- il y a 2 jours
Quantum Potential: Leaping Your Business Ahead
Catégorie
🤖
TechnologieTranscription
00:04So, hello everyone, and very nice to have so many people for our session on quantum technology and quantum computing.
00:13I'm also very pleased to essentially have like three people with me, with whom we can also today discuss the
00:21respective aspects around quantum technology and quantum computing.
00:25Just sort of quick introduction from my side, an excellent partner from McKinsey from the office in Frankfurt, also leading
00:32our activities globally in the quantum space, so therefore super happy to be here.
00:38And sort of just as an introduction, and then I have three people with me that have a lot more
00:43knowledge than also on the respective technologies and the use cases around the technology.
00:48But we were also publishing our quantum report this year, and sort of just a few things that we've found
00:55out as part of this is quantum is most probably the next Gen AI kind of revolution that we would
01:00see on the technological side.
01:02We see sort of roughly like one to two trillion in terms of value that this could enable, and the
01:07creation of at least a $100 billion market on the quantum computing side, which is both very interesting and very
01:13nice to have also three people that are actually participating in the creation of the market than also today.
01:18What we also see recently is, and therefore it's also I think nice to have a few from the more
01:25established players on the quantum computing side, is that we really also see that the respective market is now maturing
01:31with sort of most of the respective investments that we see now in quantum essentially going into established quantum companies
01:37and not so much into further startup creation.
01:39So we really see also the professionalization of the market and then finally also I think this has spurred a
01:46lot of interest both by public and private investors and thanks heaven sort of we also see that the private
01:51investments after a bit of a downturn that we had seen in recent years sort of are essentially now coming
01:56back also to the right kind of stage.
01:58So I think this is a very good outlook then also for our discussion that we want to have today
02:03and I will spare you the individual introduction by myself because I think the speakers can also better introduce themselves
02:10and I also wanted to give them a bit of time in order to introduce also their respective companies because
02:15I think it's also very interesting to see sort of what are the different both technological as well as also
02:20business aspects on this one and therefore Jan would hand it over to you and then I'm happy to take
02:25it forward from there.
02:26Sure. Thanks a lot. Great to be here and actually great to have three different technology approaches on stage. I
02:31think that's wonderful. I'm representing IQM quantum computers. We are building full stack quantum computers based on superconducting technology.
02:39What this means in the end is we have silicon chips where we bring superconducting material on top and then
02:45we form this qubit there as actually electrical circuits, microwave circles and we use let's say around 5 gigahertz frequency
02:52radiation to control them and manipulate them
02:55and then read them out. So this is what we develop as the core technology and then the business that
03:00we build behind this is that actually we build a full stack computer around it also and the software integration
03:06and we either sell the computers as on premise devices.
03:10We have there are two approaches either as educational devices to universities but then also into high performance computing centers
03:17and we are I think one of the globally leading firms when it comes to actually deployments of those systems.
03:22We have deployed many of them in Europe but meanwhile actually also in Asia and we have been making deals
03:27in the US as well.
03:28A company started in 2019. We are now 300 people mainly headquartered in Finland but also strong team in Munich
03:35and then around the world we have people as well.
03:38You want to go next?
03:39Yeah, thank you very much. Actually, I'm representing Candela. Candela is also very similar to IQM but we use a
03:45very different technology. In our case, we are technologies based on the use of photons generated by semiconductors.
03:53So basically we encode quantum information on single particle of light photons. You can manipulate them super easily by optical
04:04fibers, photonic integrated circuits that you can fabricate in any semiconductor fab, especially we have many of them in Europe,
04:11which means that basically in our quantum computers, we have a lot of components fabricated here in Europe, Germany, France,
04:19Switzerland.
04:20And we integrate all these components. We assemble quantum computers. We have delivered quantum computers in some customers, OVH Cloud,
04:30CEA, EuroHPC as well.
04:32And recently we delivered machines in Canada as well. And it took us quite a small effort to install these
04:41machines because we succeeded in making them, I would say, easily integrable in any data centers.
04:48So because we don't need extremely low temperature for running these quantum computers, it took us about a few days
04:59to just come there in any data center, install the machines and run them up.
