Nicholas Farina (00:00)
and audio is good.

Sebastian Hassinger (00:01)
Yeah, it sounds good.

Sebastian Hassinger (00:04)
Okay, Nick, thank you very much for joining me. ⁓ I've been looking forward to digging into what you're doing at Eureque and ⁓ the very unique modality of qubit you're working on. think your company is the only one in the world that's actually trying to commercialize your particular modality. before we get there, would you mind ⁓ starting with just your personal journey? I always find it interesting to find how people got to, you know,

Sebastian Hassinger (00:31)
⁓ to the point where they decided quantum computing is my thing.

Nicholas Farina (00:36)
First off, thanks for having me. It's a pleasure to be here. My personal journey begins with a theater company, which is, yeah, which is maybe not the most common answer that you're going to find on your show. But in 2000 and, let's see now, 2010, late 2010,

Sebastian Hassinger (00:46)
⁓ See, I told you it was unique. ⁓

Nicholas Farina (01:05)
I joined the board of directors of a theater company in Chicago. ⁓ And the boyfriend of the executive director ⁓ was doing his PhD ⁓ at Northwestern in ultra-low temperature condensed matter physics. ⁓ He was an experimentalist, is an experimentalist. This is ⁓ now Professor Johannes Palleman of Michigan State University. ⁓ And ⁓ we just became friends. ⁓

Nicholas Farina (01:34)
I thought the work that he was doing was incredibly neat. At the time, was software entrepreneur. So I was working in technology. ⁓ also, it's important to note that at that point in time, Johannes was not really focused on quantum computing. ⁓ He was generally interested in superfluids ⁓ and ⁓ condensed matter physics broadly ⁓ and was not

Nicholas Farina (02:03)
there were no early ⁓ indications that we would ever start a quantum computing company together. ⁓ I just thought the work that he did ⁓ and the science behind it was fascinating and perhaps a little bit more interesting than my day job ⁓ as a software founder. So we stayed in touch. ⁓ I won't belabor this. ⁓ We stayed in touch when ⁓ he went out and did his postdoc ⁓ at Caltech.

Sebastian Hassinger (02:14)
Hmm.

Nicholas Farina (02:31)
And it was during his postdoc at Caltech that Johannes became really interested in quantum computing. ⁓ And ⁓ this was between 2013 and 2015. So quantum computing was ⁓ quite a hot topic ⁓ at the IQIM over there at Caltech. And everyone was debating the pros and cons of different modalities. ⁓ And ⁓ then to put everything together in terms of both how I became involved

Sebastian Hassinger (02:48)
Yeah.

Nicholas Farina (03:00)
and also how ArrowQ was ⁓ conceived, Johannes and I, once he got the professorship, were just chatting about ⁓ life. And I said, what are you going to do with this professorship? And he said, well, know, ⁓ at Caltech folks were talking about the pros and cons of various different qubit platforms. ⁓ But ⁓ there's one that is aligned to my background.

Nicholas Farina (03:29)
that hasn't been fully explored, ⁓ but if brought to reality, could offer the best of all worlds. ⁓ So all-to-all connectivity, fast gate speeds, ⁓ exceptionally long coherence times, ⁓ basically everything that you might want. ⁓ And ⁓ I'm the kind of person ⁓ who is focused on trust in relationships ⁓ and is

Nicholas Farina (03:58)
have all was already captivated by the science behind it. And I'm the type of person to just jump into things. So after about a couple days ⁓ of due diligence ⁓ on ⁓ because my first question was the obvious one, which is why isn't anyone else pursuing this? If this is such a great idea? ⁓ And, ⁓ you know, isn't this something that a large company could copy in?

Sebastian Hassinger (04:17)
Right. ⁓ Yeah. ⁓

Nicholas Farina (04:25)
10 minutes ⁓ and ⁓ us lose our advantage if we were to build a business. ⁓ And I learned that neither one of those things ⁓ were really true. ⁓ I learned that it's an incredibly difficult field ⁓ working with single electrons floating on superfluid helium. ⁓ There's only about 10 to 15 people in the world who are truly experts on the system.

