Sebastian Hassinger: 00:31

Welcome to the new quantum era, the podcast where we explore breakthrough technologies and the people building the quantum future. I'm Sebastian Hasker, and in this episode, I'm joined by Thomas Monz, CEO and founder of AQT or Alpine Quantum Technologies. Thomas has been at the cutting edge of quantum computing from the very beginning of trapped ion technology with deep roots in both academic research and real world technology commercialization. AQT has just made their ion trap computer, IBEX, available on Amazon Braket, the quantum computing service from AWS. Now as I've talked about before, I am employed by Amazon Web Services and work in the quantum computing space, but I'm not a member of the Braket service team.

Sebastian Hassinger: 01:14

We recorded this episode a few weeks back in anticipation of this launch, and the Bracket team is using some video clips of the episode in its blog. But this is a regular podcast episode in that there's no editorial control other than my own, and the conversation is guided by my own personal interest in these topics. So you will hear in this conversation, Thomas traces his journey from his early days working on quantum networks and high precision measurements in Innsbruck to cofounding AQT to bring robust ion trap quantum computers out of the lab and into both supercomputing centers and onto the cloud. We dive into how AQT bridges fundamental quantum science and practical engineering, what it's like to move quantum systems from research prototypes to commercial devices, and the vital role of error correction and system integration in the quantum stack. We'll cover AQT's new deployments in European supercomputing centers, upcoming availability on Amazon Braket, and the next steps for scaling up towards truly useful quantum computing.

Sebastian Hassinger: 02:14

Whether you're a quantum enthusiast, developer, policy watcher, you'll find plenty of insights in Thomas' perspective on collaboration standards and what's just over the horizon for this revolutionary technology. Great. Thank you very much for joining me, Thomas. I'm very much looking forward to this conversation. Your work with AQT goes back to the very beginnings of trapped ions as as a a quantum technology.

Sebastian Hassinger: 03:09

If you could start just by introducing yourself and and talking a little bit about those those early days, your your sort of transition or journey into the quantum technology industry?

Thomas Monz: 03:21

So I started my my master's thesis in Innsbruck, initially working on quantum networks. So eventually, it's like, how do I extract a photon from an ion? So you can think of it like a network card for an ion trap quantum computer. That was my master's thesis.

Sebastian Hassinger: 03:39

Interesting.

Thomas Monz: 03:40

And then I did that roughly also the first year of my PhD. But then given the setup of the different groups, I decided to switch over to a different experiment. So then my next year was actually metrology, which means high precision measurements on a particular transition in in in calcium ions. And then with that background, I then went into quantum info working on the control of trapped where I then finished my PhD. Okay.

Thomas Monz: 04:10

And I think one part there was is, okay. I did a little bit of post hoc time, but afterwards, I was not sure which direction to head. I got an offer from a Scottish laser company. So I then worked for a Scottish laser company for some time, Like, learning, like, the sales aspect, customers, campaigns for for conferences, how to organize that. Learned a lot from from our sales guy there.

Thomas Monz: 04:36

And then in some sense, the topic came up. Look. In Innsbruck, we we've had iron traps. And and, repetitively, our assistants became professors, moved to other locations, and they always said, it would be cool. It would save me so much time if I could get the trap from you.

Thomas Monz: 04:55

And so we thought, okay. The university is not allowed to act commercially. And

Sebastian Hassinger: 05:00

then Right.

Thomas Monz: 05:01

The the question was in the room, can we potentially turn it into a business? Thomas, do you want to turn it into a business? And that was more or less the stepping stone for AQD. And then we said, like, Traps is too small. We have to think bigger.

Thomas Monz: 05:14

And that's where one thing led to another.

Sebastian Hassinger: 05:18

So interesting. So so initially, you were thinking you were gonna be almost a a vendor, a component supplier to experimental physics labs that that wanted to do ion trap experiments. That's really interesting. And so so computing is sort of building a system out of the the component that you were actually going to to bring to market. That's really fascinating.

