, okay. And even on that, on the supply chain side of things, you know, what are the critical kind of components that you need access to? Lasers you touched upon earlier. or anything else on the optics side of things?

Lasers are the number one thing from the supply chain that we need to make sure we have our fair share of and increasingly photonic integrated circuits. So PICS, as the laser systems can be integrated down into silicon, we need to make sure that we have the capacity that is out there. Right now, that's not an issue, and we're making sure that we have our fair share of the supply chain over time. And we make the little cells I showed you ourselves and we're really the only people who can make those and so we control our destiny on that piece of the supply chain.

Okay, you have your own fab to make those all.

We do, yeah, we make them internally.

Okay, and is that much of a differentiator versus peers having to partner with different providers?

It is. I mean, if you think about most of the folks in the Neutrotum world, Many of them are using our components as their foundation layer for their technology. And so it is a pretty strong pivot point that we own in the supply chain. And so I do think that is quite a differentiator. Okay, okay.

On the partner side of things, you know, new technology, you need to partner with a lot of the ecosystem. Who are the kind of key partners Inflection is partnering with and why, you know, what are you getting out of the partnership?

Well, we work very closely with NVIDIA. We made an announcement with them earlier this week on integrating some of our error correction codes into CUDA Q, which is their flavor of CUDA for Quantum. And we've done a lot of different work with them from an applications perspective. But, you know, if you think about how, going back to how AI and Quantum are going to interact, the connectivity layer between GPUs and QPUs is going to be absolutely critical. And so having a good relationship with the folks who are making all the GPUs is going to be very important. And why does NVIDIA care about quantum? Ultimately, they view it as a way to sell more GPUs over time because we'll open up new swaths of types of problems that we can send compute after, which we haven't been able to in the past. And that will mean they get to sell more GPUs into it. So they're our most important partner on the computing side of things. On the sensing, we just announced a partnership with Cisco, which is super exciting, and that will be to do some work with quantum networking and quantum repeaters, so you can think of kind of like what they do with their traditional routers, but in the quantum version. And then on the sensing side of things, we work very closely with Saffron, and Saffron is one of the leaders in the classical version of precision timing, which I'd mentioned, And so they view quantum as the next tool in their precision timing toolkit. And so we're working with them to utilize their distribution channel and start to have a replacement cycle of quantum technologies. And why that benefits them is they own a lot of the optical time transfer networking gear. And so as we put more precise clocks into data centers, it pulls through the need for more of their time transfer gear and vice versa. so it's this nice little symbiotic relationship that we have with them. And then we're very close partners with many of the agencies in the federal government, so Department of War, the intelligence community, and then the Ministry of Defense in the U.K.

Okay, okay. And so the governments are a key kind of partner for you guys, a big contributor to your revenue. I mean, long-term, do you see that they're the enabler to help enable the ecosystem, Do you see long-term that being a significant kind of contributor on the revenue side as well a very significant contributor on the revenue?

You can think of virtually all of our quantum sensing products as having Very real use cases on the battlefield for instance and so the Department of War will continue to be a big customer for us as well the Ministry of Defense and then from a quantum computing perspective I believe we'll see the Department of Energy procure systems for the national labs in the not-too-distant future, and that will expand into the NSA and others, and then, of course, the Department of War as these computers get more and more powerful. And so I think the U.S. and our allies will be big, big customers for quantum for a while. And you're right, they're occupying both a customer as well as an enabler seat today. What they're doing, what the Department of Commerce is doing in giving ourselves and a few other quantum companies some funding in exchange for equity stakes is very much kind of their enabler role, but also will be a very big customer. They are already a very big customer, but will also be a big customer in the future.

Okay, okay. And just to understand the sensing again versus the computing element. I mean, you sell both, but the government's more on the sensing side versus computing. Long-term, what does that mix look like in general?

