The New Quantum Era - innovation in quantum computing, science and technology

The quantum computing industry has been stuck at roughly 100 qubits for years — not because of physics, but because of wiring. Matthijs Rijlaarsdam, co-founder and CEO of QuantWare, explains how his company's 3D vertical chip architecture (VIO) could break through that ceiling to 10,000 qubits by 2028, and why the quantum industry needs to start thinking like the semiconductor industry if it wants to actually deliver on its promises.

Show Notes

Scaling Quantum Hardware Like Semiconductors with Matthijs Rijlaarsdam
The quantum computing industry has been stuck at roughly 100 qubits for years — not because of physics, but because of wiring. Matthijs Rijlaarsdam, co-founder and CEO of QuantWare, explains how his company's 3D vertical chip architecture (VIO) could break through that ceiling to 10,000 qubits by 2028, and why the quantum industry needs to start thinking like the semiconductor industry if it wants to actually deliver on its promises.

Episode Summary
This conversation is for anyone trying to understand why quantum computers haven't scaled as fast as promised — and what it would take to change that. Matthijs brings an unusual perspective as a computer scientist (not a physicist) who co-founded QuantWare out of TU Delft's QuTech to become the world's first commercial supplier of superconducting quantum processors.
Rather than building a full quantum computer, QuantWare sells QPUs as components — the "TSMC of quantum." In this episode, Matthijs walks through the VIO architecture that routes signals vertically through stacked chiplets instead of along chip edges, why specialization and volume economics are the only realistic path to useful quantum computing, and how the Dutch quantum ecosystem punches far above its weight thanks to consistent long-term investment.

What You'll Learn
  • Why the quantum industry is stuck at ~100 qubits — and how 90% of current chip area is consumed by signal routing, not qubits, creating a fundamental scaling wall
  • How VIO's 3D chiplet architecture breaks the wiring bottleneck by routing signals vertically through stacked silicon modules, enabling 10,000-qubit processors that are physically smaller than today's 100-qubit chips
  • Why quantum computing will be heterogeneous — different platforms (superconducting, trapped ions, neutral atoms) have different trade-offs analogous to CPUs vs. memory vs. storage in classical computing
  • The economics that make specialization inevitable — why cable costs need to drop from EUR 1,500 per line to cents, and why volume manufacturing is the only way to get there
  • How QuantWare's three business models mirror the semiconductor industry — selling packaged QPUs (Intel model), foundry services (TSMC model), and packaging services for third-party chips
  • Why the Dutch quantum ecosystem succeeds — consistent decade-plus government investment in QuTech, EUR 600M+ to Quantum Delta NL, and the WENEC report recommending EUR 9.4 billion for quantum infrastructure
  • What "Quantum Open Architecture" means in practice — how making QPUs commercially available lowers barriers for the entire industry, similar to how standardized PC components enabled the computing revolution
  • QuantWare's roadmap: VIO-40K shipping in 2028 with up to 10,000 qubits, and a path to 1 million qubits using arrays of chiplet modules

Resources & Links

Company
  • QuantWare — world's first and largest commercial supplier of superconducting quantum processors
  • VIO Technology — QuantWare's 3D vertical integration and optimization architecture
  • VIO-40K announcement — press release on the 10,000-qubit scaling breakthrough

Coverage & Analysis

Partnerships Mentioned

Ecosystem & Policy
  • QuantWare 2026 industry predictions — QuantWare's view on entering the kiloqubit era
  • QuTech — TU Delft quantum research institute where both QuantWare co-founders did their graduate work
  • Quantum Delta NL — Dutch national quantum technology program (EUR 600M+)
  • DARPA HARK program — Heterogeneous Accelerated Roadmap using Quantum Solutions; referenced by Matthijs as validation of the heterogeneous quantum computing thesis

Key Insights
"There is no path towards useful quantum computing without specialization. That is a total fantasy." — Matthijs Rijlaarsdam on why volume economics and the semiconductor model are inevitable for quantum


"The difference between EUR 1,500 and 10 cents per cable line — that's all volumes and yields." — on how manufacturing scale, not physics breakthroughs, will drive the next phase of quantum cost reduction