Sebastian Hassinger: 05:39

And were you when you were doing your PhD, obviously, you know, your cofounders are Peter Zoeller and Rayner Blatt. So Zoeller and Sourac were you know, are very well known for writing that first paper that sort of hypothesized using tri trapped ions as qubits, and then Monroe and Wineland, experimentally realizing that at at NIST. Was that something that was in the back of your mind sort of the, the the the the origin story of of trapped ions as qubits? Was that something you had been involved in at all?

Thomas Monz: 06:17

Well, that was that was actually my PhD thesis.

Sebastian Hassinger: 06:19

So to

Thomas Monz: 06:20

give you a feeling there, the the first ideas from from from Ignacio and from Peter about the so called tirac solargate, We implemented that, and then one of the first projects of my PhD thesis was how can you turn that into a three qubit, the so called toffinate. And the extension towards three qubits was part of my PhD thesis where I say, okay. This is working, but it's not working as good as it should. So we then switched to and so on

Sebastian Hassinger: 06:49

and

Thomas Monz: 06:49

so forth. And then, for instance, just to to give you a feeling, it's like I I did short algorithm to better understand what goes wrong. So this metrology background was always there in the sense of whenever I do something, can I write down an error budget to understand what were the shortcomings and how do we overcome these shortcomings?

Sebastian Hassinger: 07:10

Interesting.

Thomas Monz: 07:11

I think this this this mindset was helping a lot to push the quantum computing forward by always saying and I think that's the neat thing with ions is to say it's very pure physics. It's like AMO. It's an ion in vacuum.

Sebastian Hassinger: 07:25

Right.

Thomas Monz: 07:25

Whatever goes wrong, we can figure out almost on the mass level with some characterization. So let's understand what goes wrong. Let's fix it with new ideas, and then we should be able to lower the error rates every couple of years. And I think the ION community together with Continuum and others, we nicely successfully demonstrated that year and year again.

Sebastian Hassinger: 07:47

Yeah. Yeah. It's impressive. That's interesting. So so, you know, your your academic your your PhD thesis was directly involved in the the the beginnings of trapped ions as as quantum computing.

Sebastian Hassinger: 07:59

And I think you were you were involved in the the first controlled knot gate operation. Mhmm. Was it no. That wasn't

Thomas Monz: 08:07

So this was Malk and others that so I I learned from them. Mhmm. Eventually, I would say that it's, like, what their shortcomings were. So, why is it only I think in those days, we had, like, 91%, like, error rate of 9%. Nowadays, we are at, like, per mill level.

Thomas Monz: 08:26

Right. So how the heck can I present? What's wrong there? And then it's like, it's this it's this. We have to fix the beef.

Thomas Monz: 08:32

We have to get the laser better.

Sebastian Hassinger: 08:34

It's so fascinating. I mean, what you're describing, you know, this for the error budget and the and the incremental and and continual improvement of, the fidelity of ions as as a quantum computing platform. It's such an interesting mix of engineering and science. It's sort of deeper understanding of the scientific principles at at work and then applying engineering principles sort of for for the greater precision and and repeatability of the of the operations. It's I I I I've always struck by how tightly coupled science and engineering are in the field of quantum technology.

Sebastian Hassinger: 09:14

Is that is that sort of how you see that as well?

Thomas Monz: 09:18

It's it's very much how I experienced it. To give you a feeling, when I started, we still had a couple of like, to control laser light fields, you have these AOMs. I still were using some of the old ones because we didn't have so much money in the early days. We had them torn out of old laser printers and just mounted them on the because that's why you got AOMs for free. And we mounted them on the tables and used them.

Thomas Monz: 09:47

And I remember that back in those days, we bought the first optical laser. And and as as you can imagine, it's like up to that point, we mainly made them on our own. It's like a Right. Master's thesis would be to build your own laser. And it's like, but, you know, we know how to build it on our own.

Thomas Monz: 10:04

Why should we give a company money? And then we bought the first one, and then it was, like, so much more reliable, a good manual.

Sebastian Hassinger: 10:12

Right.

Thomas Monz: 10:12

There were improvements that we really said, no. No. It makes sense to outsource some of these engineering so that we can focus on the physics again. And so that would be root for circle where you say, it's like, let's buy something so I can focus on the physics again. And you see a little bit of the connection for AQT there where we say it's like, we have these components.