05:04On top of this, of course, we also have the software applications. These machines are available via our cloud partners
05:12and we make them open to fasten the innovation because now the challenge is for the communities to gather a
05:22community, gather academics and corporates to use these machines.
05:28And this is why we are convinced that we need to make them open to all.
05:35Thank you Henning and thank you for hosting today.
05:38And thank you to the organizers for putting together this magnificent event. It's my first time at VivaTech.
05:46So I'm Wasim Shiraz. I'm from a company called Quantinuum. We were founded in 2014.
05:53We were the first movers, I think, in this industry. It was founded as a spin-out from the University
06:00of Cambridge Accelerate program in the UK.
06:03I was part of the founding team. And since then, I was the chief operating officer of the company until
06:112021, when we acquired Honeywell Quantum Solutions to create a full-stack quantum computing company where we have hardware, middleware
06:20and application layers.
06:21We, in the early stages, we had investment from JSR, a Japanese chemicals company, which is now owned by the
06:32Japanese government.
06:33It's one of the largest suppliers of chemicals in the semiconductor industry. IBM invested and Honeywell invested.
06:41Since 2021, I've been responsible for the company's fundraising. We've raised about $750 million of new capital, principally from Honeywell,
06:52from JP Morgan, from their balance sheet, from Mitsui in Japan, from Amgen, the US biotech company.
06:58And I think I'm missing one or two others. But it's a largely industrial shareholder base.
07:04And we are, by some measure, the leader in this space.
07:09Super. Thank you.
07:11And sort of, maybe, Wasim, sort of, first question from my side, and I'll just give it directly to you,
07:18would be, I mean, we just talked also about the recent deal that your company has also closed with Qatar
07:25and sort of the additional investments.
07:27And I think sort of many people ask themselves, okay, like, what are really the use cases for the technology?
07:32Why would people invest at this stage so much money then also into quantum? Maybe you want to elaborate a
07:38bit on this.
07:39Yeah. I mean, there's lots of areas to cover here.
07:42But I think the key point is that we are living in a very, very special period where, and I
07:49think we will surprise ourselves, you know, the next 10, 15 years will be far more faster than the last
07:5510 or 15 years.
07:56And there's a lot of evidence for that.
08:00And what we are experiencing today is a fundamental shift.
08:03And that fundamental shift is what we call generative quantum AI.
08:09And this is about bringing together three of the world's most powerful technologies in the world, high-performance compute, AI,
08:18and now quantum.
08:19So, in the past, we've had telephony, we've had microprocessors, we had the internet, mobile telephony, we've all lived through
08:28this, we're living through generative AI, and now these three technologies are coming together at once.
08:35And so, there's a huge amount of acceleration that's taking place.
08:41Now, just going into AI, I think AI is wonderful, we can do so many incredible things.
08:49We can take a huge amount of data, we can run training on that data.
08:55But the problem is that AI is still doing approximations based on probabilities.
09:02And what quantum computing allows you to do is go down to the atomic level.
09:06So, once you go down to the atomic level, you can start to understand why things happen in the way
09:11they do.
09:11And the data that you can generate has the precision, the accuracy, and the reliability, which you can then fold
09:21back into your training set in order to create new solutions, reduce your time to solutions, and tackle large-scale
09:30intractable problems.
09:32So, I think that's what we can now look forward to in the next year, the following year, and beyond.
09:40And just going on to the use cases you mentioned, I think I would highlight three key use cases, which
09:46will be the first beneficiaries, I think, of this generative quantum AI.
09:50I think the first, and these are very large-scale, the first is food security, photosynthesis.
09:58Nature does this all the time.
10:00The photovoltaic process that we have today is highly inefficient, highly toxic.
10:06The fertilizer industry, I think, consumes about 5% of global energy output.
10:12The cement industry is probably very similar.
10:15I think this decade we will start to understand the conversion of light to energy.
10:25The other key area which affects every single one of us is in healthcare.
10:30You know, and I read the McKinsey paper before coming here.
10:34The average cost of bringing a drug to market is more than $10 billion, takes a decade, has a 99
10:41% failure rate.
