WEBVTT
00:00:00.880 --> 00:00:09.679
Hey, my name is Eva, and this is Abelong Quantum, the podcast that breaks down quantum technology into real-world business impact.
00:00:10.320 --> 00:00:12.320
Here we make the complex simple.
00:00:12.560 --> 00:00:14.320
No equation, no overthinking.
00:00:14.720 --> 00:00:18.239
Just insight, innovation, and a bit of cute.
00:00:18.879 --> 00:00:22.079
So don't worry, you don't need a PhD in physics to follow.
00:00:22.640 --> 00:00:29.839
In each episode, I talk to funders, scientists, and investors about how quantum is reshaping the industry today.
00:00:30.000 --> 00:00:31.839
Not in some distance future.
00:00:32.479 --> 00:00:42.320
Oh, and if you hear a black cat pouring in the background, that's my co-host, very alive shorting as cat joining the conversation.
00:00:42.560 --> 00:00:43.759
Her name is Moon.
00:00:44.479 --> 00:00:46.079
Let's get started.
00:00:47.679 --> 00:00:52.560
Welcome everyone to the next episode of Blond and Quantum Startup Series.
00:00:52.719 --> 00:00:58.240
Today I have a pleasure to host Francesco Raffaelli, a CTO of Kats Quantum.
00:00:58.799 --> 00:01:01.039
Hello Francesco, how are you?
00:01:01.439 --> 00:01:03.280
Hi, nice to meet you and thank you very much.
00:01:03.359 --> 00:01:04.239
I'm very good.
00:01:04.640 --> 00:01:05.040
Great.
00:01:05.439 --> 00:01:07.599
It's a really pleasure to have you here today.
00:01:07.760 --> 00:01:15.040
And we're gonna talk a lot about cats, what you guys are building, and also a little bit more about the industry.
00:01:15.200 --> 00:01:25.519
And on the end, I usually ask, I have this tradition to ask my um my guest about some prediction, like how you see the industry in the next, I don't know, 5, 10, 15 years.
00:01:25.760 --> 00:01:27.359
So please be ready for that.
00:01:27.439 --> 00:01:29.280
Um, but let's start from the kats.
00:01:29.519 --> 00:01:36.640
So you guys building a chip-based quantum K distribution, which we in the industry called QKD.
00:01:37.120 --> 00:01:41.920
Um, and also, as I read, quantum random number generation.
00:01:42.079 --> 00:01:47.680
Um, and this is all embedded in the hardware integrated photonics.
00:01:48.000 --> 00:01:50.959
So, again, silicon chips, right?
00:01:51.040 --> 00:01:59.760
So, this is about the theory, what I read, what I learned, but as you know, our audience are usually enterprise executive and business people.
00:02:00.000 --> 00:02:06.319
So, from that two-liner definition, they probably would not understand anything.
00:02:06.560 --> 00:02:14.639
So, if you can please kick it off from explaining in a very plain English what you guys are doing.
00:02:15.120 --> 00:02:15.680
Sure.
00:02:16.080 --> 00:02:22.080
Basically, yes, at Cats, we are building, as you said, the chip-based uh quantum key distribution systems.
00:02:22.319 --> 00:02:42.000
What we mean by that means uh we have a quantum key distribution system, so we have these systems that are used to generate and send and distribute um uh optical signals that then are used um to distribute uh keys uh to be used for cryptography applications.
00:02:42.400 --> 00:03:00.800
Uh the main differentiator of CATS is that from day one, so since our time at the University of Bristol, we knew that uh the only way to make uh quantum key distribution scalable was to build it upon technologies that are scalable.
00:03:01.039 --> 00:03:13.439
And this is why uh um we focused uh it is we spent the last uh 10 years effectively developing systems where the core technology is a photonic integrated circuit.
00:03:13.520 --> 00:03:19.840
So most technologies, most QKD systems will have uh would have uh fiber optics, effectively.
00:03:19.919 --> 00:03:36.639
Uh these are um components that are uh assembled together, they are connected by optical fibers that are good for prototypes, but they are not scalable because it would take a day to set up one, and it would take a thousand days to set up a thousand days to set up one thousand.
00:03:36.960 --> 00:03:37.439
Exactly.
00:03:37.599 --> 00:03:39.120
And they're also big, right?
00:03:39.680 --> 00:03:52.719
Also big and they are also um quite expensive um uh because every single component cost a few thousand pounds, and these sum up to become a very, very nice bill of materials.
00:03:52.960 --> 00:04:04.080
Instead, what we have done at CATS, we have uh uh spent a lot of time and effort in uh translating this uh um generation and detection of photons uh into photonic integrated circuits.
00:04:04.240 --> 00:04:15.840
Basically, these are uh these photonic integrated circuits use the same technology that is used for um computers, is the semiconductor technology used for computers, telephones, and everything.
00:04:16.079 --> 00:04:38.079
And by doing so, we have created a technology that uh um effectively enable to um have miniaturized solutions, but most importantly, is a technology that once you have demonstrated the first batch of chips, you effectively have already uh a direct route towards volume production.