"If you look at it on a cost-per-qubit basis, VIO-40K at EUR 50 million is actually a 10x reduction from where we are today. Anyone claiming they'll do it for less is just not telling something realistic." — on the real economics of scaling quantum hardware


"Imagine if you were a company today and you wanted to do interesting stuff in AI, but you first had to develop a three nanometer process to make the chips. It would be completely ridiculous. And in quantum, that's what everyone is doing." — on why vertical integration won't survive at scale


"Good companies will get funded. We have in general not been restricted by access to capital ourselves." — on navigating European deep-tech venture capital

 
Related Episodes

Sponsor
qubitsok — Cut Noise. Work Quantum. The quantum computing job board and arXiv research digest built for the community.
  • Job seekers & researchers: Subscribe free at qubitsok.com — weekly job alerts + daily paper digest filtered by 400+ quantum tags.
  • Hiring managers: Post your quantum role and reach 500+ targeted subscribers. Use code NEWQUANTUMERA-50 for 50% off your first listing at qubitsok.com/post-job.

Subscribe & Connect
Interested in quantum hardware scaling and industry economics? Subscribe to The New Quantum Era on Apple Podcasts, Spotify, or YouTube.
Stay current — sign up for our newsletter at newquantumera.com and follow us on LinkedIn and Bluesky.

Creators and Guests

Guest
Matthijs Riljaarsdam
Co-Founder and CEO of QuantWare

What is The New Quantum Era - innovation in quantum computing, science and technology?

Your host, Sebastian Hassinger, interviews brilliant research scientists, software developers, engineers and others actively exploring the possibilities of our new quantum era. We will cover topics in quantum computing, networking and sensing, focusing on hardware, algorithms and general theory. The show aims for accessibility - Sebastian is not a physicist - and we'll try to provide context for the terminology and glimpses at the fascinating history of this new field as it evolves in real time.

Sebastian Hassinger (00:01.038)
Hi, Matthias. Great to see you. Thank you for joining. I've been looking forward to having you on the podcast for a while. Can you start by, you know, you've had a really interesting journey. You're now a co-founder and CEO of a very successful quantum computing company, QuantWare. How did you get to that point? I think that's always fascinating to hear.

Matthijs Rijlaarsdam (00:24.674)
Well, first of all, thanks for having me. Great to be here. Always wanted to be here, of course. Yeah, so I'm actually a computer scientist by background. And I did, at some point during my thesis research, did research at Q-Tech, which is a large, arguably one of the best, but definitely one of the largest quantum computing research institutes here on the TU Delft campus.

on the simulation of quantum computers, on classical computers. So I'm not a physicist, so that's always a atypical in our space. Well.

Sebastian Hassinger (01:04.339)
Neither am I, so I deal with that too. Yeah. Let's talk about the physicists behind their back.

Matthijs Rijlaarsdam (01:07.884)
Well, here we are. Yeah, exactly. Let's do that. No one is listening. Yeah. Yeah, so I was doing research on trying to look at the question, what kind of quantum computer calculates that a classical computer can't calculate. That's, I guess, one of the most interesting.

Sebastian Hassinger (01:32.674)
That's the trillion dollar question.

Matthijs Rijlaarsdam (01:34.722)
Yeah, right. I mean, if you think about it, it's equivalent to the universal theory of everything in physics, right? If you can answer that exactly, you'll probably win like 10 Nobel prizes or something. Anyway, so the idea is, if you look at simulating quantum computers on classical computers, if you can simulate a quantum circuit on a classical computer, it's probably not very quantum-y. It's probably not so interesting. So anyway, I was doing that.

Sebastian Hassinger (01:46.552)
Yeah.

Matthijs Rijlaarsdam (02:03.53)
And then through a mutual friend, was introduced to my co-founder, Dr. Alessandro Bunno, our CTO, who at the time was the lead fabrication, the lead quantum chip maker of the DeCarlo lab. And we got to talking, and he kind of had this, at that time, they were working, they were...

Sebastian Hassinger (02:20.216)
Right.

Yeah.

Matthijs Rijlaarsdam (02:31.074)
I think just at the end of the Intel Q-Tech IRPAT collaboration, where they made some of the most impressive first quantum error correction demonstrations, et cetera, et And then everything about scaling up. And if you want to scale up superconducting chips, you run into certain problems. And still, the field hasn't solved those. still kind of, Intel announced at that time a 49-qubit chip.

