Filed by Horizon Quantum Holdings Pte. Ltd.
Pursuant to Rule 425 under the Securities Act of 1933, as amended, and deemed filed
pursuant to Rule 14a-12 under the Securities Exchange Act of 1934, as amended
Subject Company: dMY Squared Technology Group, Inc.
Commission File No.: 333-292737
The following is a transcript of an interview
with the CEO of Horizon Quantum Computing Pte. Ltd., Joseph Fitzsimons.
Event: Rosenblatt Securities’ Quantum Computing
Series 2026
Date: January 16, 2026
<<John McPeake, Analyst, Rosenblatt Securities>>
All right, welcome to Session 2 of the day, as
this wallpaper they made says I’m John McPeake, Senior Analyst here at Rosenblatt Securities. Good morning. Welcome to the conference
and happy Friday. This is the second session of the day. Thank you for joining us at our inaugural Quantum Series Conference today.
Next up is Horizon Quantum. This is a company
that’s bridging the gap between classical and quantum computing by a hardware agnostic software stack from the qubits to the application
layer. It’s very intriguing to me as a software analyst historically. We’re really lucky to have the CEO and Founder of Horizon with us
here this morning, Dr. Joe Fitzsimons.
Before founding Horizon Quantum in 2018, Joe held
a tenured position as associate professor at the Singapore University of Technology and Design, where he led the Quantum Information and
Theory group. And he also spent time at Merton College, Oxford as a senior research fellow in the Materials Department where he holds
a doctorate from the University of Oxford. The research focused on quantum computing architectures. He has a Bachelor of Science degree
in theoretical physics from University College Dublin and a doctorate from University of Oxford. So I feel very inferior from an educational
standpoint to our guest. Thank you for joining us, Joe.
We have kind of a rough 15, 15, 15 structure today.
We’re not going to hold that exactly 15 minute presentation, 15 minute fireside chat. Then we’ll open it up for questions via the system.
Please don’t be shy. So, Joe, go right ahead and tell us about Horizon Quantum.
<<Joe Fitzsimons, Chief Executive Officer>>
Great, thanks very much, John. So I’m sure you’re
all familiar with the customary disclaimers, but let me start off by telling you what our focus is at Horizon. We’re really focused on
getting to a point where quantum computers, where applications running on quantum computers are delivering real value for the end users
of those applications. And our focus is really on building the software infrastructure to enable this. And I would love to say that that’s
already the case today, but the reality of quantum computing is that it’s not quite at the point where applications running on quantum
computers are regularly solving hard problems. And that’s really a major focus of the field. How do we get to that point as quickly as
possible? How do we make quantum computing as useful as possible for tackling really difficult computational problems?
But at least from my side, this is a two-sided
problem. There is both a software side to it and a hardware side. And on the hardware side, things are looking pretty interesting, at
least from my perspective I feel that quantum computing has reached something of an inflection point. And there’s a couple of different
reasons for this. One of the reasons I would say a real major milestone in the field in my view is the fact that we have seen multi-round
error correction achieved for the first time in late 2024, where error correction was actually improving the performance of qubits. So,
the lifetime of the quantum information was being extended.
Now, does that mean we’ve managed to squeeze
out all of the errors yet? And by we I mean the community, not Horizon. No. Quantum computers are still swamped by errors, but there is
really interesting kind of threshold behavior in quantum computing where if you’re above the error correction threshold, if your
operations are not quite precise enough, then when you try to do error correction, you make things worse. And now, although we’ve
known how to do error correction for 30 years, it’s only really in the last year or so that it has become – started to become
possible and now things are starting to get better.
And as you start to improve, if you get say a
twofold reduction in your error rate, but your error encoding can correct up to nine errors, meaning that you need 10 errors before you
get an actual error in the encoded information, then each factor of two improvement in the physical error rate is a factor of 1,000 improvement
in the error rate on the encoded information. So once things start going in the right direction, they can go pretty quickly. And it’s
also the case that the overhead you need to do this kind of error correction has come down substantially. In the last four years, there
has been a major change in our understanding of quantum error correction codes.
There has been the discovery of asymptotically
good codes, which have just much better properties that – than was previously taught. So for a long time it was thought that you
would need maybe 1,000 or even 10,000 physical qubits for every one error protected qubit. And that number has fallen off a cliff with
the introduction of quantum LDPC codes that really have much better parameters than the standard surface code that had been explored before.
So now instead of thinking of 1,000 or 10,000
to 1, the community is starting to look at 20 to 1, 10 to 1, 5 to 1. And that just means the complexity of the overall quantum computer
you need to do, you need to build to get to error free computation or at least massively reduced error rates is a lot less complex than
we thought it was five years ago. And we’re at the point now where several different frontier systems are convincingly hard to simulate.
So it doesn’t really matter how many GPUs you can scrape together, you can’t really simulate what’s going on, on some of the best systems
that are kind of currently being demonstrated.
And then the last thing that’s happening
is that there are multiple different platforms emerging and maturing at the same time. So for a long time you had maybe superconducting
qubits and trapped ions that were right at the front and I’d say probably in the commercial world, they’re still the largest
or most advanced systems that are available. But you’re also seeing photonics and neutral atoms maturing rapidly as well. So there
are multiple different paths emerging to get to large scale quantum computers. And all are kind of making simultaneous progress, which
to me at least is very exciting.