10:42I think we can now start to look forward to bringing drugs to market, not for the billions, but for
10:48the tens of millions in a matter of months rather than years, with better quality outcomes.
10:54And I think by the end of this decade, if we can start to understand DNA information, we can then
10:59start to, which I think is incredible, which is understand the cause of disease rather than treating the symptoms.
11:07That's a huge, huge market.
11:09And then the third one I would point out is sustainability.
11:13You know, we've been trying to sequester carbon from the atmosphere for decades now.
11:19And I think quantum computing in combination with high-performance computing and AI can help us or bring us closer
11:28to the holy grail of clean, unlimited energy.
11:31Next-gen batteries, better catalysts to solve the energy problem.
11:34So I think these three shifts are immediate and we'll see the development of these new industries in the coming
11:45decade.
11:46Cool.
11:47Thank you, Wasim.
11:48And I think sort of looking sort of at these great prospects in terms of use cases, sort of in
11:56many of the conversations that I at least have also with some of our senior leaders, it's always, yeah, yeah,
12:01Henning.
12:01But sort of, you know, sort of quantum computing is always 10 years from now and it will be the
12:06same in 10 years again.
12:07So why are you worried about the respective technology?
12:11And then I always need to explain sort of what would then also be the real evidence and the real
12:16progress that has been made recently that is actually substantiating that this timeline is not anymore 10 years from now.
12:22But maybe, Valerian, sort of, do you want to say a few words concerning also what kind of real progresses
12:26have you seen and what would essentially tell you, like, okay, this is not the next time when we need
12:31to say, okay, like, it comes in 10 years from now?
12:33Yeah, basically, I completely agree with Wasim.
12:37And I would like to add, we will maybe have surprises, maybe good surprises about the acceleration of appearance of
12:47large-scale use cases where everybody, like, you know, kind of a Sputnik effect on quantum technologies and it might
12:56be before the end of the decade.
12:58Or we can also have bad surprises about the competition between classical and quantum, we don't know.
13:04So the fact is, we clearly see an acceleration of the progresses in the field and the way we connect
13:15quantum computers, the quantum processors, the building block, with other technologies, GPUs, and HPCs as well, which it's the beginning
13:25of the connection between the different systems.
13:28And now, actually, we are developing a lot of new frameworks in order to increase this impact and open, finally,
13:41quantum technologies, quantum computing to AI communities, AI developers.
13:46So, for example, we just published a new framework, which is dedicated for AI developers that would like to implement
13:55quantum neural network.
13:56So, neural network would be based on quantum-inspired mechanisms, such as we would like to invite non-quantum users,
14:12non-quantum experts to test and to develop use cases on these machines.
14:17Because, like for continuum at Candela, we are completely convinced that AI and quantum will have strong connections together.
14:28And quantum will be used for the next generation of AI, and AI will be used as well in order
14:37to fasten and accelerate the development of quantum computers.
14:42This is already the case in our quantum computers. We have some machine learning techniques, such as we can do
14:47some error mitigation using some AI techniques as well.
14:52Thank you. Maybe, Jan, you want to elaborate a bit on this as well, since I think we've recently seen
14:58a number of announcements.
14:59We just discussed this also before, sort of, be it from Microsoft, be it from Google, be it also from
15:05others, sort of, how do we essentially differentiate a bit, sort of, what is the hype and what is the
15:10actual reality?
15:11Yeah, I think, also coming back to this question, how do we convince people that it's not always these ten
15:15years away?
15:16I think we need to show them clearly that the gap is closing.
15:20And there are two aspects on the high level. One is the efficiency of algorithms.
15:24So, kind of, what is the resources that a given algorithm needs to be executed?
15:29And, for example, just a few weeks ago, there was a breakthrough announced by Google.
15:33They were estimating the resources that you need for this code breaking, the RSA encryption.
15:39Previously, it was thought that you need maybe around 20 million physical qubits on the chip to run it.
15:44And they brought it down to 1 million. So, a factor of 20 increased just like that.
15:48So, this is one way to close the gap, is to make the algorithms more efficient.
15:52And the other one is actually, yeah, to develop and provide the hardware.