Thomas Monz: 10:33

Like, we can we can accelerate ion physics by giving people ion traps that work, but we can also accelerate quantum information processing by giving people access to good quantum computers.

Sebastian Hassinger: 10:47

Right.

Thomas Monz: 10:48

To say it's like you don't have to fiddle around with noise or something special because it just go out of the box.

Sebastian Hassinger: 10:54

That's really interesting. Yeah. It it's it's always struck me that HGT is very still very rooted in the fundamental science. I mean, you're you're you're at Innsbruck. You've got very tight relationships with Rayner Blatt with Peter Zola and other active researchers and their groups.

Sebastian Hassinger: 11:15

And it you know, I think even recently, you did some work with a the what was that? Complete quantum field theory in two dimensional it was a simulation of of of particle fill of or an element of particle physics. Is that right?

Thomas Monz: 11:34

Yep. Yeah. Well, that's the cool thing. It's it's it's we we have let's put it that way. From a quantum algorithms point of view, there is only a selective small number of algorithms where we really have an exponential benefit from using quantum filters.

Thomas Monz: 11:52

And so I think we need to support thinking outside the box. And if you think about, like, HPC is like the idea is quantum computers will help in some fields of high performance computing devices.

Sebastian Hassinger: 12:04

Right.

Thomas Monz: 12:04

What are these fields which they are currently used for? Well, many things where quantum plays a role in the first place. It's like chemistry, material science, but also high energy physics. And so if we think about quantum chromodynamics and so, why not think about quantum devices to use and investigate these problems? And interestingly, it's like it's it's not how to say?

Thomas Monz: 12:25

It's not completely, like, clear which path will lead forwards in a in a good Right. And and you could say some of the work, why not go beyond two levels? Think about simulations where I use that. It's it's not a bug. It's a feature.

Thomas Monz: 12:40

I can directly more nicely map the underlying physics on a controllable device and then push the field forward. Interesting.

Sebastian Hassinger: 12:50

And so, I mean, you know, I always remark on on how incredible Feynman's statement in 1981 was that, you know, if if you wanna simulate nature, you're nature is not classical. You're going to need a quantum computer. So that's that's where you're seeing sort of the these devices are quantum many body, you know, by by their nature, and therefore, you should be able to map quantum problems onto them. Just makes intuitive sense. And that's certainly as you said, there's only Shores and HHL and a few other Grovers and a few other, isolated algorithms we were aware of that may have sort of abstract applicability.

Sebastian Hassinger: 13:34

But, but quantum simulation, we know for certain they're they're better at that than classical computing. And and you just mentioned HPC. That's been a major focus for AQT for quite some time, sort of defining the place for quantum computers in the HPC, the broader classical HPC ecosystem. Is that what motivated sort of the relationship with LRZ and and other supercomputing kind of centers?

Thomas Monz: 14:02

So for me, it was actually a lit well, went along those lines. It's it's because of the proximity to University of Innsbruck. When we started to think about the quantum computer, I went to the IT department and I said, can I please have a look at the high performance facilities? And they were like, sure. But what do you want to see?

Thomas Monz: 14:23

And I was like, I I don't know yet. I want to get a feeling for the room. And so I went in there, and all I noticed was, like, many different colors and everything blinking, but everything was constantly 19 inch boxes. And so I said, if we build something, we have to build something which is compatible with this 19 inch rack standard because this is the infrastructure that customers are used to.

Sebastian Hassinger: 14:48

Right.

Thomas Monz: 14:49

I I can't just change, let's say, the the power block and say, I go for 287 volts. No. It's like we have two twenty and so on. And so for us, it was like, okay. Let's build it in these 19 interacts.

Thomas Monz: 15:05

The very moment we had that, Europe started to say is, like, if we want to use quantum computers, how could we where would we use them? It's like and and probably, it's it's about computing. We need to bring it closer to computer scientists. What do computer scientists ask for? They usually say, I want to have access to high performance computing.