But at least from my perspective, the hardware
alone is not enough. These things without software to run on them, without the software infrastructure to develop applications to take
advantage of them, they’re exquisite physics experiments, but they’re not yet really computers, they’re not yet programmable
systems that you can really turn to tackle hard problems. Of course, there are programming frameworks for them, but if we really want
to make use of these systems, we need to get to higher levels of abstraction, we need to make it much easier to develop applications for
them.
Where we sit as Horizon, I mean, it’s somewhat
self-aggrandizing, but I like to think we sit at the heart of the quantum computing ecosystem in that the software infrastructure we build
connects developers to end users and to hardware providers in the sense that we are building the programming languages and compilers for
developers to develop complex applications to take advantage of quantum computers. We’re building the queues in the runtime environment
to be able to execute those programs on hardware and we’re focused on supporting as wide a range of hardware as we possibly can.
And we’re building the deployment infrastructure so that when you finish – when the developer has finished creating an application,
they are able to wrap it up as a web API, to easily integrate into any user-facing application, into any user-facing interface, so that
their end users can make easy use of this without them having to manage the deployment infrastructure themselves, without them having
to stand up servers, and so on.
So as I see it, there are basically three barriers
to getting to really useful applications on quantum computers from the software side. Of course, there is also a hardware angle to this
where we need to build more accurate and more capable systems. We need to increase qubit count, we need to reduce the error rates in the
hardware.
But on the software side, there are real challenges.
The first, I would say, is that if you want to take advantage of quantum computing to solve a hard problem, you really need to take advantage
of quantum interference. If you are not taking advantage of this effect from quantum physics in your program, then your program could
have been run on a conventional computer, it could have been run more reliably and more cheaply on a conventional computer.
So you need to take advantage of this, this kind
of quantum physical effect. But this isn’t something that humans have an actual intuition for. In particular, we just don’t
interact with quantum phenomena in our daily lives, we don’t experience quantum interference. So the way we think about approaching
problems is much more like how we do calculations on pen and paper, things like this, which pretty closely maps how we approach tackling
problems with conventional computers. Just as with a pen and paper, if you are doing a calculation, you maybe look down at the page, you
read some information into your brain, make a simple manipulation, and then record the result. The CPU and a conventional computer is
doing much the same reading in a bit of information from memory, doing some simple manipulation on it, and then storing the result back
in memory.
So how are we going to take advantage of these
quantum effects in order to more efficiently process information? I would say the only proven way to do this is essentially to bang your
head against the wall for 10 years, trying dozens of different ways to solve a problem, maybe failing 99 times, and succeeding only once.
And that maybe gets you to a quantum algorithm, but it’s slow, right? It takes a lot of time to build up the expertise to do this,
it takes a lot of time to get good at doing it. The number of people that have been involved in more than two quantum algorithms is only
a few hundred people worldwide.
The next problem you have is that the hardware
is extremely diverse. So you have different kinds of platforms based on different kinds of underlying physics. In some cases, the instruction
sets are very different, there are different constraints in terms of how programs need to be constructed. And in some cases, even the
underlying physical model is different. Sorry, I should say the underlying computational model is different.
So, for example, in photonic quantum computing,
you really need to do all of the computation via measurements, because photons just don’t directly interact with each other. And that’s
a major difference from how you would approach it with a superconducting system or a trapped-ion system, for example. And then the third
challenge you have is that the programming languages really lack abstraction. So by that, what I mean is that many of the existing frameworks
for programming quantum computers are ultimately piecing together a circuit, one logic gate at a time.
And maybe they’re leveraging the infrastructure
of Python to do that in a programmatic way. But it’s still like trying to design an application or to build an application by designing
an integrated circuit to implement it. If you had to put together your applications one NAND gate at a time, it would be very challenging
to build anything that looks like modern computer software. So at least in my view, to get us to the point where quantum computers are
solving hard problems that are really creating value for end users, hardware’s only half the – is only half the solution. You also
need software to get there. You need to build the software infrastructure to enable developers to tackle these problems more efficiently.
So what we have been focused on at Horizon, essentially
since inception, is the question of can we make programming quantum computers more like programming conventional computers. And the real
reason we want to do this is because as I’ve said, there’s only a few hundred people that have shown any kind of success in constructing
quantum algorithms. But if you look at the number of people out there in the world that write code regularly, it’s a very large number.
So, if you were to look at the number of active GitHub accounts or something like that, you’d see this as many tens of millions of people.
So in that situation, you think how are we going
to approach this problem of building quantum applications for the domains that really depend on them. So – or that can really stand
to benefit from quantum computing. So domains like pharma, for example, domains like AI and machine learning, but also domains like finance
and the energy sector, because the people that are domain experts in those fields are not experts in quantum computing, and the people
who are experts in quantum computing are not experts in those fields.
So if you think about the way of trying to address
this, currently there’s maybe three different approaches. You can try to view it as an education approach where you try to teach people
more about quantum computing and hope they become able to develop their own applications. You can view it as a libraries approach where
you kind of pre-program some quantum applications and make them accessible through a Python library or something like that, so that it
becomes easier to access them, easier to apply them. But this maybe isn’t so flexible in terms of allowing people, allowing developers
to tackle new problems in quantum computing that maybe haven’t previously been explored.
And the third approach you have is kind of a professional
services approach where it becomes a consulting problem. You get together a room full of experts in quantum computing and then you approach
the potential end users of quantum computing and look to solve that problem for them. And that can work fine, it can make a lot of sense
for high value problems. But the reality is it’s not very scalable, it’s not how most modern software is developed.