15:54And there, it's a lot about showing that companies are actually delivering.
15:59Because, of course, in the early stages, everyone was making promises.
16:02But now it's time to actually show and deliver.
16:05And many companies have hardware roadmaps out there.
16:09And in our case, the way we do it is actually by, of course, developing the technology
16:13and then deploying it to customers to do exactly what was Seam said,
16:17to connect it, make the connection between quantum, HPC and AI.
16:21So, to give you one example, we have been delivering a system to a big supercomputing center in Munich.
16:27And it's physically there in the computing center next to the racks full of CPUs and then GPUs.
16:32And now the users of these computing centers, they're running algorithms
16:36where actually they do a lot of work still on the software side,
16:39where it is about fragmenting then the problem into a part that goes to the GPUs,
16:44another part that goes to maybe CPUs, and then some part that goes to the quantum level.
16:48And one nice application there is, for example, the simulation of molecules
16:54and the chemical properties of molecules.
16:56And what people reach there right now, in our case, for example,
17:02is that they reach this so-called chemical accuracy.
17:04So, this is kind of the gold standard in chemistry of how good your result needs to be.
17:09And they reach this now for, I think it was an eight atomic molecule or something.
17:13So, this is where we are, and it's executed in a way that, indeed, part of the problem was done
17:17on the conventional computers and part of the quantum computers.
17:20And I think this is the way, in the end, to create trust amongst investors,
17:23amongst people here in the room and elsewhere, that actually the gap is closing,
17:28companies are delivering on the roadmap, and then we also see these kind of algorithmic developments as well.
17:34Cool. And maybe, Jan, sort of also getting a bit into, you mentioned the work that you're also doing
17:41with some of the supercomputing centers.
17:44You've been one of the quantum companies in Europe that has been very successful
17:49at also deploying a number of the respective machines,
17:51and we've also seen that there have been a number of investment rounds also into your company.
17:58Unfortunately, sort of this is a bit the odd one out currently on the European side.
18:02So, maybe you just want to elaborate a bit sort of also on what does it require in order to
18:09ensure
18:10also further investments, and sort of what is a bit sort of also what would need to happen
18:14then also on the European side in order to overcome a bit the challenge that we see
18:18that currently US companies, at least sort of the majority share of them, is growing at a faster pace.
18:24And, of course, I mean, this is also a little bit from your thing that I'm a strong believer
18:28in European technology and the future of Europe.
18:31And when we started the company, we have been thinking a lot also with policy makers,
18:36how can we make sure that it is possible to grow tech companies in Europe and from Europe?
18:41And, of course, we should not limit ourselves to Europe.
18:44I mean, we have global ambitions, but still, this is the core of what we do is in Europe.
18:49And the example that we took was basically SpaceX.
18:52And if you look at SpaceX in the beginning, they sold mostly to NASA and US government organizations.
18:58And most of those rockets that they shoot up, they exploded a few seconds after the start, basically.
19:03So, there was not much real use of the technology.
19:06And this is probably where quantum computing was at the day when we started this.
19:09But still, I think it's a good investment for governments to make to procure systems from companies like ours.
19:17Because these revenues then can be used to convince investors that there's actually real traction on the one hand side.
19:23But it also brings a lot of structure and clearness to the organization.
19:28Because most of our companies, you mentioned this, you're a spin-out from Cambridge and you're here from the Paris
19:33ecosystem.
19:33Also in our case, we are a spin-out from a university.
19:36So, we have mainly researchers working in our companies and they're not used really to this kind of commercial way
19:41of doing things.
19:42And the contracts that we made actually forced us that we define a clear product with clear specifications, with a
19:48clear delivery timeline.
19:49And then there's a contractual penalty if you don't deliver.
19:52So, I think it has lots of benefits.
19:54And from this perspective, this is now the way we have been using and there are great programs in Europe
19:59actually to advance procurements.
20:02And this is something that is now being taken actually also up in other regions.
20:05We see a lot of traction also in APEC and the US for this.
20:10Yeah, I would like to emphasize that actually based on this model, I think the US are in advance in
20:17terms of programs.