Thomas Monz: 15:26

And this could be yes. There are facilities from Amazon where I can get access to to AWS and so on, But some of them are more on-site. And so for us, there was really, a way of saying, okay. If this is what people are used to, lower the entrance barrier. If they can access HPC, ideally, you can envision it.

Thomas Monz: 15:48

Like, how do they don't want to know. Think about your laptop. It's like you you might not care. Is it an inkling, IMD, whatever chip is in there? You could say, I want to run something on a HPC facility but I offload some task to a quantum accelerator in one corner of the HPC center.

Thomas Monz: 16:06

I don't care as long as it's faster and better. Right. And that was this idea of let's bring it into HPC because this hybrid infrastructure is something that the customers want. They don't want to know the details that should be Right. About the API.

Thomas Monz: 16:19

And the library takes care of the different hardware, but it should just work.

Sebastian Hassinger: 16:23

Right. I mean, that that's such a critical stage in the transition from scientific research to deployed technology is that abstraction and allowing the end user to use the device for a purpose of their own without caring about, you know, you know, tuning the lasers and the the optical table and all the rest of the the idiosyncrasies that are required for the scientific research. I mean, that that level of control, that sort of intimate knowledge of how it's operating is exactly the focus of scientific research, and it it doesn't really have a place in technology. So so so was the was the deployment at, of the first, device, I think the form factor is PIE is your acronym, right, for the the form factor. Was that at University of Innsbruck or LRC?

Thomas Monz: 17:12

So the first system was deployed as, we like, historically, it's like we set some up in the very first office, which was, like, on university grounds. And the first thing we learned about stability is when we moved to this office, which I'm currently in, which forced us to move it a couple of 100 meters, but doesn't matter whether it's a 100 meters or a 100 kilometers because we had to, like, put this box into a truck, move it across the building. Right. Get it out. Get it down the stairwell.

Thomas Monz: 17:39

And so the first time we did it with a customer, as you said, was in Munich.

Sebastian Hassinger: 17:44

Ah.

Thomas Monz: 17:45

Okay. And that went really, really smooth. And then very swiftly afterwards, we actually got the second system installed in Botsna in Poland.

Sebastian Hassinger: 17:54

And we

Thomas Monz: 17:54

just built a system, get it on a truck, drove to Poland, and that's it.

Sebastian Hassinger: 18:00

And that Poznan, that's the EuroHPC center Yeah. Which is part of the EU's effort to deploy, I think, six or so different devices in different your HPC centers. Right?

Thomas Monz: 18:12

I think it's by now, it's 12 because the problem is hope. So initially, it was more quantum computing focused, but now in on purpose also includes quantum simulators. Okay. It makes sense because you could say I'm I'm a computer scientist. There are different devices, different ways of controlling them depending on which problem I have, what's the ideal, how to say, solution, chip, or platform to solve it.

Thomas Monz: 18:41

And for some of them, we'd say maybe a digital device is not the right way forward. Maybe we want to do boson sampling or something.

Sebastian Hassinger: 18:48

Right.

Thomas Monz: 18:48

So if we have access to more, it's it's only benefiting, in that case, the European customer base.

Sebastian Hassinger: 18:56

Yeah. Yeah. So so the the the deployment at at LRZ in in Germany, that's part of the Munich Quantum Valley, which involves TU Munich. And there's been quite a bit of work in looking at at sort of what are the challenges and how can they be systematized for for integrating quantum computing with with classical HPC. Was that something that AQT was involved in as well?

Thomas Monz: 19:26

So partially, it's like we we we said we want to focus on on the computer per se. The hardware inside, we want to make the quantum computer work, and then we'll happily go up to the level of the API.

Sebastian Hassinger: 19:41

Mhmm.

Thomas Monz: 19:41

And then it depends a lot. That's also what we learned from discussing with HPC people, which I think many many many people in the HPC domain are aware of. Like, they're everyone has his own or her own ideas on how to do a scheduling task, like how to distribute the workload. And so we noticed it will be hard for us to say it's like, this is the right way forward because they have already pretty good ideas on what they think is best given their infrastructure. And so with that, we'll go up to the level of the API.