So from my perspective, if we can make this much
more like programming conventional computers, then we get to a point where we’re really able to do a lot more. We’re able to open up quantum
computing to a much wider audience. So what we’ve been trying to do is to build a pathway from programs written for conventional computers
all the way down to an accelerated implementation running on quantum computers.
That means we’ve had to put together the technology,
we’ve had to develop the techniques to automatically construct quantum algorithms from code written for conventional computers. And we’ve
had some success with that. We’ve had some standalone demos, we’ve some patents granted on this process, but we’ve also had to build the
compiler stack and the languages to go from a very high level description of a quantum program all the way down to an concrete implementation
on hardware. And that has meant overcoming many of the limitations of existing systems. Many quantum computers that are available today
are not capable of doing things that you would take for granted on even the simplest conventional computer. They’re often not able to
do loops, for example, and when it gets to things like dynamic memory allocation, that’s completely non-existent in quantum computing.
So this process of going from a very high level
description of a quantum program down to implementation on hardware is what’s available today in Triple Alpha that we work with
mostly hardware partners with the systems in early access.
So our users are really mostly from the hardware
companies and we’re working with them closely to see what we can do to better support their hardware with our tools. Because our
focus is really on getting to a real concrete advantage or we’re solving hard computational problems as soon as we possibly can.
As I say, what we’re trying to do with our
system, with our programming languages and with our runtime environment is to extend the capability of hardware systems. So if you look
at many of the systems that are available today, current generation systems are often not able to at least in their commercial offerings
are not able to do things like mid-circuit measurement or control flow, or at least those features may be poorly supported or not uniformly
supported across hardware providers.
And once you get down to things like dynamic memory
allocation, which is usually something that the operating system takes care of for you, finding the space to store a variable for your
program that’s not implemented at all. That has to happen at compile time in quantum computing. And that only works when you don’t
have while loops. So it only works because we don’t have very capable quantum computers today.
But when you access these same systems through
our tools, you’re able to express programs that make use of all of these features in our languages and you’re able to execute
those programs on hardware today using our runtime environment. We’ve also become, I believe, the first quantum computing company
– sorry, the first software company, I should say, to start operating its own quantum computers. So we have a hardware testbed in
Singapore where we’ve been working to integrate our software stack very tightly with the control systems, so that we can get as
close as tight as efficient an implementation of quantum programs as possible. So that we’re minimizing the overhead to be able
to do things like these loops, recursive function calls, to be able to implement things like dynamic memory allocation in real time on
the system.
Where we stand as a company, our business model,
what we’re trying to do is essentially focused on building not just the infrastructure for developing programs, but also for distributing
and executing them. So applications essentially run through our infrastructure and that puts us in a position where we can charge in a
kind of AWS like way, where we’re charging based on the usage of the resource that the program is running on and the value of the
resource you’re accessing, which is I would say a very familiar model to anyone that builds web applications today.
And it’s kind of aligned with value creation
for the developer in the sense that if you build an application and you use it only once or twice, there’s no significant cost to
that. But if you have an enormous success, you build an application that is creating real value for the end users, it’s getting
used all the time, then hopefully we’ve created a lot of value for you as the developer, and it’s fair to charge accordingly.
So I will leave it there. But just to recap, what
we are focused on as a company is really trying to become the default software layer through which quantum applications are developed
and deployed. And we’re trying to do this in a way that insulates the developer from the risk of changes in hardware, so that they
do not have to place a bet on which hardware is going to win the race to becoming large scale, low noise quantum computers.
Thanks very much. I’ll leave it there and
happy to move on to the questions.
<<John McPeake, Analyst, Rosenblatt Securities>>
Thanks, Joe. That was great. I am asking every
participant in the conference today, I’m going to ask every participant in the conference today the same question and see what your
answer is with two time horizons 2030 and 2035, if one, obviously we’ll talk about it in terms of relative to 100%. How likely do
you think by 2030 and then by 2035? And if you have another date, you want to tell me, that’s fine too. How likely do you think
it is that we’ll have quantum computers that are doing something that is actually commercially productive by 2030 and then by 2035?
And again, if you think it’s longer than that, let me know.
<<Joe Fitzsimons, Chief Executive Officer>>
This is a great question. And actually I usually
prefer not to give my own answers on this because I think anyone person is going to be off, I mean, some people might get it right, but
there’s a lot of unknowns in this. There’s a really – in my view, there’s a report that’s very useful that’s written each
year for the Global Risk Institute by two professors in the area, Michaela Mosca and Marco Piani, that tries to anticipate the timeline
to quantum computers. The report’s mostly focused on when do we expect to see quantum computers that can break cryptography in particular
RSA 2048. And so what they do is they go out and they ask about 50 different experts in the field, they send them survey and it kind of
asks you to give your confidence interval for seeing a quantum computer that could break RSA2048 in 24 hours in different time bins.
<<John McPeake, Analyst, Rosenblatt Securities>>
Yes.
<<Joe Fitzsimons, Chief Executive Officer>>
And so they collect that information. I find that
really useful. And you see the crossover there. If you look at that, the majority opinion goes from thinking it’s less likely than not
to more likely than not in the 10 to 15-year window. And if you stare at it really closely, it looks like maybe it’s around 12. At least
that’s how it appears to me. But what I would say is that’s kind of last year’s report, they also asked the question of when are we likely
to see commercial applications of quantum computing? And there the crossovers in the three to five-year window. So that will be ahead
of 2030 for first commercial applications.