20:18You mentioned the NASA program and recently DARPA has been launching a lot of programs for quantum technologies.
20:25I don't know if Continuum is part of it.
20:27But I can testify that in France, six months ago, we started a procurement program piloted by the Defense Ministry
20:38on quantum technologies.
20:40So, this is basically a 10-year program called Proxima.
20:45This is like, I would like to say that this is kind of prototype.
20:49That's going to be a first step to bring us to develop useful quantum computers for defense, for the French
20:57defense.
20:58But that's not going to be enough.
20:59And we would like to invite European Union to consider and elaborate this kind of procurement program.
21:06And we are not talking about R&D here. We are talking about innovation programs like on the same model
21:14as what DARPA can do and what the US and NASA can do as well at European scale.
21:20And I think this is what's missing now for quantum, for space tech as well.
21:26But Europe should rapidly innovate on this dimension and maybe this is also a call for, for example, a second
21:36flagship or quantum act.
21:38I don't know how we can call it.
21:41Because now it is time to think and have our policymakers in Europe to consider what's going to happen in
21:51Europe for the second part of the decade.
21:55But I think sort of, Wassim, maybe sort of since we already referenced and also the respective activities also on
22:01the US side, maybe you also want to comment a bit sort of on how you also see like the
22:05current climate on the US side and sort of also the respective changes also with the initiatives that have now
22:11been put forward also by the Trump administration.
22:12Sure.
22:13I mean, but it's global.
22:15I mean, I think the, so every nation of any consequence today has a quantum computing program.
22:28And I think you have to dig deep to understand or to look for any other technology that or any
22:37single technology that has had this type of attention and funding by governments.
22:44And the reason for that is that because it's not just another mere application.
22:48This is a transformational technology that will have huge impact on every single sector of society and economy.
23:00And the interesting thing is that we, as I mentioned, there's this unprecedented shift taking place in high performance computing,
23:14AI and quantum.
23:15And it's coupled with this huge mobilization of capital.
23:20And so, you know, you're looking at tech spend and infrastructure spend in the US, which will hit, I think,
23:28a trillion dollars on an annual basis by 27, 28.
23:33And I think that is the pattern going forward.
23:38So, yeah.
23:40Thank you.
23:40I think, well, yeah, you had already elaborated a bit sort of also on the talent and the respective programs
23:48that would be required in order to build and also the quantum companies sort of.
23:52Do you actually see that we have enough talent and also available and sort of what would need to change
23:57in order to in order to get it?
24:01So, basically, for Candela, we are looking for new talents all the time.
24:07Now we have this is a real challenge, which is on our shoulders.
24:10So, the company is already 150 percent.
24:14So, it's quite large company.
24:16But, yeah, clearly, we are looking for good talents, especially in quantum.
24:23So, it's possible.
24:24It was easy for us because we are coming, we are spinning out from a hardware lab.
24:28So, we have a good knowledge, we have a good reputation in quantum hardware.
24:33And it's easy for us to recruit a lot of engineers, PhDs, postdocs to join the labs.
24:42The main challenges for the future is going to rely on the next architectures of quantum computers and how we
24:50can improve these architectures, how we will improve the way quantum computing, quantum computers will be efficient.
25:00You mentioned the fact that, recently, Google labs improved the algorithms, the show algorithms, such as, actually, we can decrease
25:10the resources.
25:12This is a challenge, actually.
25:14I think, basically, I'm convinced that we have still many things to be optimized and we don't have enough talents
25:23and researchers on this topic, especially on this topic in Europe.
25:26And, in France, I think this is the worst.
25:31We are good in hardware, but algorithms and theory, I think we could do better.
25:41This is why, actually, there is this initiative in France, which is Choose Europe, to invite researchers to come and
25:50participate to this technology.
25:51And there are a lot of funds and grants that are available here, especially in quantum algorithms and theory.
26:00So, yeah, it's half yes and no.
26:03We have good talents.
26:05We have resources.
26:06But, of course, we need to accelerate, especially on the topic of architecture, which is the case at Candela.
26:12We have a lot of open positions on this field.