Nicholas Farina (04:56)
and it would take quite a long time for a large company to catch up to what doing. And the fact that there were so few people in the field answered the question as to why nobody else is doing it. Additionally, for some history on the qubit type, the history dates back to a theory paper in science from Mark Dickman, who is still a professor at Michigan State University.

Nicholas Farina (05:22)
and Phil Plattsman at Bell Labs, who has since sadly passed away. And this was in 1999. So the first proposal of electrons and helium for quantum computing goes back to 1999. ⁓ And there were initial experimental efforts. ⁓ They did not work ⁓ for various reasons. And this was in the early ⁓ 2000s. ⁓ So the reason it can work now ⁓ is ⁓ the toolkit around quantum computing. ⁓

Nicholas Farina (05:52)
the control methods we have, the equipment you can buy off the shelf. I mean, we didn't have ⁓ dry fridges then, we didn't have ⁓ CQED, which is how we control our qubits. ⁓ So without inventions like these, we would not be able to do the work. And that's in fact why it failed originally. So I said, this sounds like a good business, let's do it. Yeah. ⁓

Sebastian Hassinger (06:11)
So ⁓ the cavity, sorry, yeah, right. So, okay, there's a lot I want to unpack there, but just before I move on, the CQED, that's cavity ⁓ quantum electrodynamics circuit. Okay, okay. ⁓ So you're using the control methods that were developed around the transmon essentially.

Nicholas Farina (06:22)
⁓ Circuit quantum electrodynamics. ⁓ Yes. ⁓

Nicholas Farina (06:34)
That's exactly right, yes. ⁓ So we owe a lot to those folks. ⁓

Sebastian Hassinger (06:36)
Yeah, OK. OK, so. ⁓

Sebastian Hassinger (06:41)
Yeah. Okay. So, but the, ⁓ you, you ⁓ breeze past the, the, the fact that the Orc qubits are single electrons ⁓ and they're floating on ⁓ a liquid, ⁓ gas space. It's hydrogen, right? Or is it helium?

Nicholas Farina (06:59)
It's super fluid helium.

Sebastian Hassinger (07:01)
A helium, yes, okay. Yeah, so there's a little pool of liquid helium ⁓ and a little electron sitting on it like a water bug, basically. ⁓

Nicholas Farina (07:13)
You know, I've never heard it described that way, but that's accurate, ⁓ broadly speaking. ⁓ Fewer legs. ⁓ And ⁓ we love, of course, the fact that electrons are small because it allows us to ultimately pack up to, ⁓ just based on the density, can fit about a million plus qubits, potentially 10 million physical qubits on a

Sebastian Hassinger (07:18)
⁓ Fewer legs probably. ⁓

Nicholas Farina (07:43)
single chip. ⁓ the system doesn't have, which is one of the other benefits, it doesn't have a need for modular interconnects, ⁓ at least until you get to 10 million plus qubits. ⁓ At which point I ⁓ will be in a different position than we are today. ⁓ Slightly. That'll be a good problem to have. ⁓ So yes, ⁓ there is a CMOS, the system is also CMOS compatible.

Sebastian Hassinger (07:44)
Hmm.

Sebastian Hassinger (07:53)
Right. Right. Right.

Sebastian Hassinger (08:00)
Sure, yeah. ⁓ That would be a good problem to have.

Sebastian Hassinger (08:05)
Yeah.

Nicholas Farina (08:12)
⁓ So we start with the CMOS ⁓ produced control chip. ⁓ And this chip is the foundation ⁓ of our quantum processor ⁓ and has microchannels etched into it ⁓ for the electrons. ⁓ And then what we do is we put a layer of bulk superfluid helium at the bottom ⁓ of the cells that we encase the ⁓ chip in.

Nicholas Farina (08:41)
And then what's really nifty about superfluid helium is that it will crawl up walls as a superfluid. ⁓ So it crawls up the walls of this ⁓ copper cell we have. ⁓ And then it coats ⁓ the ⁓ CMOS produced chip with a thin layer of liquid helium. ⁓ And then ⁓ the electrons are fired using filament.

Sebastian Hassinger (08:48)
Hmm.