Thomas Monz: 20:15

We'll help you with integrating at that level how you want to integrate it in detail. That's up something that we leave up to you. For instance, Martin Schulz and Robert Wille are doing really, really great work there on how to integrate, like, different platforms in a heterogeneous structure into HPC with quantum. And I'd say it's like, if if you want to extend the podcast, robot would be really cool, like, and partner to having this discussion on how to bring that on board.

Sebastian Hassinger: 20:48

Yeah. Absolutely. Absolutely. And so, I mean, that's a really interesting sort of, as you said, less many, many, powerful lessons, can be drawn from all those experiences in deploying and integrating quantum computers with classical computers. How do you see you know, now you're bringing AQT to the Amazon Braket service on AWS.

Sebastian Hassinger: 21:15

How do you see the lessons that you've learned with on premise deployments in relation to sort of deployment in in, you know, the hyperscaling cloud?

Thomas Monz: 21:25

So for us, I think the having this API and saying this is how far we need to go and being willing to say, like, everything beyond we leave to the partner is very much similar to what we have to do as we work with Bracket. So it's like, here is the handover point. You take care of your part. You take care of our part. Many it helped us a lot to have, say, on the API level, like, numbers, characteristic available so that, let's say, in that case, bracket can extract the meaningful information that was already there.

Thomas Monz: 22:01

Like, all the monitoring capabilities have already been developed in the context of HPC and are now Right. That for bracket. So the overhead was notably how to say? I wouldn't say it was nothing, but there was a reasonable good foundation to build on top.

Sebastian Hassinger: 22:19

That's excellent. That's excellent. And are there opportunities for collaboration or experimentation that you think, you know, a global cloud provides that that are different from those those on premise deployments?

Thomas Monz: 22:36

I think the the cloud has a notably broader width of customer requests. You know? It's like it could be commercial customers that say, I want to have a special feature, and this is, like, a a very technical feature, but it's it's very well founded on economic reasons. And you could have a very cool, interesting idea from scientific people, but more along the lines, I I want to try something out and having this feature would make my life easier. And so we we in I'd say in HPC, the focus is notably tends to be from what we saw so far, notably more scientific and academic

Sebastian Hassinger: 23:20

Right.

Thomas Monz: 23:20

Right. Which is fine. I think we will learn a lot about the more commercial, how to say, ways people think about quantum

Sebastian Hassinger: 23:30

Right.

Thomas Monz: 23:30

It's not about how they use it, but it's also what the expectations are on what it should do or what information should be available or how it should react, which is a little bit different, and and we're happy to learn.

Sebastian Hassinger: 23:45

Interesting. Yeah. That I mean, that makes sense because supercomputing centers tend to be in The US. Of course, they're they're typically operated by Department of Energy, for example, and they're they're primarily focused on scientific computing as their their main area of of activity. So that makes sense.

Sebastian Hassinger: 24:05

And and the cloud is clearly motivated by you know, that's where businesses run their IT is is in the cloud. So it's much more commercial in nature. Do you foresee sort of the the mic you know, the the presence on on the global cloud being something that helps accelerate your plans for, you know, your road map towards fault tolerant quantum computing and and scalability? Is that it's sort of the the agenda at this point?

Thomas Monz: 24:37

So it's the I mean, the revenue for sure will help. Yeah. No. But I think one one one technical aspect really is end users is like they want to benefit from error correction, but they don't want to do error correction. Mhmm.

Thomas Monz: 24:55

And I think in that sense, it will be interesting to talk with various parties to say, is there someone who says is, like, here is this let's call it three qubit problem, but how do I map the three qubit problem onto a three logical qubit problem with encoding, decoding, and so on? Right. Is that gonna be a service that's, like, maybe part of Brexit, maybe part of a third party? How do we make sure that these solutions like, say, how do we integrate them in this entire chain? So then it's let's say, if you would be the end customer, you say, I don't care what you have to do.

Thomas Monz: 25:31

Just do this thing. And at the end of the day, don't tell me that you had 300 qubits for your logical solution because I gave you a three qubit task. I want the three qubit answer.

Sebastian Hassinger: 25:43

Right. Yeah. Yeah.