Sebastian Hassinger (02:36.269)
Right.

Matthijs Rijlaarsdam (03:01.218)
the state of the art is maybe pushed to 100 or so. IBM will be in 120 qubits in 2028 still. And so they had this concept that we now call VIO, for vertical I-O. We got to talking, we hit it off, and we were able to spin out that theme and that concept as one of the IP of the Intel collaboration. And the rest is history. Here we are.

Sebastian Hassinger (03:06.904)
Yeah.

Sebastian Hassinger (03:27.81)
Hmm. Yeah. Yeah. I mean, that's, really interesting. So you just started, you sort of launched VIO, I guess, at towards the end of last year at Q2B. you did a keynote where you talked extensively about it, but it really was the thinking rate at the very launch of the company. guess that's 2020 or 2021, roughly five years ago, right? Wow. That's really interesting.

Matthijs Rijlaarsdam (03:49.536)
Yeah, that's right. mean, it's right. We basically, so, you know, we worked intentionally in silence on VIO for about five years to, you know, we spent out a team which, you know, best QPURs texture team in the world. But the idea was to kind of build up this head start, get this far enough along.

Sebastian Hassinger (04:00.085)
Mm.

Matthijs Rijlaarsdam (04:14.358)
that we actually have data and can show it. And it's not just a cool render and a concepts. And in parallel builds up this market share, right? Like we launched immediately with the world's first commercially available quantum processor, which was only a five-quid chip, right? So Prano with the idea that the more volume we can get and the more we can change the predominant business model in this industry, that is preparing the market for when we introduce VIO.

Sebastian Hassinger (04:16.494)
Hmm.

Matthijs Rijlaarsdam (04:44.127)
interesting moment, indeed, like as of December, we're talking about it publicly and, you know, really looking forward to the impactful make on the world for that. Yeah.

Sebastian Hassinger (04:49.122)
That's exciting. That's really exciting. So you mentioned sort of the product line of full integrated QPUs or packaged QPUs like Contralto. And that's sort of analogous to like an Intel business model, right? You're selling chips that people can use. And I think mostly they're being used in like, I know the Israeli Quantum Computing Center has a Quantware chip.

and a quantum machines integrated system that they use for research. There's other systems like that in, I think, Korea and Colorado and others where it's test beds and research devices that are powered by quantware. Is that sort of accurate?

Matthijs Rijlaarsdam (05:35.71)
I mean, that's definitely part of our business. So today, the majority of our revenue comes from, indeed, as you mentioned, QBUs, Quantware-designed, Amperbergated, and packaged processors. We sell those primarily to commercial entities that typically are in the building, developing quantum computers to then sell those computers or the resulting compute. And yeah, some of the examples you mentioned, I think, are good examples of that.

Sebastian Hassinger (05:38.894)
Yeah.

Sebastian Hassinger (06:01.069)
Okay.

Matthijs Rijlaarsdam (06:05.57)
And yeah, and as you mentioned, this Intel business model, I mean, when we started, we still even used that in our public marketing and now, know, for obvious reasons, we were kind of, you know, use different analogies. But I think the analogy still holds, right? the exact same economic principles of the classical compute space, where this drive to specialization, you know, increased volumes, which led to, you know,

near monopolies for Intel, for instance. That is the exact same laws of economics that apply to this space.

Sebastian Hassinger (06:42.296)
Really? you think, I mean, you think that that will happen if the modalities converge on superconducting or even if we have multiple modalities that have viable commercial use cases in, in, parallel with one another.

Matthijs Rijlaarsdam (06:57.282)
So I think, mean, everyone will say, of course, my platform is best, right? But actually, I think our view is a lot more nuanced. So we think that quantum computing will be heterogeneous in the sense that a lot of these platforms are very sensible, but they just have different trade-offs. But you can even see, I think this is becoming a more mainstream view. Like, there was this Hark.

Sebastian Hassinger (07:02.55)
Of course.