And as I say, that’s last year’s report, so two
to four years. And I would say I’m pretty aligned with that, probably a little bit on the optimistic side. So I would hope to see things
maybe a little bit sooner than that. The other thing that has happened since then is that in the year since the report, we’ve seen maybe
four different claims to have achieved a meaningful application of quantum computing. And of course, whenever a claim like this comes
out, it takes some amount of time for the community to come to a consensus as to whether that has actually been achieved or not. And at
least in my view, this is a little bit like if you look at computers playing chess or playing Go, that you have this gray period where
it’s not clear whether the computers are better or worse than humans, right?
So there’s a long period which humans could always
beat computers at chess, could always beat them at Go, and then there’s a gray area, Deep Blue beats Kasparov or AlphaGo beats Lee Sedol,
where you get into a point where there’s argument, right? There’s no clear consensus. People are saying, well, just on the day, and maybe
if it was a different player, it would have been a different outcome and so on. But then you get, you fast forward a year or two and it
becomes unambiguous. It becomes very clear that – I mean, it’s very clear today that computers are better at both of these games
than humans are.
<<John McPeake, Analyst, Rosenblatt Securities>>
Right.
<<Joe Fitzsimons, Chief Executive Officer>>
I mean, unbeatably so. I think we will see that
with quantum computing as well. I think we’re entering that gray area where there will be some back and forth and discussion as to whether
it has really been achieved or not. And I don’t think there will necessarily be a very clear example, but we will move on pretty quickly
to a point where it is very clear that quantum computers are beating conventional computers for specific tasks.
<<John McPeake, Analyst, Rosenblatt Securities>>
So if I parse your response there. So 2030, you’re
optimistic that by 2030 we’ll see a commercially advantageous application somewhere?
<<Joe Fitzsimons, Chief Executive Officer>>
Yes, I think so. And I would probably be more
optimistic than that, but there’s a lot of variance on that, right? There’s a lot of uncertainty because you’re making predictions about
things that are not yet known.
<<John McPeake, Analyst, Rosenblatt Securities>>
Well, remember, I’m a financial analyst and I
make predictions every day. I’m forced to operate in an environment where I’m lucky if I can get to 52% and then I still have to make
decisions or make recommendations. And that’s just unfortunately the world we live in. So I appreciate your – and would that be
most likely probably in chemistry, because I don’t think breaking security is necessarily constructive. It could be commercially viable
and advantageous for a three letter agency or a sovereign or something that, yeah.
<<Joe Fitzsimons, Chief Executive Officer>>
Chemistry is significantly easier than cryptography,
that’s right. So if you think about Shor’s algorithm, really you’re talking about billions of logic gates to be able to implement
it.
<<John McPeake, Analyst, Rosenblatt Securities>>
Right.
<<Joe Fitzsimons, Chief Executive Officer>>
So it is – you need quite a mature quantum
computer to be able to do that. There are chemistry applications that come in at fewer cubit counts and at fewer gates. So certainly you
would expect to see maybe chemistry applications beforehand. Of course, there are other areas that people are interested in, condensed
matter physics and particle physics. I mean, I would be shocked if we didn’t see a pretty clear advantage for some of the calculations
that we in quantum computing do all the time to figure out how to build better quantum computers and better software for quantum computing.
I mean, I know from my own experience building software for this that there are a lot of places where we could harness quantum computing
within our own software stack that would make it more efficient.
It’s a little way off. We’re not –
the hardware is not quite at the point of being able to realize that yet. But I think this is a unique position we have as a software
company that we can start to see where in our own tool chain we will be able to make use of that. So I think you’ll see early applications,
chemistry, things like that, things that are close to quantum mechanics are most likely to be the first ones, I would say.
<<John McPeake, Analyst, Rosenblatt Securities>>
And you mentioned the circuit depth on Shor’s.
I know that’s a kind of a moving target depending on the parallelism and the algorithm and that type of thing. I’m just curious
in the chemistry realm, how useful depths, what are they like? I don’t recall.
<<Joe Fitzsimons, Chief Executive Officer>>
It’s different. There’s a whole range
of different chemistry algorithms. There’s also this separation between types of problems. So we know that you can efficiently simulate
like molecular dynamics, any kind of dynamics efficiently with quantum computers. Computing spectra, on the other hand, like working out
lowest energy states and things like this is not something that you’re guaranteed to be able to do efficiently on a quantum computer.
And indeed they don’t always get populated
in nature. So you can make certain kinds of molecules, certain kinds of materials that don’t cool efficiently, so they get trapped
in these long lived metastable states and they just don’t reach their ground state. And so although you can efficiently simulate
whatever you could see in the lab, it’s much harder to make predictions about progress on spectra in general, for example.
<<John McPeake, Analyst, Rosenblatt Securities>>
Right. Okay. So let’s talk a little bit
about your company versus generalities. I apologize there, but I’m just asking all the panelists today because I want to see what
people are thinking about the time horizons. How do you choose who to work with? I mean, you have limited resources and you have a scarce
resource, which is software talent in quantum, triple-alpha. I guess, you have early access requests from more than 40 major corps, 80
universities, 10 quantum software companies, national labs. How do you prioritize and how do you make sure you get leverage, right? Because
you don’t want to have 16 instantiations of work that doesn’t scale over time and has to be supported.