Nicholas Farina (09:11)
into this environment ⁓ and the electrons ⁓ are trapped naturally above the liquid helium, ⁓ a few nanometers above the liquid helium, ⁓ and are attracted to their own image ⁓ beneath the helium. ⁓ And then we control them ⁓ via the CMOS control chip is how we are able to manipulate ⁓ these electrons ⁓ and ⁓ using a CQED ⁓ toolkit.

Nicholas Farina (09:40)
So it's ⁓ a combination of multiple modalities when you kind of it up. ⁓ It looks, and I will mention one more important detail ⁓ about our journey. So we had originally been looking at using the emotional state of the electron to make our ⁓ QPU, but our processor. ⁓ Well, so we changed to spin qubits. ⁓ So that is a different

Sebastian Hassinger (09:44)
Yeah. Yeah.

Sebastian Hassinger (10:03)
Like a spin qubit.

Sebastian Hassinger (10:08)
okay. Okay. that's angular. ⁓

Nicholas Farina (10:10)
type of motion right. ⁓ No pun intended. ⁓ then in 2001, excuse me, in 2021, ⁓ Professor Steve Lyon from Princeton joined us ⁓ as our CTO. ⁓ And ⁓ Steve Lyon has been working on this system since around 2003 ⁓ and wrote ⁓ a the seminal paper on using ⁓ spin qubits

Nicholas Farina (10:40)
with electrons on helium in 2003. It was published in 2006. Steve said, look, I'd love to join you folks, but you've got to work on the spin state because there you can get coherence times that are drastically longer, 10 seconds or greater. The predictions are for even longer than that, but to be conservative, 10 plus seconds.

Sebastian Hassinger (10:42)
Okay.

Sebastian Hassinger (11:04)
Wow, so this is sort of an off the shelf number, it's 10 seconds.

Nicholas Farina (11:08)
Yes, that's a relatively conservative estimate. ⁓ yeah, it's a combination of spin qubits. ⁓ It looks also like trapped ions, ⁓ especially the microwave controlled versions.

Sebastian Hassinger (11:13)
That's amazing.

Sebastian Hassinger (11:21)
Hmm.

Sebastian Hassinger (11:25)
Right, I was going to say, ⁓ so the control of the electrons is by microwave field that you're creating ⁓ in the, as you said, the CMOS chip that is, ⁓ that's the container for the liquid helium, essentially. Huh, wow.

Nicholas Farina (11:40)
That's right. Yep. And it's all contained in a copper cell in a cryostat.

Sebastian Hassinger (11:46)
Right. And is the challenge to scaling, I imagine it's getting the microwave control fine tuned enough that you can actually accurately move around so many electrons in that space in an accurate way with good control.

Nicholas Farina (12:04)
That's right. That's a key challenge that we've worked on. We recently published a paper showing that we can control up to a million ⁓ qubits with only 50 control lines. So we are getting better ⁓ at solving this problem. ⁓ And ⁓ the nice thing about these electrons is that they are mobile, which also makes them compatible with any type of future error correction.

Sebastian Hassinger (12:05)
Yeah.

Sebastian Hassinger (12:18)
Wow. Okay.

Sebastian Hassinger (12:22)
Huh.

Nicholas Farina (12:33)
because I think the error correction that exists in ⁓ five years is going to look very different than the error correction of today. I don't think that's a very controversial statement. ⁓ If you want me to come on here and say something controversial, I'm sure I can figure something out. ⁓ I think, yeah, we'll get there. But I think saying that error correction will get better is not that controversial. ⁓

Sebastian Hassinger (12:33)
Right.

Sebastian Hassinger (12:40)
Yeah. Yeah.

Sebastian Hassinger (12:44)
No. ⁓

Sebastian Hassinger (12:50)
We'll get there. We'll get there. ⁓

Sebastian Hassinger (12:58)
No, no. And I think you're right. It's ⁓ entering this phase of co-design where the topologies of the chips are being adapted to new approaches ⁓ for lower ⁓ overhead ⁓ error correction. mean, QLDPC is sort of the case in point, right? I mean, that long distance coupling is driving ⁓ architectural decisions. And as you were saying, like the advantage for you, it's similar to neutral atoms in that

Nicholas Farina (13:14)
Exactly.