Matthijs Rijlaarsdam (07:25.302)
project from DARPA recently announced, this is becoming more mainstream. you can think of this as certain platforms, say, neutral atoms or trapped ions, they have fantastic fidelity and really long coherence times. But they're very slow operations. And so probably they're quite suitable for more memory type applications. And then you have

Sebastian Hassinger (07:26.166)
Yes, yes.

Sebastian Hassinger (07:51.405)
Right.

Matthijs Rijlaarsdam (07:53.71)
say superconducting, which is very fast, but as a counter to that very error-prone, you need a lot of error correction. But it's very fast. And if you want to do quantum compute at scale, the speed of operation really matters. There was this kidney paper from last year that with a million superconducting qubits, you can crack an RSA key in about a week. Well, imagine being 1,000 to a million times slower than that. For compute, it's not going to be great.

Sebastian Hassinger (08:20.748)
Yeah. Yeah.

Matthijs Rijlaarsdam (08:23.106)
I think the analogy here is, again, going to classical computing. You wouldn't want to process your data on your hard drives, that'd be super slow, but you wouldn't want to store a Blu-ray video on your CPU, or your RAM today is even worse.

Sebastian Hassinger (08:34.998)
Right. Yeah.

Sebastian Hassinger (08:40.898)
right in your ram.

Sebastian Hassinger (08:47.758)
Exactly, it's way too expensive. Okay, that makes sense. And you think within those parallel modalities, there's still, there's the scaling advantage when you're a specialist in one of those modalities, you can deliver performance at a scale and economics that others can't match.

Matthijs Rijlaarsdam (09:09.442)
Yes, and I would even say that more strongly, there is no path towards useful quantum computing without specialization. That is a total fantasy. I'll give an example. We announced VIO 40K, and we also announced that we will retail that for starting at 50 million euros.

Sebastian Hassinger (09:13.902)
Hmm.

Sebastian Hassinger (09:19.138)
Right.

Yeah.

Sebastian Hassinger (09:37.581)
Mm-hmm.

Matthijs Rijlaarsdam (09:37.826)
And the response to that was, a lot of people said, that's a really large amount of money. But you need to look at that on a cost curve per qubit basis. And actually, this is a 10x production in price on a per qubit basis from where we are today. So actually, if you look at it that way, anyone claiming that they'll do it for less is just not telling something realistic. Like a 10x production in two years' time is hyper aggressive.

Sebastian Hassinger (10:03.406)
Right.

Sebastian Hassinger (10:06.786)
Yeah. Yeah.

Matthijs Rijlaarsdam (10:06.998)
And it goes for the other parts of the stack as well. Take high-density ribbon cables, these cables that you need. You need quite a lot of them to get to large-scale quantum computers. One of those cables, yeah, well, in VIO 4DK, you'll need 4DK of those to power, say, a typical tunable.

Sebastian Hassinger (10:21.72)
Sounds like tens of thousands actually.

Matthijs Rijlaarsdam (10:33.056)
qubits with tunable copper design will take about four signal lines per qubit on average. So therefore for 10,000 qubits, you need 40,000 lines. Well, today these are about one and a half thousand euros a line. And so that needs to come down pretty aggressively. But if you look at it from a first principles basis, what's a cable? Well, it's a thin layer of some kind of plastic material with a thin layer of superconducting metal and then a thin layer of plastic.

Sebastian Hassinger (10:36.632)
Right.

Matthijs Rijlaarsdam (11:02.658)
That's it. so on a material basis, the material floor is a cent. Let's call it 10 cents. Let's get some room for margin. the difference between 1,500 and 10 cents, that's all volumes and yields. And so I think this industry is going through a transition where this whole framework around, it's not science, it's engineering anymore. Well, actually, it's much more

Sebastian Hassinger (11:09.4)
Right, right, right.

Sebastian Hassinger (11:19.586)
Right. Right.

Matthijs Rijlaarsdam (11:32.898)
transitioning now to economic cost curves and bringing up those volumes because, mean, technically you can make a lot of those cables. It just needs to be much cheaper.

Sebastian Hassinger (11:35.564)
Mmm.

Sebastian Hassinger (11:41.366)
Right, right, right. And so you just said sort of, it's not science, it's engineering. Now, your VIO architecture, you see a clear path to getting to tens of thousands of qubits with that architecture.