<<Joe Fitzsimons, Chief Executive Officer>>
Yes. So at the moment our focus is on trying to
get to a real advantage as soon as we possibly can. And that mostly means working with the hardware companies directly. You also want
to make sure we’re in a situation where we’re able to serve the broadest range of customers as possible. And that’s
also kind of aligned with working with the hardware companies first because ultimately it’s the combination of hardware and software
that gets you to a real solution to these problems. So we certainly want to make sure that we are enabling the hardware companies to succeed
as much as possible.
And in some sense their success helps us as well.
So that’s where our main focus is at the moment in terms of how we prioritize things. We’re really interested in what paths
there are to large scale quantum computers in a reasonable timeframe and to low noise systems. Our focus is really on the kind of error
corrected side of quantum computing, the low noise regime rather than on variational techniques. So that’s where our kind of alignment
is.
<<John McPeake, Analyst, Rosenblatt Securities>>
I got you. And people can look at – look
to Bill Gates and say, this is someone that made a fortune on software. And the hardware companies could certainly look to Bill Gates
and say he made a fortune on software. While the hardware companies, particularly in the PC space, didn’t make a lot of money. Really
their margins are much lower and they’ve been commoditized on the hardware side of the x86 market at least, I’m going somewhere
with this. And when I’m old enough to remember back when many computers had – everyone had their own operating system. It
could be a Prime computer, Wang computer or whatever it might be. And even some of the workstation companies, Sun had their own OS, although
it was a flavor of Unix. So the hardware companies, I would think see software as maybe where the value could get created. How do you
convince them that it’s a good path to let you and Horizon Quantum kind of own that for them and get them to market faster? I’m curious
what those conversations are like.
<<Joe Fitzsimons, Chief Executive Officer>>
Well, look, I would say, if you look at what happened
historically in PCs, for example, as you made that analogy. It’s not just Microsoft that did well. So out of the Wintel alliance, both
Intel and Microsoft did extremely well. So there you have both the hardware companies and the software companies in a mutually beneficial
relationship where the software is driving adoption of the hardware, and the hardware is what enables the software. So it’s a very synergistic
relationship in that sense, I would say. Certainly hardware companies have software efforts too. I mean, in some sense they’d be crazy
not to. But it’s very hard to do two things well and to be focused on two different parts of the stack. At the same time both when you’re
in this enormous struggle to get to full tolerant computing to get to these large scale, low error rate systems where your focus really
has to be on driving the performance of the physical systems.
And on pushing that side forward, and also be
focused on, hey, what does the future of quantum computing software look like? So I would say, what you see with a lot of hardware companies
is that they definitely have focus on the hardware where, of course, their focus should be, but are also interested in how do you enable
program of that hardware and how do you get to applications running on top of that. And I think ultimately, we all recognize the importance
of getting of software and of getting to a real quantum advantage as soon as possible. And I think, if you’re given the option of getting
to a real quantum advantage sooner, that’s a winning proposition in most cases.
<<John McPeake, Analyst, Rosenblatt Securities>>
That makes a lot of sense. And along those lines,
how much can you accelerate, like I’m sure you go, I would imagine, I would imagine you go into some quantum hardware company that calls
themselves full-stack and they may have a software team there that’s somewhat overlapping with what you do. How much can you accelerate
the time to market versus an internal team? I mean, it’s hard to make generalizations, I bet, but…
<<Joe Fitzsimons, Chief Executive Officer>>
These kind of things are often collaborative.
So what I would say, is normally we’re focused on the stack. Although with our test-bed system, for example, we go all the way down to
the control systems and to the shape of the pulses that are getting sent to the processor. We extend up in the stack far beyond pretty
much any other effort in terms of building up higher levels of abstraction. So we’ve just introduced Beryllium, which is an object oriented
programming language. And most quantum programming approaches up to now have really been at the level of circuits, at the level of manipulating
quantum circuits, even if there are adaptive circuits where there’s some amount of executing, some operations conditioned on a measurement.
By pushing up this way, by enabling general control
flow, by getting to an object oriented programming language, we get to unlock a kind of network effect around quantum programming. Because
you can start to build libraries. Because with an object oriented programming language, you have the idea of objects and classes. You
can start to define objects, you can start to define classes that represent specific kinds of information built up of lower level objects,
built up of qubits and numbers and things like this.
And then you can build new classes on top of those
that are using the classes you’ve previously defined. So you get to kind of leverage up to get to higher and higher levels of abstraction
so that you move away from talking about, how do I represent information in qubits and manipulate it and shuffle them around the chip,
to moving up to expressing your program much more in terms of how am I manipulating different types of abstract information?
And that that’s getting to a level far beyond
where most hardware companies are focused today. But it gets to a point where it becomes much easier to build complex applications, and
where you do have this network effect where the developer maybe does not have to make a bet on which hardware is going to win and they’re
able to build complex applications and execute them on a wide range of hardware.
<<John McPeake, Analyst, Rosenblatt Securities>>
I love the approach, actually.
<<Joe Fitzsimons, Chief Executive Officer>>
Oh, thank you.
<<John McPeake, Analyst, Rosenblatt Securities>>
Got you. So the business model at scale, I would
think looks like a enterprise software model, infrastructure software model with fairly high gross margins and fairly high EBIT margins.
Is that the way I should think about it?
<<Joe Fitzsimons, Chief Executive Officer>>
So what I would say is that there are –
this depends to some extent on the way in which the quantum computers are accessed. So there’s – I would say at the moment most
quantum computers that are used commercially are cloud based.