Thomas Monz: 25:46

And in that sense, I think broadening also with our services, broadening the playground for, for instance, software developers, like quantum software, quantum service developers, and saying there is a there is an additional system. We might be more approachable to compared to others to say it's like, can you add a certain feature? And we'd say it's like, we're working on that anyway. We do a beta something custom accent. We give you this custom feature, and we say this feels like it's going the right direction.

Thomas Monz: 26:17

Let's do a little bit more.

Sebastian Hassinger: 26:18

Right.

Thomas Monz: 26:19

And then we would say we can offer something cool, convenient, nicely, fully integrated with bracket, which makes the life of end users easier, and they benefit from higher performance with no extra complexity. I think if you can enable that, like, with our hardware, but with the broader quantum ecosystem, And this is something we want to push for because it's like everyone will benefit from that.

Sebastian Hassinger: 26:45

Right.

Thomas Monz: 26:45

Then we are moving in the right direction.

Sebastian Hassinger: 26:48

Right. Yeah. I mean, you've already had, an active partnership with Classique. I know and they're you know, they are another partner of the bracket service. They use bracket as a way to to access hardware.

Sebastian Hassinger: 27:02

And I I think you're right. I mean, I I think the exciting thing about watching the hardware companies, the software companies sort of using bracket as sort of, the the GM bracket used to call it, like, the town square for the quantum industry. Everything you know, all the interactions benefit everyone actually, as we work out, you know, how to actually get this stuff to to the point where it has commercial value for our respect our our collective customers. Is there, you mentioned error correction. Is there a a scheme that AQT sort of favors in terms of error correction given the the characteristics of your platform?

Thomas Monz: 27:43

I think that's the the huge benefit of trapped ions in the first place that, okay, shuttling has its drawbacks, but at least we can relatively easily get long range interactions in Right. In the platform. And with so we don't this allows us to be very, very agnostic to codes. And if you think about my own history, I started I think it was still part of my PhD thesis together with Philippe where we did the normal three qubit repetition code. And then it was Marcus Muller who is now in who said, like, no.

Thomas Monz: 28:18

No. No. No. You need to think about topological codes. And so we started working on color code and scene error correction and so on.

Thomas Monz: 28:27

And nowadays, people are leaning heavily towards LDPC codes, which have, again, like, these long range interactions, which are something neat and and how to say native to trapped ions.

Sebastian Hassinger: 28:39

Right.

Thomas Monz: 28:40

But it's also been amazing to see, like, the changes which took, like, a few years only. If you think about it's like we had the surface code and the nice results from Google with larger code distances, and and it feels like almost the entire like, a large amount of the community has switched to LDPC within a or so. And Yeah. And I think if the next person stumbles upon an even better solution two, three years, I think for us with IONs, we don't have to change the chips each and every time.

Sebastian Hassinger: 29:14

That's interesting.

Thomas Monz: 29:15

Yeah. Yeah. Can can have a, let's say, say, more, software enabled robust platform where changes to the encoding doesn't necessarily mean we have to start from scratch to build a new trap.

Sebastian Hassinger: 29:27

Yeah. Yeah. Interesting. And and you you mentioned shuttling. So I think the the question that the ion traps, are are all faced with is is what is the answer to scaling up beyond a single trap?

Sebastian Hassinger: 29:44

And, you know, there are some approaches that like, the racetrack at Quantinuum, and others. What's the the AQT strategy for for getting to sort of those hundreds or thousands or even beyond in terms of numbers of qubits?

Thomas Monz: 30:00

So the the traps per se, I mean, the the first goal needs to be two d. That's like the the essentially, the number of shuttling operation depends on how many dimensions you have available in your

Sebastian Hassinger: 30:11

Right.

Thomas Monz: 30:12

Rule of thumb. There are some prefects, but you could say instead of a linear string, if I have something two d, it's only square root. And if you go three d, it's the third root and so on. Right. This is something which we push for, carefully selecting in that sense also how much overhead do I have to deal with.