Matthijs Rijlaarsdam (11:55.927)
Yes.

Sebastian Hassinger (11:57.272)
That's incredible. That's really cool. you'll walk me through, like, I mean, we mentioned VIO a few times, like what exactly in, you know, in broad strokes, what is VIO?

Matthijs Rijlaarsdam (11:59.596)
I mean, I mean, I'm sorry.

Matthijs Rijlaarsdam (12:10.562)
So VIO is a 3D cube view architecture. And what I mean with that is it's a way to make a very large quantum processor or an array of qubits out of many smaller modules. And a VIO module, what it typically looks like, it's a qubit plane consisting out of a few qubit chiplets, so a big chip out of many smaller chiplets.

that resides in an array of vertically placed chips. And so that is basically a cube of silicon. And that is really important for a variety of reasons. But one of them is it gives you real estate to integrate other parts of the quantum computer in this stack of chips, which if you get to really large qubit counts, actually becomes a very important thing to resolve. Because otherwise, you just simply run out of space.

Sebastian Hassinger (13:00.855)
Hmm.

Matthijs Rijlaarsdam (13:10.53)
It also gives you a way because it is modular to actually make these things because of yields. you want to make a large, let's say you want to make a chip with a million qubits, well, that is just technically infeasible, but you can make a chip of say maybe 100 qubits and you can attach those together. And then third, going from science to engineering to like industrial scale up.

Sebastian Hassinger (13:27.82)
Right. Right.

Matthijs Rijlaarsdam (13:35.872)
Because it is silicon based, it is producible at a massive scale because we just use normal semiconductor tooling and processes. And that combination is really important.

Sebastian Hassinger (13:43.854)
So are the chiplets individually fabbed and then combined into a chip and then the chips are combined into these layers or is the chip the sort of the unit that you're fabbing and then combining the chips into the layers?

I guess I'm just asking is a chiplet just like is a chiplet a standalone unit with a certain number of cubits that then gets connected to other chiplets?

Matthijs Rijlaarsdam (14:10.914)
Typically, yes. And you'd want that to be that way because it again makes it mass-producible. It's just this repeatable unit cell that you can produce at very large scale, therefore cost collapse. And I think what's really important to emphasize is that, so right now, the field's stuck at 100-qubit chips, the size of an A4 piece of paper. And so you can't really get to...

Sebastian Hassinger (14:12.27)
Hmm.

Matthijs Rijlaarsdam (14:38.666)
your skills of say 10,000 qubits and beyond, which we will need to do something, you know, actually economically relevant without resorting to things like networking. Whereas if you do, and this is already proven in, you know, in experiments in real life, a connection on qubits between two chiplets is as good as a connection between two qubits.

only single chip. So intra-chip fidelity is as good as inter-chip fidelity, which means that you can scale up without losing compute power. And therefore, it is exponentially better than a network-based alternative.

Sebastian Hassinger (15:05.613)
Hmm.

Right.

Sebastian Hassinger (15:16.834)
Right.

Sebastian Hassinger (15:23.702)
So these are superconducting leads that bridge from one chiplet to another and one chip to another. They're all interconnected with superconducting leads just as the qubits themselves are on the chiplet. Right.

Matthijs Rijlaarsdam (15:30.838)
Yeah, yeah.

Matthijs Rijlaarsdam (15:36.578)
It is just, from the perspective of the user, it is just one very large chip.

Sebastian Hassinger (15:41.902)
Yeah, yeah. You said a cube. I'm sort of imagining this, this, you know, cube of qubits floating in the bottom of a dill fridge with 40,000 leads coming off it.

Matthijs Rijlaarsdam (15:50.092)
Why, maybe we should clarify that a little. It's little cubes that we then attach together. So the eventual thing is more of a square, if you will.

Sebastian Hassinger (15:56.971)
Okay.

Sebastian Hassinger (16:01.088)
Okay, okay. That's amazing. okay, so each of those units that you're connecting together, how many qubits is sort of in a chip that's made up of chiplets?

Matthijs Rijlaarsdam (16:13.23)
That's a design choice to some extent. And what that will typically also depend on is what are the yields of your qubit chiplets and what are the yields of attaching those together, right? There'll just be some optimum there.