<<John McPeake, Analyst, Rosenblatt Securities>>
Yeah.
<<Joe Fitzsimons, Chief Executive Officer>>
So people are accessing them over the internet.
And it makes a lot of sense to operate in an AWS like model. There not necessarily where you’re operating the hardware yourself, but where
you’re charging based on the usage of your APIs for the deployed program. Then you’re able to charge kind of proportionate to the value
of the resource that’s being accessed and the amount of usage of that resource.
<<John McPeake, Analyst, Rosenblatt Securities>>
Sure.
<<Joe Fitzsimons, Chief Executive Officer>>
As you move over to on-prem deployments, I would
say there’s a pretty well-trodden path there with products like Windows Server and VMware where you’re charging kind of annual recurring
licenses, but you’re doing it in a way that is again tied to the capabilities of hardware, in the sense that you’re charging based on
the number of cores, for example. So I would say there’s pretty well established software pricing models in each of those categories.
<<John McPeake, Analyst, Rosenblatt Securities>>
And in terms of the cost to provide the service
though, it’s a software like model, right? I mean your COGS, if you will, the cost to provide revenues…
<<Joe Fitzsimons, Chief Executive Officer>>
Yeah, absolutely.
<<John McPeake, Analyst, Rosenblatt Securities>>
Yeah.
<<Joe Fitzsimons, Chief Executive Officer>>
Absolutely. I mean, yeah, when it’s third-party
hardware, absolutely, it’s a – we’re providing software, we’re providing the server. Well, I mean we’re running it on servers, but
yes, we’re just providing a software there. We operate our own testbed system as I’ve said in Singapore, we anticipate making that also
available through our tools. But even with that, our CapEx on that, the components were about $2.3 million, something like that. Sorry,
that’s a pretty rough number. But to piece together that system which is very different from the kind of CapEx you have if you’re a hardware
company and you’re designing your own chips, for example.
<<John McPeake, Analyst, Rosenblatt Securities>>
And in this type of model you really want to get
developers at an early stage and get them used to your tools. Are you trying to get into educational institutions and have you considered
open sourcing any components of this sort of the way IBM did with Qiskit, curious?
<<Joe Fitzsimons, Chief Executive Officer>>
So I mean certainly, educational institutions
are important to us. We have not yet added academic institutions to early access, but that is something that we would anticipate doing
in the future. We’re participants in a number of networks including for example the Trinity Quantum Alliance in Dublin where we’re –
where Trinity College in Dublin has been running a master’s programming in quantum technologies. And…
<<John McPeake, Analyst, Rosenblatt Securities>>
Yeah.
<<Joe Fitzsimons, Chief Executive Officer>>
So we’re definitely engaging with academia, I
would say on that side. We haven’t quite added them to early access yet, but it is certainly something we would anticipate there.
<<John McPeake, Analyst, Rosenblatt Securities>>
And is it something where you want the product
to be more mature. Before you do that, what are you thinking about there?
<<Joe Fitzsimons, Chief Executive Officer>>
So initially our focus has been on working with
the hardware companies because that’s where we are actually able to overcome technical problems to make those systems more useful. And
that has been the first step for us anyway.
<<John McPeake, Analyst, Rosenblatt Securities>>
And that they pay you to do that, is that the
way I should think about that?
<<Joe Fitzsimons, Chief Executive Officer>>
With the hardware companies, no, that’s not our
intention at all. We want to ensure that we are the best software platform available to…
<<John McPeake, Analyst, Rosenblatt Securities>>
Okay.
<<Joe Fitzsimons, Chief Executive Officer>>
…be able to program any quantum computer.
<<John McPeake, Analyst, Rosenblatt Securities>>
Yeah.
<<Joe Fitzsimons, Chief Executive Officer>>
And we think we’ll win on the merits on that.
<<John McPeake, Analyst, Rosenblatt Securities>>
So that’s like R&D basically to attach to
those systems.
<<Joe Fitzsimons, Chief Executive Officer>>
Absolutely. So what I would say is that the most
important thing in quantum computing right now is getting to a real quantum advantage.
<<John McPeake, Analyst, Rosenblatt Securities>>
And is it your view that we’ll have multiple modalities
in the end? Or will it be suddenly PsiQuantum has a million gate computer – a million qubit – million perfect qubits in 2030
and everyone will be like, we have to use that and everything else will shut down. I’m curious, these are questions I get covering the
space.
<<Joe Fitzsimons, Chief Executive Officer>>
I think what? Well, it’s always hard making predictions,
particularly about the future, but what I find very exciting about the current time is that you have multiple modalities that are all
maturing in parallel. So the fact that you have a viable path for photonic systems, for trapped-ion systems, for superconducting systems,
and for neutral atom systems to scale – to scale up significantly and get to lower noise is really somewhat unique.
I think if you look to the far future and say,
well, is one modality going to win? I would imagine we would go through a period with hybrid systems before then. If you look at computers,
for example, they have, over the history of computing, different modalities have been used for storing information over different timescales.
So the technology you use to build a processor has been very different from the – for the technology you use to build long-term
storage like a hard drive or a tape or whatever it happens to be.
And I don’t really see a reason for quantum computing
not to go that direction other than the fact that quantum computing is difficult to connect components in this way. It’s difficult to
make trapped-ions talk to superconducting qubits, for example. But were that ever to be overcome, then it would make a lot of sense to
see hybrid systems. But there’s challenges.