Sebastian Hassinger (13:25)
You've got this entirely plastic space. You can move your qubits around however you want to create almost a pseudo architecture within the device. That's really interesting. ⁓ How do you carry out gate operations on your qubits?

Nicholas Farina (13:44)
⁓ So gate operations, ⁓ initially we're using ⁓ the dipole-dipole interaction. ⁓ This is going to have gate speeds equivalent to ⁓ ion traps. ⁓ And then to make the gates ultimately faster, we will move to likely to exchange gates ⁓ to speed up gate operations significantly. But that's a little bit...

Nicholas Farina (14:13)
down the roadmap. Our first processor will use ⁓ the dipole-dipole to qubit gates.

Sebastian Hassinger (14:14)
Okay.

Sebastian Hassinger (14:21)
Okay. And what's an, what's an exchange gate? How does that differ from a dipole dipole?

Nicholas Farina (14:26)
⁓ So ⁓ with Todo Exchange Gates, we basically need to make everything smaller. ⁓ And these are well known ⁓ in the spin qubit world. ⁓ So that's why when I say we're sort of a mix of a bunch of ⁓ different types of modalities, ⁓ this exchange qubit concept comes from the spin qubit world. ⁓ And as long as ⁓ we're very confident that it's something we can do,

Sebastian Hassinger (14:35)
Okay.

Sebastian Hassinger (14:42)
Yeah.

Nicholas Farina (14:56)
because we essentially have to shrink everything down in size, ⁓ which we can do at a tier one foundry ⁓ as we move on.

Sebastian Hassinger (15:03)
Hmm. Okay. Gotcha. Okay. Interesting. Um, and so you mentioned, you know, uh, um, the, the processor that you're building, what, what's your current status? Do you have an operational device? Is it, you know, what's the scale that you're, you're able to produce at this point?

Nicholas Farina (15:21)
⁓ So we do not have an operational device to answer the question directly. I'm always a believer in being very ⁓ honest about where we are in our development. ⁓ But we've done a few things, thank you. ⁓ We've done a few things ⁓ along the way. ⁓ One, I have to say watch this space ⁓ because it's under embargo, but ⁓ within a couple, ⁓

Sebastian Hassinger (15:33)
Appreciate it. ⁓

Nicholas Farina (15:49)
couple of weeks or a month, ⁓ we have ⁓ an exciting paper coming out from a top-tier journal ⁓ on one of our main achievements ⁓ in progress. ⁓ One thing that I can talk about openly ⁓ is we have demonstrated the largest CMOS-based architectural ⁓ of a chip, where we were able to run ⁓ essentially a mock surface code on that.

Sebastian Hassinger (15:56)
Great.

Sebastian Hassinger (16:10)
Hmm.

Nicholas Farina (16:15)
and we can control 2000, it's an ensemble, 2432 electrons and down to packets of one in a CCD-like structure.

Sebastian Hassinger (16:15)
Hmm.

Sebastian Hassinger (16:20)
Okay.

Sebastian Hassinger (16:30)
Hmm, interesting. So you're saying as an ensemble, so you're not addressing them as individual qubits. Right, so this is an engineering milestone is what, is that the right way to characterize it as sort of an engineering milestone in terms of standing up your control systems in that architecture? Okay.

Nicholas Farina (16:37)
Yes, these are not 2,400 qubits, but they're future qubits.

Nicholas Farina (16:51)
Yes, ⁓ that's a ⁓ fair way to characterize it. Our belief, which is a little contrarian, ⁓ so now I will say something a little bit interesting, ⁓ is that scaling is actually the hardest part of building a quantum computer. ⁓ And that if you look at the last 10 years, ⁓ we've made tremendous progress in qubit quality ⁓ in terms of ⁓

Sebastian Hassinger (16:55)
Yeah.

Sebastian Hassinger (17:02)
Good. ⁓

Nicholas Farina (17:20)
Gate Fidelities are a leading indicator, but of all aspects, almost of qubit quality. And we've made great strides in error correction and in more efficient algorithms. But one area that we haven't really made a lot of progress in over the last 10 years has been scaling fast enough to get to the hundreds of thousands and eventually millions of qubits that we'll need.