00:04:38.240 --> 00:04:47.519
So once we have a first batch of say 30-50 units, then it's just a matter of repeating the order and then you suddenly scale to the thousands of units.
00:04:47.680 --> 00:04:57.040
And the nice feature of that is that the cost is as opposed to with fiber solution, the cost is not linear and the cost decreases as the volume increases.
00:04:57.279 --> 00:05:00.480
And this is uh really the nice aspect of our exploration.
00:05:00.959 --> 00:05:08.319
So you're basically saying that in the next couple of years you may have a manufacturing similar to semiconductors.
00:05:08.399 --> 00:05:23.759
So, again, for summary, we have a small chip which is much more scalable because of the production and will be cheaper in the future comparing to those bigger rocks that right now need to be installed um somewhere in the hardware and they are pretty expensive, right?
00:05:23.920 --> 00:05:25.759
Yeah, so that's what I understood.
00:05:25.920 --> 00:05:28.959
Now tell us, Francesco, why do we need it?
00:05:29.279 --> 00:05:30.879
Why this is needed for?
00:05:31.360 --> 00:05:32.240
Yeah, sure.
00:05:32.399 --> 00:05:43.839
Uh the the main the really the the the the reason of existence of QKD is the fact that um uh computing power is evolving at exponential speed, as we know.
00:05:44.000 --> 00:05:49.600
Everyone knows about uh AI, but there is also a lot of people also know about quantum computing.
00:05:49.839 --> 00:06:00.800
Quantum computing is a technology that, amongst its big uh promises, has also uh the potential of uh breaking the existing um cryptography.
00:06:01.040 --> 00:06:15.120
Existing current uh cryptography is based on effectively mathematical problems that are very hard to solve by uh standard computers, but for which quantum computers unfortunately seem to be tailor-made.
00:06:15.199 --> 00:06:27.360
So quantum computers seem to be incredibly efficient, will be incredibly efficient in five to ten years' time in breaking these uh existing um uh cryptography.
00:06:27.600 --> 00:06:36.079
Quantum key distribution is one of those technologies that uh uh is understood to be secure against quantum computing attacks.
00:06:36.319 --> 00:06:38.319
And the reasons are several.
00:06:38.560 --> 00:06:50.560
Two main reasons are the first one is that QKD has this nice uh feature of being um able to detect any attempt of ease dropping.
00:06:50.639 --> 00:07:01.759
So if someone tries during the exchange of a key tries to gather information about the keys, a QKD system can detect this type of hacking.
00:07:02.079 --> 00:07:12.800
And the other aspect of QKD is that it fundamentally lives in a different dimension to the one that quantum computers can attack.
00:07:13.040 --> 00:07:18.480
Quantum computers attack software and QKD lives in optical fibers.
00:07:18.639 --> 00:07:31.360
So essentially, there are two substantial reasons for why QKD is important, and these are all rooted to the fact that quantum computers will be soon able to break existing um uh cryptography.
00:07:31.600 --> 00:07:52.160
There is also another technology which is called uh post-quantum cryptography, which is effectively, which is effectively um different new kind of algorithms that have been uh developed and they will um replace the existing ones for which there are not known quantum attacks.
00:07:52.399 --> 00:08:00.160
And therefore they are effectively a complementary, there will be a complementary um solution to QKD.
00:08:01.600 --> 00:08:06.959
These will have different um type of application, different scale of of deployment.
00:08:07.519 --> 00:08:30.079
Post-quantum cryptography will cover the majority of the cases, but it won't be sufficient for uh those situations where security is paramount, such as banking, critical infrastructure, such as um power plants and similar applications is where uh QKD uh will find its role to support the security.
00:08:30.480 --> 00:08:31.839
So the reason, yeah.
00:08:32.080 --> 00:08:45.759
Yeah, so basically, if I just got you saying is that for most of standard enterprise post-quantum cryptography, it can be enough because it's working on the software layer, which um normally quantum computers attack.
00:08:46.000 --> 00:09:06.240
But for crucial infrastructure, so we're talking here, I don't know, finances, banking, uh, a lot of people asking about crypto these days, maybe uh pharmaceutical, maybe defense, um, those companies and system will need on top of that um also the hardware protection, which is um QKD.
00:09:06.639 --> 00:09:11.120
Yes, the reason being that would they have both or only the hardware one?
00:09:11.440 --> 00:09:12.639
They would probably have both.
00:09:12.799 --> 00:09:28.960
The reason uh for why QKD is necessary because in a similar way as uh for which the current RSA protocols are being replaced, no one can guarantee that this post-quantum algorithm will be secure for how long?
00:09:29.200 --> 00:10:00.559
We know that the exponential development uh of computing power, which is not necessary, just the direct uh attack on the algorithm, as quantum computers will do, but also the detection of faults in how an algorithm has been written, are developing so fast that relying on a single uh software approach, which is not proven to be secure but deemed to be secure, is a too risky approach for the for certain critical industries.