Thomas Monz: 30:33

Can I get, say, all the control inside? Like, how many interaction zones do I have? It doesn't help if I have, let's say, say, I have a million but if I have only one interaction zone Right. Like, so so I need to integrate it photonics, and we have activities along those lines because in particular in Europe, there has been this chipset now which allows us to work with various partners to do integrated photonics, waveguides, and so on. But I think on the long run, similar to what we are doing here is, like, we have to think about how do we interface quantum computers, not just within a chip and make it larger, but what what, like, Ryan, I used to say is, like, there is scaling down.

Thomas Monz: 31:16

So, like, we have to make structure smaller. There is scaling up, which means control larger and larger system size, but you also have to think about scaling out. So what do I do if I have, let's say, HPC and clusters? I have two computers. I need this network favor.

Sebastian Hassinger: 31:32

Right.

Thomas Monz: 31:33

And with IONs, I think it's really neat because it's controlled to a large degree already with light, and we can get them out. And we have already demonstrated in Innsbruck, like, remote entanglement between ion traps across the camera. Mhmm. And so in that sense, being able to say it's not just, like, say, within the same room, but we, in principle, can also send it over larger distances, even telecom compatible. That makes trapped ions a really, really interesting platform Yeah.

Thomas Monz: 32:02

Essentially, other platforms are struggling with.

Sebastian Hassinger: 32:05

Right. It's a little bit easier than transaction in a superconducting regime, for example. So you you mentioned two dimensions. How do you plan to to realize a two dimensional trap? Because normally, I mean, the the, you know, the the optical trap is a single is just a one d sort of array.

Sebastian Hassinger: 32:27

Right?

Thomas Monz: 32:27

So no. No. That's no problem. So, like, opening a a junction, for instance, colleagues in mines have already been working on that. Colleagues in The UK have been doing that.

Thomas Monz: 32:37

So, like, you can have junctions. You can move left, right, up, down. Okay. It will be more of a question, I think, like, which easy to overlook. One part is, like, controlling the INTS, but how do I do all the multiplexing and so on?

Thomas Monz: 32:51

That if I let's say, if I have a million segments, I can't bring a million cables inside. Yeah. So and and I think that's a nice connection to what we said in the beginning. It's like, you yes. You can build your own diode laser.

Sebastian Hassinger: 33:05

Right.

Thomas Monz: 33:05

But you want to if you buy it, you can focus on the physics again. And so if you now say it's like, I have partners which give me the multiplexing capabilities, which give me the integrated photonics and integrated light control, then I can focus on the quantum processor again. Mhmm. And now you see it's like you you there are different building blocks that you have to fix and make available to the community in order to scale up and tackle the problems which are on the quantum side and not so much necessarily on the classical control side.

Sebastian Hassinger: 33:37

I see. I see. Interesting. So okay. So you're you're you're right on the verge of deploying on Amazon Bracket.

Sebastian Hassinger: 33:46

By the time this podcast is out, you will be on Amazon Bracket. Beyond that deployment, there's obviously going to be, you know, an initial phase of sort of settling into that. What do you see as sort of the next next major, you know, objectives or goals for AQT in the in the coming year or two?

Thomas Monz: 34:05

So I think the the key part is we have to AQT, but the community as a whole, we have to provide better quantum computers, which offer, like, a benefit for end users. It I think as long as we stay in this, it's, let's say, scientifically, technologically interesting, it's nice. But if we want to unlock economic value and an actual market, we have to build devices where people say it's like it's better to use this quantum device compared to using a classical device. And I think there, the advantage for us with INCs because we have lower error rates. The overhead on how large the system actually have to be Right.

Thomas Monz: 34:51

Can be significantly smaller, and this is something that we are pushing for.

Sebastian Hassinger: 34:57

Really interesting. That's fantastic. Well, Thomas, thank you very much for your time. This has been really interesting, and I'm very much looking forward to seeing, you know, what uses people find for the AQT hardware on the Amazon Braket service. Thank you.

Sebastian Hassinger: 35:12

Thank you for listening to another episode of the podcast, a production of the New Quantum Era hosted by me, Sebastian Hassinger, with theme music by OCH. You can find past episodes on www.newquantumera.com or on bluesky@ newquantumera.com. If you enjoy the podcast, please subscribe and tell your quantum curious friends to give it a listen.