Nicholas Farina (17:50)
for really interesting applications. So the way that we tackled building a quantum computer was instead of ⁓ starting off with the two-qubit gate ⁓ and sort of hoping and praying to find ways to scale from there, we started out ⁓ with a, ⁓ in reverse, with ⁓ what does a million-qubit ⁓ quantum processor look like ⁓ and how might one build that? ⁓

Sebastian Hassinger (17:52)
Yeah.

Sebastian Hassinger (18:19)
Mm.

Nicholas Farina (18:20)
That is how we came to the belief that you need to use ⁓ CMOS ⁓ because that's the only way humanity has ever shown we can build that many features. ⁓ Yes, exactly. That's the only, right. So if you're able to use that, then why not use that? If that technology has been invented at the cost of over a trillion dollars over 40 years, then why not use that technology? ⁓ So yeah.

Sebastian Hassinger (18:29)
Right. We can build trillions of things ⁓ per second. ⁓

Sebastian Hassinger (18:38)
Yeah.

Sebastian Hassinger (18:41)
Right.

Sebastian Hassinger (18:47)
Right, right. ⁓ And how would you compare your approach at EuroQ to ⁓ a standard spin qubit? Pick your favorite spin qubit, because that's also leveraging CMOS. It's also doing either ⁓ single transistor or single electron or very small number of electrons per transistor, which is actually a qubit. ⁓ it sounds like there's a fair number of similarities. What would you say is ⁓ the biggest differences?

Nicholas Farina (19:16)
⁓ Thank you for giving me ⁓ a fairly straightforward question. ⁓ There's ⁓ one key difference. ⁓ Silicon is and likely, mean, material science is an evolving field, ⁓ but silicon is an imperfect material. ⁓ And this causes rise to ⁓ trap charges, ⁓ valley splitting, ⁓ and a number of issues. ⁓ Additionally,

Sebastian Hassinger (19:34)
Right.

Nicholas Farina (19:44)
with spin qubits, ⁓ which ⁓ we have infinite respect for. ⁓ So this is nothing against the world of spin qubits. ⁓ This is purely the engineering challenge with them, ⁓ is that your qubits are stuck inside of the silicon. So they are not mobile. ⁓ And what we do differently ⁓ is ⁓ adding this layer of liquid helium, which ⁓ may sound exotic or difficult, but it's actually quite easy.

Sebastian Hassinger (19:49)
Sure.

Sebastian Hassinger (20:00)
Mm-hmm.

Sebastian Hassinger (20:09)
Mm-hmm.

Nicholas Farina (20:14)
and was first done in the ⁓ 1960s. ⁓ this is a, particular system, ⁓ electrons on helium is very well understood. ⁓ So adding this layer of ⁓ superfluid helium, just think of it as a protective barrier that takes the electron ⁓ spin qubit out ⁓ of the silicon ⁓ and into ⁓ a perfect environment that's free of

Sebastian Hassinger (20:32)
Right, right.

Sebastian Hassinger (20:42)
Hmm.

Nicholas Farina (20:44)
disorder or defect.

Sebastian Hassinger (20:46)
Right, right. Yeah, and when you were describing the sort of the physical ⁓ manifestation of it, did occur to me that in my mind, I was sort of imagining a bathtub full of liquid helium with electrons floating on top. ⁓ Yeah, but the way you described it sounded much more like a protective film. ⁓ And I get, mean, the superfluids are really cool. And that's a really interesting, you know. ⁓

Nicholas Farina (20:56)
Yes. I always like clarifying what it is. ⁓

Sebastian Hassinger (21:12)
benefit of the superfluid behavior that it actually climbs up the walls of that chamber and coats all the surfaces evenly. That's really fascinating.

Nicholas Farina (21:19)
Yes, ⁓ my favorite fun fact is if you put superfluid helium in a glass and stirred it, it would keep stirring forever. ⁓ It would keep moving. It would never ⁓ ever stop. ⁓ Yes, exactly. ⁓ You've invented your own one. ⁓

Sebastian Hassinger (21:34)
perpetual motion machine. ⁓ It is real. ⁓

Sebastian Hassinger (21:40)
Yeah, nice. ⁓ So, OK, so let's see. ⁓ The operating environment of the chip itself, does that also need to be, I imagine, since you've got liquid helium inside the chip, you probably need to be in ⁓ a dilution refrigerator as well.