<<John McPeake, Analyst, Rosenblatt Securities>>
Sure, I could talk to you all day, I have so many
questions. But do you think that the developers. Is it your view that developers, smart software algorithm writers will overcome or make
do with limited hardware faster than the hardware can deliver the qubits that are required to solve hard problems? Kind of like, the Gidney
Shore implementation, that type of thing. Do you see more of that happening?
<<Joe Fitzsimons, Chief Executive Officer>>
Yeah, I do. I think as the systems become real,
particularly if the error rate comes down, then you’ll see a lot of – you’ll see a lot of work trying to make programs fit within
existing systems. And you’ve seen this, this is like a proud tradition in the history of computing, right? Where there are all sorts of
tricks, like there’s this very famous – this very famous trick in graphics that there’s – there’s this peculiar constant that
appears in a bunch of code and it gives a shortcut to division.
And so you see a lot of these tricks emerging.
And Craig has definitely been at the forefront. I mean, if I was to think of anyone in quantum computing that is coming up with tricks,
it’s definitely Craig. So, yeah, I think that’s definitely the case. But interestingly, quantum computing has hit a kind of funny milestone
in that now the largest quantum computers have just eked past the number of bits in an Atari 2600. So it gives you some kind of sense
of where things are at. It’s less reliable, but about the right number of bits.
<<John McPeake, Analyst, Rosenblatt Securities>>
Yeah, and he’s a researcher. Wait till they have
your tools in the market more broadly available to all of the minds that are on planet Earth. I mean, I think these new technologies have
been doing this since the 90s. They always – they always have unintended consequence. There’s things that happen that I don’t think
about. I never thought the Internet would allow, I don’t know, food delivery or whatever.
I thought about it as a cheap pipe because I wasn’t
imaginative enough. So anyway, I would like – I didn’t realize this, but there was some access difficulties in the beginning of
the session. So I just want to let everybody know that the Meat Max platform that we use is nice in the sense that you can just go back
and listen to the replay fairly easily. So if anyone had problems getting on this session, you can go back and it has a DVR like function.
You can just listen to what you missed.
So, Joe, thank you very much for joining.
<<Joe Fitzsimons, Chief Executive Officer>>
Thanks, John. I appreciate it.
<<John McPeake, Analyst, Rosenblatt Securities>>
Yeah, very good.
About Horizon Quantum
Horizon Quantum’s mission is to unlock broad
quantum advantage by building the software infrastructure that empowers developers to use quantum computing to solve the world’s
toughest computational problems.
Founded in 2018 by Dr Joe Fitzsimons, a leading
researcher and former professor with more than two decades of experience in quantum computing, the company seeks to bridge the gap between
today’s hardware and tomorrow’s applications through the creation of advanced quantum software development tools. Its integrated
development environment, Triple Alpha, enables developers to write sophisticated, hardware-agnostic quantum programs at different levels
of abstraction.
About dMY Squared
dMY Squared Technology Group, Inc. (“dMY
Squared”) is a blank check company whose business purpose is to effect a merger, capital stock exchange, asset acquisition, stock
purchase, reorganization or similar business combination with one or more businesses.
Additional Information about the Business
Combination and Where to Find It
In connection with Horizon Quantum’s previously
announced business combination (the “Business Combination”), Horizon Quantum Holdings Ltd. (“Holdco”) and Horizon
Quantum have filed a registration statement on Form F-4 relating to the Business Combination and certain other matters (the “Registration
Statement”), which includes a preliminary proxy statement of dMY Squared and a preliminary prospectus of Holdco with respect to
the securities to be offered in the Business Combination. After the Registration Statement is declared effective, dMY Squared will mail
a definitive proxy statement/prospectus to its shareholders as of a record date to be established for voting on the Business Combination.
The Registration Statement, including the proxy statement/prospectus contained therein, contains important information about the Business
Combination and the other matters to be voted upon at a special meeting of shareholders of dMY Squared (the “Special Meeting”).
This communication does not contain all the information that should be considered concerning the Business Combination and other matters
and is not intended to provide the basis for any investment decision or any other decision in respect of such matters. dMY Squared, Holdco
and Horizon Quantum may also file other documents with the Securities and Exchange Commission (the “SEC”) regarding the Business
Combination. dMY Squared’s shareholders and other interested persons are advised to read, the Registration Statement, including
the preliminary proxy statement/prospectus contained therein, and, when available, the amendments thereto and the definitive proxy statement/prospectus
and other documents filed in connection with the Business Combination, as these materials will contain important information about dMY
Squared, Horizon Quantum, Holdco, and the Business Combination. The documents filed by dMY Squared, Holdco and Horizon Quantum with the
SEC also may be obtained free of charge upon written request to dMY Squared at dMY Squared Technology Group, Inc., 1180 North Town Center
Drive, Suite 100, Las Vegas, Nevada 89144.
Participants in the Solicitation
Horizon Quantum, Holdco and dMY Squared and their
respective directors, executive officers and other members of their management and employees, under SEC rules, may be deemed to be participants
in the solicitation of proxies of dMY Squared’s shareholders in connection with the Business Combination. Investors and security
holders may obtain more detailed information regarding the names, affiliations and interests of dMY Squared’s directors and officers
in dMY Squared’s Annual Report on Form 10-K for the fiscal year ended December 31, 2024, filed with the SEC on April 3,
2025 (the “dMY Annual Report”) or its subsequent quarterly reports. Information regarding the persons who may, under SEC rules,
be deemed participants in the solicitation of proxies to dMY Squared’s shareholders in connection with the Business Combination
is set forth in the proxy statement/prospectus for the Business Combination. Information concerning the interests of Horizon Quantum’s,
Holdco’s and dMY Squared’s participants in the solicitation, which may, in some cases, be different than those of their respective
equityholders generally, is set forth in the proxy statement/prospectus relating to the Business Combination.