26:15It seems, Jan, this is a topic that is also close to your heart, so maybe you just want to
26:20chime in.
26:21I mentioned already in the beginning that we are selling computers for universities.
26:24And, actually, here, this is not to get rich, in a sense.
26:27We really went down with a price point as much as we could, so that universities actually can afford it,
26:32because we see the pain on the talent side.
26:34And we want that people learn quantum computers on real machines.
26:38And we not only need kind of quantum scientists, but we need people also coming, then, from other disciplines and
26:44supporting this industry, because it's going to be a complete new industry.
26:47So we need, obviously, the computer science, the software architects.
26:51We need electrical engineers.
26:53We need people who actually produce the machines, ship the machines, who do all the service in the end.
26:58So that's going to be kind of a huge gap once the industry really picks up, and it's already now
27:04a challenge.
27:04And this is why we are dedicating a lot of effort on the education side.
27:07It's not only that we are actually having these computers for universities, but on top of this, we have developed
27:13a whole platform.
27:14It's called IQM Academy.
27:15It's free for all to just go there and log in and learn quantum in a playful way, because we
27:20want, actually, that everyone who has an interest and who gets excited,
27:23actually with very low barriers, can kind of at least get some level of education on quantum computing and then
27:29hopefully work in the industry later on.
27:33Maybe, Wasim, sort of, since you're also looking now sort of at scaling the company quite significantly and sort of
27:41then also need to recruit the respective talent, sort of,
27:44how do you see it also with this, with, of course, sort of, on the one hand side, sort of,
27:49we see a lot of researchers coming from the U.S. into Europe,
27:52but on the flip side, of course, sort of the respective conditions on the U.S. side are now much
27:56better.
27:57Sort of, how do you see the talent availability then also for the scaling of your company?
28:01So, just following on from the points, I mean, the quantum computing represents, I think, a extremely important and existential
28:11challenge for national security and economic competitiveness.
28:15And so, there's a huge amount of investment, as I mentioned before.
28:20We're living through a period of unprecedented capital mobilization in this space, and quantum computing is the core of that.
28:26I think you only have to look at the announcements made by Microsoft, by IBM, by the Treasury Secretary in
28:33the United States,
28:34and various others about the importance of this sector and the amount of investment that's required.
28:39Interestingly enough, the U.S. does not necessarily have the outright leadership as we've seen in technologies over the last
28:5050 or 60 years.
28:52And so, you know, Finland is a leader in cold temperature supercomputing.
29:01And you have other centers in Germany.
29:04The U.K., for example, has a tremendous tertiary education sector in quantum computing.
29:11There's talent in Japan, in China, in Russia, India, Australia.
29:16And what I see is quite a fragmented space where, although, yes, certain companies and certain countries might have leadership
29:25in certain aspects of the hardware,
29:27but if you look deep into the supply chains, including labor and supply of talent, that's truly global.
29:34So, how this will play out, I think, is going to be really interesting.
29:39Cool. So, I think this was a very interesting discussion.
29:45And I think sort of one of the deep beliefs that I at least have on the quantum side is
29:49that the best is yet to come,
29:50in terms of also the respective advent of the technology.
29:54So, one final question sort of also on the list from myself sort of which was about what is actually
30:01the vision,
30:02what will be enabled by 2030 and how your companies would also then work towards or would contribute then also
30:10to the vision that we would have for 2030.
30:13I'm not sure. Maybe, Jan, you essentially want to start on this one.
30:16I know it's a broad question, but please do.
30:18I mean, in a sense, of course, we are a strong believer in quantum computing.
30:22And I believe quantum computing will be, in a sense, everywhere, but not in a way that we see it
30:27or that we carry the quantum computer maybe with us.
30:29Because the way I see compute development in general is that we have these more centralized data centers.
30:35And then we, as humans, we interact through whatever kind of interface it is.
30:39Right now, we do already a lot on our phones and other things.
30:42Maybe in the future, we have some glasses on or we have a chip in the brain that then connects
30:46to the center where the actual compute power is.
30:48But I think, as Vassim was elaborating in the beginning, there are so fundamental use cases that will be enabled
30:56by quantum.
30:57And some of them will be on the kind of daily basis.