Nicholas Farina (21:59)
That's right. ⁓ We ⁓ currently operate at around 10 millikelvin. ⁓ We put out a paper in Physical Review X recently ⁓ that showed that we can control and sense ⁓ electrons on helium ⁓ at 1 Kelvin, which is considerably warmer ⁓ and could use smaller ⁓ refrigerators. ⁓

Sebastian Hassinger (22:04)
Okay.

Sebastian Hassinger (22:17)
Hmm.

Sebastian Hassinger (22:24)
Right. Right. ⁓ Compared to superconducting which need millikelvin ranges. Yeah.

Nicholas Farina (22:30)
Right. And to be very clear, we currently do and likely will for the foreseeable future, but we have opened the door with this paper ⁓ to potentially ⁓ having a future of operating at higher temperatures. Hot cubit exactly. Which ironically will help with heat dissipation. ⁓ but yes, ⁓ that was a ⁓ work that we were proud of last year as well.

Sebastian Hassinger (22:44)
of hot cubits. ⁓

Sebastian Hassinger (22:51)
Hmm. Hmm.

Nicholas Farina (22:58)
Last year was after, so the company is about a decade old and last year was our highest velocity year, I would say, in terms of things all coming together.

Sebastian Hassinger (23:02)
Hmm.

Sebastian Hassinger (23:06)
That's great. That's great. And I suppose, you ⁓ you were ⁓ an early sort of ⁓ participant in the formation of the community that's emerged out of the Chicago Quantum Exchange and the five universities that ⁓ joined that initially in the two national labs and all of the regional focus on ⁓ quantum as an economic development.

Sebastian Hassinger (23:32)
topic that ⁓ seems like your accused really benefited from that set of resources and the community that's built up around you.

Nicholas Farina (23:39)
⁓ Absolutely, we have. decided to make Chicago was an intentional decision. ⁓ We had a national search for our headquarters. ⁓ And ⁓ because of the time we were in this was back in 2021, we had folks in New York, Princeton and Michigan ⁓ and ⁓ wanted to ⁓ have a blank slate and find the best place in the country. That was ⁓ convenient.

Nicholas Farina (24:09)
to ⁓ build our business. And this was even before some of the ⁓ new and frankly sweeping initiatives that the state of Illinois has taken like the quantum and microelectronics park ⁓ and the 500 million plus that the ⁓ governor JB Pritzker has put into quantum technologies. ⁓ And there's, as you mentioned, the Chicago Quantum Exchange.

Sebastian Hassinger (24:34)
Yeah.

Nicholas Farina (24:38)
⁓ There's a group called P33, which is ⁓ in World Business Chicago ⁓ and Illinois Economic Development ⁓ Coalition. That's a rebranded Intersect Illinois, so I hope I got that one right. ⁓ But all of these are incredible resources, ⁓ and we found that between the local talent ⁓ and the coordination between all these different groups that are happy to provide resources, we found

Sebastian Hassinger (24:41)
Right.

Sebastian Hassinger (24:51)
Hmm.

Nicholas Farina (25:08)
that it was the best place. And I think that's borne out to be true because now there's been continued investment. The state of Illinois is actually an investor in ArrowQ directly. So they are on our cap table ⁓ and indirectly ⁓ by ⁓ providing, you know, their shared services here and whatnot. So, yes, I ⁓ am not being compensated to say this, but I really do think Chicago is the best place to build a quantum technology startup.

Sebastian Hassinger (25:15)
Yeah.

Sebastian Hassinger (25:20)
great.

Sebastian Hassinger (25:38)
That's great. That's true. We just had Harley on the show ⁓ last week actually was the episode was posted and that was a great conversation too. I've been ⁓ involved, ⁓ you know, ⁓ since I was on the IBM side of the table with IBM joining CQE very early on. I've been ⁓ watching closely and it really is, it's a model for, for the other regions that aspire to build a quantum ecosystem in many ways. So

Sebastian Hassinger (26:04)
You're on the path, you had this high velocity year in 2025, you're on this path to developing a full, functional system. Do you have a sense of how many qubits you want to ship in that first system that outsiders get to kick the tires on?