Disclaimer
Past performance by Horizon’s or dMY Squared’s
management teams and their respective affiliates is not a guarantee of future performance. Therefore, you should not place undue reliance
on the historical record of the performance of Horizon’s or dMY Squared’s management teams or businesses associated with them
as indicative of future performance of an investment or the returns that Horizon or dMY Squared will, or are likely to, generate going
forward.
Cautionary Note Regarding Forward-Looking
Statements
This communication includes “forward-looking
statements” with respect to dMY Squared, Holdco and Horizon. The expectations, estimates, and projections of the businesses of Horizon
and dMY Squared may differ from their actual results and consequently, you should not rely on these forward-looking statements as predictions
of future events. Words such as “expect,” “estimate,” “anticipate,” “intend,” “may,”
“will,” “could,” “should,” “potential,” and similar expressions are intended to identify
such forward-looking statements.
These forward-looking statements involve
significant risks and uncertainties that could cause the actual results to differ materially from the expected results and are subject
to, without limitation, (i) known and unknown risks, including the risks and uncertainties indicated from time to time in the dMY Annual
Report, dMY Squared’s other filings with the SEC, and the Registration Statement, including those under “Risk Factors”
therein, and other documents filed or to be filed with the SEC by dMY Squared, Holdco or Horizon Quantum; (ii) uncertainties; (iii) assumptions;
and (iv) other factors beyond dMY Squared’s, Holdco’s, or Horizon Quantum’s control that are difficult to predict because
they relate to events and depend on circumstances that will occur in the future. They are neither statements of historical fact nor promises
or guarantees of future performance. Therefore, actual results may differ materially and adversely from those expressed or implied in
any forward-looking statements and dMY Squared, Holdco, and Horizon Quantum therefore caution against placing undue reliance on any of
these forward-looking statements.
Many of these factors are outside of the control
of dMY Squared, Holdco and Horizon Quantum and are difficult to predict. Factors that may cause such differences include, but are not
limited to: (1) the occurrence of any event, change or other circumstances that could give rise to the termination of the business combination
agreement, dated as of September 9, 2025, among dMY Squared, Holdco and Horizon Quantum (the “Business Combination Agreement”)
and/or the private placement of approximately $110 million of Holdco’s Class A ordinary shares (the “PIPE Transaction”);
(2) the outcome of any legal proceedings that may be instituted against the parties following the announcement of the Business Combination
and the Business Combination Agreement; (3) the inability to complete the Business Combination, including due to the failure to obtain
approval of the shareholders of Horizon Quantum and dMY Squared or other conditions to closing the Business Combination; (4) changes to
the structure of the Business Combination that may be required or appropriate as a result of applicable laws or regulations or as a condition
to obtaining regulatory approval of the Business Combination; (5) Horizon Quantum’s ability to scale and grow its business, including
through the use of proceeds of the PIPE Transaction, and the advantages and expected growth of Horizon Quantum; (6) the cash position
of Horizon Quantum following closing of the Business Combination; (7) the inability to obtain or maintain the listing of Holdco’s
securities on the New York Stock Exchange, the NYSE American, or Nasdaq following the Business Combination; (8) the risk that the announcement
and pendency of the Business Combination disrupts Horizon Quantum’s current plans and operations; (9) the ability to recognize the
anticipated benefits of the Business Combination and PIPE Transaction, which may be affected by, among other things, competition, the
ability of Holdco to grow and manage growth profitably and source and retain its key employees; (10) costs related to the Business Combination;
(11) changes in applicable laws and regulations or political and economic developments; (12) the possibility that Horizon Quantum may
be adversely affected by other economic, business and/or competitive factors; (13) Horizon Quantum’s estimates of expenses and profitability;
(14) the amount of redemptions by dMY Squared public shareholders; (15) difficulties operating Horizon Quantum’s quantum processor
and the possibility that the quantum processor does not provide the advantages that Horizon Quantum expects; (16) the ability to successfully
or timely consummate the PIPE Transaction; and (17) other risks and uncertainties included in the “Risk Factors” sections
of the dMY Annual Report, dMY Squared’s other filings with the SEC, and the Registration Statement and other documents filed or
to be filed with the SEC by Horizon Quantum, Holdco and dMY Squared. The foregoing list of factors is not exclusive. You should not place
undue reliance upon any forward-looking statements, which speak only as of the date made. Horizon Quantum, Holdco and dMY Squared do not
undertake or accept any obligation or undertaking to release publicly any updates or revisions to any forward-looking statements to reflect
any change in their expectations or any change in events, conditions, or circumstances on which any such statement is based, except as
required by law.
No Offer or Solicitation
This communication does not constitute a solicitation
of a proxy, consent, or authorization with respect to any securities or in respect of the Business Combination. This communication also
does not constitute an offer to sell or the solicitation of an offer to buy any securities, nor will there be any sale of securities in
any states or jurisdictions in which such offer, solicitation, or sale would be unlawful prior to registration or qualification under
the securities laws of any such jurisdiction. No offering of securities will be made except by means of a prospectus meeting the requirements
of Section 10 of the Securities Act of 1933, as amended.