Sebastian Hassinger (16:14.774)
I see. Okay.

Sebastian Hassinger (16:26.774)
Yeah, yeah. Again, back to the economics, as you said, it's like, it's the yield economics. that's really interesting. So, okay. So, we mentioned sort of the, the package CPU QPU business. you have two other sort of models for engaging with the industry. One is more of a Foundry services, right? So you're actually, collaborating with, vendors who are seeking to build a full quantum, computing system.

and you're providing the chip foundry capability much in the same manner as TSMC provides chips to Apple and Nvidia and others, right? Is that a good analogy?

Matthijs Rijlaarsdam (17:06.996)
I think so, yeah. Yeah, I think one way to think about our QPU business is that that is, you know, our internal design teams designing QPUs optimized for quantum error correction on our own platform. But we can also take in indeed customer designs on our own platform. And we can even take in QBit's chiplets of customers on the VIO platform.

Sebastian Hassinger (17:09.368)
That's really cool.

Sebastian Hassinger (17:25.485)
Right.

Sebastian Hassinger (17:35.554)
Hmm.

Matthijs Rijlaarsdam (17:35.82)
So it's really these three verticals in which we operate.

Sebastian Hassinger (17:38.688)
Interesting. Interesting. So somebody has, they think they have, you know, the magic qubit design, you can actually integrate that into the chiplet or if they want to leverage your qubits but have a different sort of, you know, base system architecture, you can adapt VIO2 to that.

in the way that you connect the individual chips. That's really, really interesting. So you're working with, I think with Seq already on sort of this type of collaborative model, right?

Matthijs Rijlaarsdam (18:11.554)
So, mean, we have a bunch of customers and Seek has been a wonderful customer and partner basically from our early days. we, know, I think the work that they're doing on SFQ has the potential to be really important.

Sebastian Hassinger (18:26.05)
Hmm. That's very cool. And then, mean, when you describe that sort of range of services, I guess the lightest services that you provide is just packaging, right? Like somebody brings you their full QPU, just the silicon or the chip itself, and then you package that up into something you can integrate into a full system design. that another sort of aspect?

Matthijs Rijlaarsdam (18:51.936)
Yeah, so we're really positioning ourselves to get the entire industry to scale using VIO because this is the bottleneck that is keeping back the quantum industry. And so we believe it to be for quantum computers to make the positive impact on the worlds that they can make, they need to be larger and they need to be so in an economical way. And that last bit, a lot of...

people in industry typically tend to forget. But getting there in an economic way means making it at industrial scale, at high volumes with very high utilization rates and efficiencies. And so therefore repositioning in a way that we can get everyone to scale using VIO. And there's of course, there's some flywheel effect in there, right? Like the more customers we get on our platform, the better the product is that we can provide to our customers.

Sebastian Hassinger (19:22.656)
Right.

Sebastian Hassinger (19:29.24)
Yeah. Yeah.

Sebastian Hassinger (19:40.62)
Right. Right.

Right, right, right. So I mean, this is essentially an open systems approach to quantum processors. How difficult is that to sort of try to establish or champion an open standard or open systems architecture at this super early stage of a technology?

Matthijs Rijlaarsdam (20:09.644)
So it's been surprisingly easy. So we call this the quantum open architecture, indeed. It's been surprisingly easy because it just massively lowers costs and development times to start building these systems and then actually to do interesting stuff, right? Imagine, again, I think the analogies to the semi-con world are really interesting to look at. Imagine if...

Sebastian Hassinger (20:12.162)
Hmm.

Matthijs Rijlaarsdam (20:37.122)
You were a company today, even if you were a large company today, and you wanted to do interesting stuff in AI, but you first had to develop a three nanometer process to kind of make the chips, then design them and drive the drive. It would be completely ridiculous. And in quantum, the reason that everyone was vertically integrated is that that was the state of the industry.

Sebastian Hassinger (20:51.243)
Right.

Sebastian Hassinger (20:55.596)
Yeah. Yeah.

Matthijs Rijlaarsdam (21:05.612)
But as you get to larger scales, it becomes completely ridiculous that everyone is doing that themselves. That will just be too inefficient.