31:00There are things, especially the group of optimization problems that you can think about where in life you want to
31:06have something optimized.
31:07It's basically everywhere.
31:08It not only starts in the traffic jam here in Paris, but you can think along the lines.
31:13So, I think this is one thing.
31:15But what really fascinates me is that we will see use cases that we don't even think about today.
31:21Because the real power and the real promise of quantum computing is that these machines can compute something that no
31:27other machine will ever be able to compute.
31:30And I think we are probably not creative enough to see what's coming.
31:34Like, when the internet was created, I think no one had exactly this vision where we are today.
31:38And this is, for me, the really fascinating field that as now quantum continues to grow in a Moore's law
31:44fashion.
31:45And this is what I strongly believe in.
31:46There will be decades of development on the technology and acceleration of the compute power.
31:51We will actually see use cases that we don't even dream about.
31:53And that's what really fascinates me.
31:55And this is why it's so hard to predict what's going to happen in, whatever, 2035 or so.
32:01Thank you.
32:02Do you want to take it from me?
32:03We clearly see, I completely agree with Jan, we clearly see an infraction point in the quantum industry.
32:12We just run out from this phase, which is the first phase with the first machines that are available for
32:19academics,
32:20just showing that some technologies are available on the market.
32:24And now we are entering in a new phase for the coming year, which is at Candela.
32:30We are completely convinced by the utility of this phase that's going to be hybrid AI and quantum computing machines,
32:38where quantum computers will be relying on GPUs and classical processors.
32:44And we will be deploying applications on these machines.
32:48We will be deploying algorithms that are going to be maybe 90% on classical techniques and maybe 10%
32:57on GPUs.
32:57And the combinations of these will bring value, will be useful for research, and it will bring an acceleration in
33:06the topic.
33:06So that's going to be an intermediate phase before the fault-tolerant quantum computers, based on the fact that we
33:14will have more and more logical qubits.
33:17So recently we've just seen that logical qubits are now available.
33:23Now it's time to scale up with the number of these qubits, make them more accessible and controllable for users.
33:32So it's not for now.
33:33But we clearly see that that's going to be within three or five years.
33:38So that's going to be before the end of the decade.
33:41We believe in this.
33:42We believe in the fact that before fault-tolerant quantum computers, we will have intermediate machines based on an hybrid
33:50approach.
33:50And the challenge for this hybrid approach will rely on the fact that these machines will be available for users,
33:59and users won't have to take care about the quantum side.
34:04So there is a large challenge on the software part, on the programming framework,
34:09make it compatible with biters, scikit, such as actually users, developers.
34:15They need to completely forget about the fact that they're going to rely on a quantum processor,
34:22such as they will focus on the top, on the applications, on the mathematics, on top of this.
34:32So just to summarize, and I share the views on the panel, my view is that AI, not in its
34:40current form,
34:41but in its next generation form, will be transformational.
34:46And it will be transformational in the context of being powered.
34:51So HPC GPU is being powered by quantum computers.
34:56And I don't think we are years away from that.
35:00I think later this year, 2025, this autumn, at the end of the summer,
35:06we as a company, Quantinuum, will launch the world's first quantum supercomputer with 50 logical qubits,
35:14which will allow us to do things that have never been done before.
35:19And our first demonstration of this hybrid AI, high-performance compute,
35:24and quantum will be launched at the same time in the area of drug delivery.
35:30And there's a very strong likelihood that drug delivery using quantum computers,
35:36working with high-performance compute and AI,
35:40will change the basis upon which the entire pharma industry is based upon.
35:45And that is going to take place this year.
35:50It takes years to find the right delivery catalyst,
35:54and I feel this is the year that we'll see that shift.
35:57Then, in late 2026, early 2027, we will cross another important threshold.
36:04We don't need to know what it is. It's 100 logical qubits,
36:08which, in combination with high-performance compute and AI,
36:12will allow us to start to solve much larger commercial problems.
36:16So, I think there's a lot to look forward to.
36:19What I'm looking forward to, personally, is personalized medicine,
36:23new batteries, new chemicals, new food resources, all thanks to quantum.