Nicholas Farina (26:14)
system. ⁓ Do you have a sense of how many?

Nicholas Farina (26:24)
About 10,000 will be our debut device ⁓ is what we're shooting for. ⁓ And that's only made possible because of CMOS. ⁓ And because of the work that we've already done in scaling. So ⁓ once we have perfected our two qubit gates, ⁓ we are able to, in principle, drag and drop them onto the...

Sebastian Hassinger (26:26)
Wow. Okay.

Sebastian Hassinger (26:36)
Yeah.

Nicholas Farina (26:50)
chip that we've already built, which is called Wonder Lake, ⁓ the one that we can control the 2400 electrons on currently. ⁓ we're able to use that chip already has operation zones built in ⁓ for gate operations in the future. So we take our two qubit gates, we put them ⁓ onto that chip, ⁓ and ⁓ then we have a quantum processor. ⁓ And with CMOS can scale quickly.

Sebastian Hassinger (26:52)
Mmm.

Sebastian Hassinger (27:04)
Okay.

Sebastian Hassinger (27:18)
So when you say two cubic gates, ⁓ put the cube. So the two cubic gates that you're working on, that's almost essentially like firmware, it be the control sequence required to carry out a particular two cubic gate. Is that, is that the right way to think of it?

Nicholas Farina (27:32)
⁓ That's the right way to think about it, yes. ⁓ But for us, we're ⁓ still ⁓ at the level of, ⁓ okay, we've proven out that we can work with electrons on helium ⁓ using CMOS, that we can have exquisite control over them down to the single electron ⁓ level. ⁓ Now, working on building ⁓ two cubic gates that will then

Sebastian Hassinger (27:35)
Yeah, interesting.

Nicholas Farina (28:00)
slot into the architecture that we have already built, ⁓ if that makes sense. ⁓ So going in reverse.

Sebastian Hassinger (28:03)
Right. Yeah. Do you have sort of a public roadmap? Do you have a date that you have declared you're shooting for? ⁓

Nicholas Farina (28:11)
We're shooting for late 2028, ⁓ which ⁓ can be breaking news for your ⁓ podcast here. ⁓ We do not have a public roadmap. We, of course, have a private one. just because we ⁓ like to under promise and over deliver, ⁓ and ⁓ putting up a public roadmap, it does hold you accountable, which is nice. ⁓ But ⁓ on the other hand, we want...

Sebastian Hassinger (28:18)
Excellent. ⁓ Excellent.

Sebastian Hassinger (28:31)
smart. ⁓

Sebastian Hassinger (28:38)
Yeah.

Nicholas Farina (28:41)
⁓ to deliver something really cool ⁓ to the community ⁓ without hyping up too much beforehand. ⁓ But we are very excited about what we're doing ⁓ and think it could be the best of all worlds.

Sebastian Hassinger (28:43)
Yeah.

Sebastian Hassinger (28:48)
Do you?

Sebastian Hassinger (28:53)
10,000 cubits in 2028 would be very, very exciting. Do you have a sense for what error corrections codes you'd be sort of shipping with that device?

Nicholas Farina (29:05)
⁓ Yes. ⁓ And ⁓ to be clear, ⁓ in the end of 2028, we'd be taping out our 10,000 qubits. So it wouldn't be on the cloud. We would not be cloud ready by the end of 2028, but we would ⁓ be at the tape out point of it and have everything designed. ⁓ And then ⁓ in terms of error correction, ⁓ well, I mean, I'm a big fan for obvious reasons of ⁓ quantum LDPC ⁓ codes.

Sebastian Hassinger (29:15)
Okay. Okay. Right, right, right.

Sebastian Hassinger (29:34)
Mm-hmm.

Nicholas Farina (29:35)
⁓ And if I had to guess, if I had to bet, would say that a variation on those ⁓ would be what we will end up using. ⁓ we have all-to-all connectivity, so are compatible with them. ⁓ But I don't know what the future holds for error correction, ⁓ which is why we're so ⁓ proud ⁓ that the system can work with any different type of error correction.