Sebastian Hassinger (21:08.014)
great.

Right.

Right. It's interesting from that perspective, you could say that every disruptive technology is actually an incremental progress, right? Cause you're building on every prior sort of incremental increase in like the smaller and smaller leads in lithography for semiconductors, for example, that's a many, many decade effort that got to a GPU or a modern arm chip or whatever.

Matthijs Rijlaarsdam (21:41.9)
Yeah, and I think indeed like, you know, it's compounding improvements that, you know, if you lower cost every year by 30%, yeah, then at some point it becomes completely unbeatable. And if you get, you know, it's the same again with chip making. you, whether you make 10 or a hundred chips, your marginal costs barely change. It's all about depreciation, right? And so if you sell commercially a hundred chips, your competitor sells 10, your margins are roughly.

Sebastian Hassinger (21:46.829)
Mm-hmm.

Sebastian Hassinger (21:54.872)
Right.

Sebastian Hassinger (22:04.493)
Right.

Matthijs Rijlaarsdam (22:11.426)
know, 10 times better. and, but that's a flywheel, right? Because if, you know, that also, that means that you can get many more, you can just make a much better product and reinvest much more in R &D. And that's, yeah, over time, that, that, that, that difference gets so big that you can dominate a space.

Sebastian Hassinger (22:13.1)
Right, right.

Sebastian Hassinger (22:30.082)
Yeah, I suppose that really kicks in once the chips that you're producing that we're shipping in quantum computing actually have some commercial application, right? mean, then you've got a product that there'll be some demand in the market for repeated, like, you know, the customers will multiply as opposed to today where it's, you know, there's a small number of customers that tend to be really at the moment, concentrated in the public sector or at least public private sector.

where ecosystems are trying to be NQCC or IQCC, as I mentioned before, those types of efforts to build quantum industry ecosystems. But at some point there'll be quantum data centers and quantum applications that the enterprise in some industry or other will actually want to access and pay money for. And then your economy is really kicking to overdrive.

Matthijs Rijlaarsdam (23:24.482)
Yeah, and I think what's really helped with that view is the current hyper skilled data center build outs, right? you know, not getting into the whole discussion whether that's, know, whether we're over building there or not, but that's, know, like north of $7 trillion cumulative CapEx investment by 2030. And if you look at it, you know, in that way, like, you know,

Sebastian Hassinger (23:32.226)
Mm-hmm.

Matthijs Rijlaarsdam (23:52.534)
The thing that restricts that is access to electricity. so quantum computers, hyperscaled quantum computers, so we're talking maybe quantum computers of a billion dollars plus, like very large pieces of compute and infrastructure. In that context, not only is a billion dollars reasonable, but more importantly, quantum computers will provide exponentially more compute in certain use cases per dollar or per watt.

Sebastian Hassinger (24:13.379)
Yeah.

Matthijs Rijlaarsdam (24:21.666)
let's interchange these. that means that even if that percentage is small, I would argue that percentage is going to be quite large. But the honest answer is we don't exactly know, right? Going back to that would be equivalent to a universal theory of everything in physics, if you could answer that exactly before end. But even if it's a couple percents, that is a massive market.

Sebastian Hassinger (24:37.644)
Yes. Yeah.

Sebastian Hassinger (24:45.27)
Yeah. Yeah. Yeah. Yeah. Yeah. You mentioned access to power being the bottleneck, but there's another bottleneck, which is access to capital. As a European company, I've heard other European founders say that in quantum, it can be difficult to access the scale of capital, especially in hardware from the European.

private markets compared to the American markets. Is that something that sort of added friction to the growth of quantware?

Matthijs Rijlaarsdam (25:19.874)
I mean, obviously, capital markets in Europe and also in the Netherlands are not as developed as in the US. I don't think that's a controversial statement. Something I am very active in advocating for and trying to help develop, especially here in the Netherlands and Europe. However, it also, of course, depends a little bit on the

the rate of progress and the momentum you have as a company, you know, good companies will get funded. And yeah, so I think, you know, for us, I think it would be better if the European ecosystem would develop further. I think there's a lot to be gained by that. We have in general not been restricted by access to capital ourselves though.