36:29And I firmly believe that we are now moving into this accelerated scientific discovery age.
36:39So, the balance of this year and next year, I think, is going to be the most interesting,
36:44certainly in my lifetime. I'm sure for many others in this room.
36:47I don't think we'll feel the difference on a day-to-day basis,
36:50but I think things are going to change, and they will change forever.
36:56Maybe, Jan, you also want to elaborate a bit sort of on how would then also IQM contribute
37:01that also to the vision that you've alluded to?
37:04Yeah, I mean, the way we see it is that the compute power is going to grow,
37:09and, of course, we will be able to deliver bigger and bigger machines.
37:13Something that we have on our roadmap is actually to also work much more on the software side,
37:19because what we think what is needed now is actually a much bigger developer community to bring in,
37:25so people who don't have a quantum background but who are excited about developing software,
37:30compiler solutions. So, what we are going to do is actually we are going to be much more open
37:35and modular on the software side and create a platform there and a playground for developers
37:41to bring this crowd in because we strongly believe that the more people who contribute,
37:47the more development there is, and this will benefit, ultimately, everyone.
37:51So, I think there's a big need for this, and many people I talk to at events like this
37:56who have a background in this field, they ask me, okay, how do you program a quantum computer?
38:01How does it work? Do I have to understand quantum physics and this kind of stuff?
38:05So, we have to take away this feeling that quantum is something like black magic
38:09and nobody understands it because that's not the case. In the end, it's just a computer.
38:13You program it through Python or C++ or whatever you like as your preferred way of doing it.
38:19So, this is a big focus that we will have in addition to creating amazing things like with the logical
38:25qubits
38:26is actually to make it much more easier and much more accessible for people out there to use it and
38:31create adoption.
38:35Just before, we briefly discussed about the difference of capitals between the US and Europe,
38:41and this is a recurring topic. About the quantum side, I would like to say that it also relies on
38:47us,
38:48as European researchers and startups, to optimize everything.
38:53We must imagine, develop, and build resource-efficient computers.
39:00Myself, I don't believe in broad-force approach which can be proposed by some US players because, actually,
39:11they didn't take the time to optimize the architectures and to minimize the required resources
39:17because they got a lot of capitals. They are supported by funds with billions and billions of dollars.
39:25In our case, in Europe, we don't have the same ecosystems to raise billions in capital,
39:33but we have technologies, we have talents such as, actually, we can propose more efficient architecture,
39:44we can propose resource-efficient quantum computers, and we will make, actually, cost-efficient quantum computers available in the market.
39:51This is what I strongly believe. We don't have any choice. European quantum industry will exist,
39:59and it will exist by proposing the most efficient architecture and quantum computers available in the world,
40:05and we will, actually, have the better ratio between capital and performance.
40:13This is what I completely believe. We can be capitally efficient like this.
40:20Cool. Thank you, then, also, for the contributions to the discussion, also, here.
40:26I think, sort of, what we've heard, sort of, gives us a lot of encouragement that, sort of, the best
40:30is yet to come,
40:31and that we will not just see growing technology players on the U.S. side, but possibly also on a
40:37worldwide basis,
40:39and that this is not just the technological development and then, sort of, will have risk consequences,
40:45but, sort of, that this is a new technology that will really also enable future use cases that are beneficial
40:50for the world in its entirety.
40:53And I think, sort of, we got a number of very nice use cases that were mentioned around better drug
40:57delivery,
40:58also solving the food problems on this earth, also solving our sustainability challenges.
41:03So, I think it's a great outlook that we would have in terms of what quantum computing could then also
41:07enable.
41:07And I think, sort of, with the three ones that we are also participating in this discussion,
41:11we also should have confidence that it's not the next time that we will need to say,
41:15okay, like, quantum computing comes, like, ten years from now,
41:18and in ten years we will sit here with a different crew and have exactly the same discussion.
41:22So, with this one, I would like to thank you all also for participating in the event today
41:27and for sharing your thoughts on this very interesting topic.
41:30We would also thank you, sort of, for the time that you have also spent together with us.
41:33Thank you.
41:34Thank you.
Commentaires