Sebastian Hassinger (29:49)
Right. Right.

Sebastian Hassinger (30:01)
Yeah. Yeah.

Nicholas Farina (30:02)
⁓ because there will be new and improved ones in the future.

Sebastian Hassinger (30:06)
Totally agree, totally agree. And sorry, you did touch on this, but remind me, what would it take for you to distribute across, like a workload across multiple 10,000 qubit chips?

Nicholas Farina (30:20)
⁓ So each, ⁓ that's a great question. mean, each ship ⁓ will serve ⁓ independent, can work independently. ⁓ We're building as far up as the control stack. ⁓ So our go-to-market strategy, which is related to that question, ⁓ is ⁓ to get up. ⁓ Let me take a step back actually. You know, I think the way that the ecosystem is developing is really interesting.

Nicholas Farina (30:49)
and ⁓ watching new companies form or watching new results come out of academia. ⁓ And ⁓ that ⁓ may ⁓ end up deciding ⁓ how far up the stack we ultimately go ⁓ based upon who we might be able to partner with ⁓ at various levels of the stack. So that is TBD, ⁓ but that's generally the right way to think about it.

Sebastian Hassinger (31:03)
Right.

Sebastian Hassinger (31:09)
Gotcha.

Sebastian Hassinger (31:14)
Yeah.

Nicholas Farina (31:19)
⁓ and ⁓ we ⁓ work up to the control layer right now.

Sebastian Hassinger (31:26)
Okay. And that control layer stops at a single chip so that you don't have a notion of chip to chip, just, you know, links ⁓ to fly qubits across from one chip to another.

Nicholas Farina (31:40)
Ultimately, ⁓ we look to do everything on a single chip, ⁓ on ⁓ a ⁓ monolithic design. Yes, exactly. And I think that's one of the key advantages of the system, ⁓ is ⁓ that we can get millions and millions of qubits on ⁓ one single chip. So for on-prem installations, which ⁓ have become more popular, ⁓

Sebastian Hassinger (31:44)
Hmm. right. You're saying you just, you'll just expand the density instead of trying to scale out. Okay. Yeah, that makes sense.

Sebastian Hassinger (32:00)
Yeah. Right.

Nicholas Farina (32:08)
Of course, the cloud, think, will always probably be the leader in how folks access quantum computers. But we certainly have had folks ask about on-prem installation. ⁓ And for that, ⁓ we have a key advantage by, you know, just need, you do need a refrigerator. ⁓ You need a fridge. But as long as you have a fridge, you just need one. And ⁓ you need one chip in that fridge. And ⁓ the helium

Sebastian Hassinger (32:32)
Hmm.

Sebastian Hassinger (32:35)
Yeah.

Nicholas Farina (32:38)
⁓ sort of recirculates itself. So it's not like you need someone consistently adding more ⁓ helium into your system. So it could be offered. It's not like you need an AeroQ specialist on ⁓ premise all the time.

Sebastian Hassinger (32:43)
Right. ⁓ Topping it up. Not like the wet fridges.

Sebastian Hassinger (32:50)
Right.

Sebastian Hassinger (32:53)
Interesting. That's really cool. Well, so you've got this paper that's embargoed right now that's in a peer-reviewed journal. coming out within, you said, a month. Great. Excellent.

Nicholas Farina (33:01)
within about a month. ⁓ So we'll be excited for this ⁓ and we'll be able to share more about the system and our achievements.

Sebastian Hassinger (33:12)
Fantastic. Well, it's really, really interesting to hear about Nick. It's like I said, I've been dying to hear more about this modality because it's, ⁓ think you're the only company that I know ⁓ that is ⁓ the sole developer of a modality. Everybody else has competition within a modality. So it's a really unique position. It's an interesting bet. It makes a lot of sense. ⁓ And I'm very much looking forward to that 10,000 qubit machine sometime after 2028.

Nicholas Farina (33:19)
Ha

Nicholas Farina (33:39)
That makes two of us, Sebastian. So ⁓ yes, I very much am too.

Sebastian Hassinger (33:46)
Thank you so much for joining me.

Nicholas Farina (33:47)
⁓ Thank you so much for having me. I really appreciate it.