Sebastian Hassinger (26:14.146)
Okay, good. Yeah, I mean, you mentioned, I mean, there's a lot to gain. mean, in a way, I think that the, you know, there's something like 30 countries that have national quantum strategies around the world now was, you know, last I looked on the Cureca site, it was something like 70 billion, I'm sure it's more than that now, committed publicly, you know, who knows what the actual sort of scale up to, but I feel like the Dutch ecosystem is

one of the examples of how to do it. mean, you came out of, as you said, Q-Tech, which is part of Delft, which is the D in quantum delta. And you've got QDNL participations on your capital, as an investor. And it feels like the skill base you're able to recruit out of at Q-Tech really is kind of a dream team of talent.

when you look at the experience of your sort of founding team and the people you've been able to recruit, what do you think that sort of the Dutch sort of punching above its class, for lack of a better term, it's like the impact of the Dutch ecosystem seems outsized compared to the size of the country. What do you think that sort of is attributable to?

Matthijs Rijlaarsdam (27:38.018)
So first of all, I think you're right. And I'm not saying that because I'm, you know, Dutch and standing here in that ecosystem. I think you're right. I think the reason for that is consistent long-term investment and policy. And Q-Tech, think, is a prime example of that, right? Like that is well over a decade of governmental investment at a serious size and also attracting serious international companies.

to do their research here, right? Microsoft, Intel, et cetera, et And then following that up with more than 600 million to quantum delta and NL, which of course for the Netherlands, but for any country is a sizable amount of money. And I think the key thing now is, I think really this is kind of a make or break moment. We now have this very diverse ecosystem with all kinds of.

Sebastian Hassinger (28:06.508)
Yeah. Right.

Matthijs Rijlaarsdam (28:32.63)
parts of the ecosystem, parts of the value chain, especially some of you around Delft. And we have this real opportunity to become this value chain concentration points, which has very economical benefits, but also deeply national strategic benefits. And the key thing for that, you mentioned it already, is

access to large scale of capital, right? Like in the end, know, this is not a, know, these companies will need hundreds of millions of capital. That's just the nature of this business. The market is very much big enough to support that. And, but that needs to be, you know, that needs to be there in time and it needs to be accompanied by the rights, you know,

governmental policies around it. And of course, we're now getting to a size where we'll be in different countries as well. we will internationalize. But I think what was a really positive sign is a few weeks ago, there was a, the WENEC report was published. This is a report written by the former CEO of ASML, comparable to the Drager report for Europe, if you're familiar with that. Where one of the recommendations was to invest 9.4 billion euros

Sebastian Hassinger (29:27.203)
Right.

Matthijs Rijlaarsdam (29:50.004)
in quantum internet islands. And I think really that is the skill we need to be thinking of.

Sebastian Hassinger (29:54.68)
That's incredible. Are you lining up already for that?

Matthijs Rijlaarsdam (29:59.394)
I mean, obviously, this is a report that will then go to the government and to politics. And so I don't expect there will be 9.4 billion on our bank account tomorrow. But I do think it's a very positive sign that the kind of momentum and the long term policy support that Quantum has here in the Netherlands and in Europe.

Sebastian Hassinger (30:03.212)
Yes, I know. I know.

Sebastian Hassinger (30:10.325)
Yeah.

Sebastian Hassinger (30:20.652)
Yeah, that's fantastic. So, okay, so this is all super exciting. Can you leave us with any sort of prediction forecast? Where are you going to be with QuantWare, with VIO in a year or two in terms of your roadmap?

Matthijs Rijlaarsdam (30:39.714)
So in 2028, the first VAO for the Gator race will be shipping to customers, which that is up to 10,000 qubits. And so 100 times more and 100 times denser than anyone else. And yeah, we're really looking forward to see the impact that that will make.

Sebastian Hassinger (30:47.086)
So that's 10,000 qubits.

Sebastian Hassinger (31:03.906)
That's incredible. Okay, well, you're gonna have to come back when you ship that. Because I want to talk about it. That's great, Mathias. Thank you so much for your time. This is really exciting. I've been a long time admirer of QuantWare, so it's great to have a chance to dig into it with you.

Matthijs Rijlaarsdam (31:19.618)
Pleasure to be here. Thanks for having me.