I do believe over time, both from the government as well as just overall, compute will be the bigger piece of our revenue, just because I think that it will unlock market opportunities that are significantly larger than the sensing market. But still, the sensing market will be very, very large in and of itself. I think one important thing to take away, though, and I think I made this point at the beginning, but just to hammer it home, the underlying technology is the same. You can just think of them as varying levels of complexity of what you can do with that underlying technology. So our quantum clocks and our quantum RF are almost like less complex versions of the computer and as we get better at commercializing those less complex versions it really directly leverages our ability to build better quantum computers okay okay maybe switching over in the couple of minutes we've left just onto the financial model what what should investors are sure how should they think about the financial model what targets have you given how can investors track what your progress so we guide the two metrics and I think those are the two most important metrics for investors to focus on. One is revenue. And so we've got it to $45.1 million in revenue in calendar year 2026. So you should hold us to that and make sure we hit that. And then the other is the logical qubit number, which isn't a financial, but it's our most critical R&D metric. And so we have published our targets for that number. And I think that's a fairly important metric to follow, only because it really is the number when you get to 100 logical qubits, that's when we will be able to do useful things. And there's a lot of other numbers thrown around out there, but I think logical qubits encompass all of those numbers. And so that's the one that you really should focus on and honestly should focus on across the quantum industry. In terms of other financials to keep your eye on, I think cash burn is always an important one. Obviously, our goal is to become a massively cash flow generative business over time. Right now, we aren't that yet, But we are much less cash-consumptive than any other quantum company. And so I think keep an eye on our cash burn relative to our cash envelope, probably. So we've got $550 million on the balance sheet, and we're burning something in the mid-40 million this year. And we will invest in the business. That will go up next year, but I think you can expect to see us being much more capital-efficient than most other players. Okay, okay.

Again, just typically at this stage of a technology development, you have to kind of create the market from an M&A perspective. How do you think about that M&A, particularly given where Quantum is at? Any other kind of core technology that you really need to bring in-house? Just if you could talk about that.

Well, I mean, I was an investor for 18 years before coming to the company full-time, and I have seen my fair share of M&A done. And unfortunately, some of the times it works really well, but a lot of times it doesn't. And so I go into M&A with a high degree of, not skepticism, but I just have a very high bar for doing it. And so I wouldn't expect to see transformative M&A from us, but I would potentially expect to see tuck-ins. Because the kind of M&A that does tend to work are nice technology tuck-ins. And so there are, let's say, opportunities to potentially solidify leadership in some of the sensing verticals that we focus on and also own more of the supply chain, particularly in photonics. So if we were to do any type of M&A, it would likely be in those types of areas.

Okay, okay. And between, as we talked about earlier, 2028, that's when we really kind of approach the kind of hockey stick with those 100 logical qubits. What's the kind of biggest risk between now and then? for the business? Is it just engineering, supply chain?

How would you frame it? You know, there could be some unforeseen engineering challenge that we just have not encountered yet that will, for some reason, delay our ability to get to 100 logical key bits. I don't see that on our roadmap at this point in time, but that certainly is always possible. There's the possibility that another modality has some kind of leapfrog moment and gets there significantly faster than we do, but for the things that are in our control, I think I feel pretty good about our ability to get to the 100 logical kubit level.

Okay, okay. Any questions out there? Please go ahead.

Folks are targeting around the same time that we are for 100 logical kubits, and so there are a variety of, so a lot of people have set that date or thereabouts as their target. I think it's probably more indicative to look at where people are today and where they've And so there are only a handful of companies that have shown that they have any logical And so we ended last year at 12, we'll be at 30 by the end of this year. The only company that has posted a higher number, I believe, is Continuum, who has a plan to get to 50 logical qubits and we've said we're going to sell a machine to the state of Illinois that will be able to do 50 logical qubits in 2027 so I feel like it's those are the you know us and Continuum are two of the companies that are in the race right now I don't think any of the superconducting companies have showcased logical qubits yet the photonic companies have not as far as I can tell and so I'd say those are the folks I'm Q has a very aggressive sort of hockey stick in their logical qubit roadmap, but I don't know if they have shown any. It's a smaller number at this point in time. So Quera has, and so they're another quantum, they're another, I was talking just about the publicly traded companies. Quera is another neutral item company that has been a leader in logical qubits. The 98 number is, I wouldn't take that one at face value, otherwise they'd be very close to be doing something useful. And so there's, so that's probably like, I'm not sure how they're defining the logical qubits there. But all that said, I have a lot of respect for the Quera team. They're a strong neutral item team.

Awesome. Well, we'd have to leave it there.

All right.

CEO of Infections. Thank you.

Thanks, everybody.