00:10:00.879 --> 00:10:01.039
Okay.
00:10:01.759 --> 00:10:03.919
And last question on that topic.
00:10:04.000 --> 00:10:14.240
So, do you think um the QQD and PQC will coexist, or maybe one will dominate in 10 years?
00:10:14.399 --> 00:10:16.080
For example, they will coexist.
00:10:16.240 --> 00:10:45.919
They will coexist because there will be situations because the um what people maybe don't it's difficult to quite to picture, but the complexity of the networks is huge, and therefore there are several situations where uh one will be sufficient, there are several situations where the other would be more practical to deploy and faster to deploy, and therefore we expect that the future security landscape would be very heterogeneous with different solutions for different uh uh specific uh cases, so they will coexist.
00:10:46.240 --> 00:10:56.799
And and you mentioned um telephones as one of the examples, and I know that you have some very fruitful cooperation with some of the big telecoms.
00:10:56.879 --> 00:11:00.000
Maybe you can I will leave it to you to talk about it.
00:11:00.080 --> 00:11:09.919
Um, but I wanted to ask you in general, why should telecoms operator invest in QKD instead of the software-based encryption?
00:11:11.279 --> 00:11:12.399
This is a good question.
00:11:12.559 --> 00:11:23.840
So, the reason why um so telecom operators effectively offer are those that offer the service to the banks, to the health sectors, to the uh energy grid.
00:11:24.399 --> 00:11:36.240
And uh the reason is because as providing the service, they want to provide the most uh secure service, and this is a combination of uh QKD and MPQC.
00:11:36.399 --> 00:11:48.240
So uh again, they provide the service to the to the banks, say, and they will make and based on the specific case, they will uh implement one or the other or both solutions.
00:11:48.559 --> 00:11:56.240
I think that's and if you think about your place in the Spark cooperating with telecoms, for example.
00:11:56.320 --> 00:11:57.840
So, how do you place yourself?
00:11:58.000 --> 00:12:04.799
You're more like a telecom vendor or more like a semiconductor company, or maybe something else, maybe security company?
00:12:05.279 --> 00:12:06.639
Yeah, it's a good question.
00:12:06.799 --> 00:12:10.080
I think uh we are uh a bit of the these three.
00:12:10.320 --> 00:12:18.480
We did what we sell, we sell um systems, solutions that fit within the existing telecom equipment.
00:12:18.720 --> 00:12:28.960
So, what we sell, we sell our QGD system that might fit uh into uh the equipment that already exists in the telecom uh network and therefore that can be integrated.
00:12:29.039 --> 00:12:35.759
So, in that sense, we provide components for uh the telecom equipment um industry.
00:12:36.240 --> 00:12:40.639
We also uh sell uh a lower level down into the supply chain.
00:12:40.720 --> 00:12:56.080
We also sell uh simple components that provide a different level of integration where people might explore different applications that is not necessary, the standard ones, where they can take our own chips, maybe packaged, and develop their own uh solution.
00:12:57.120 --> 00:13:04.639
And these are it's like a development kit, yeah, kind of like a develop development kit with the core component, exactly.
00:13:04.720 --> 00:13:16.080
We can imagine our us selling the chip, but also a development kit, and therefore that can use can be used for different applications that is beyond what we have imagined, for example.
00:13:16.240 --> 00:13:16.720
Yes.
00:13:16.960 --> 00:13:20.240
And these are two different layers in the supply chain.
00:13:20.399 --> 00:13:34.799
These are all technologies that uh entail uh security, and therefore uh the kind of approaches that we use in developing our hardware and software stack are those of uh we try to follow those of HSMs and security hardware companies, yes.
00:13:35.039 --> 00:13:40.799
So there's basically two different products, depends on the need on the company that is a client.
00:13:40.960 --> 00:13:57.200
Uh, what's about the uh random numbers generators that yeah, and we have uh um as part of our journey we have developed uh uh random number generators, which was in fact my uh started off with my PhD thesis, which is uh from Bristol, again from Bristol.
00:13:57.440 --> 00:13:59.759
From Bristol, from Bristol under uh Jonathan Matches.
00:14:00.399 --> 00:14:06.799
And yes, the quantum random number generator is a very interesting device that again we developed on on chip.
00:14:07.039 --> 00:14:14.720
So we have a very nice little component that produces very high quality random numbers, and this can be used in a variety of solutions.
00:14:14.960 --> 00:14:39.120
Originally, a catch was developed to be to provide the the randomness, the entropy for our quantum key distribution systems, but also the quantum random number generators can find applications in several uh other solutions such as classical cryptography, but also Monte Carlo stochastic simulations, anything, effectively any solution that requires very high quality and provable randomness.
00:14:39.200 --> 00:14:42.559
Uh, that's what the quantum random number generator um can offer.
00:14:42.720 --> 00:14:50.320
And you mentioned your PhD and your ten spend in the University of Bristol, and we talked about this a little bit before um recording.
00:14:50.399 --> 00:15:06.159
I wanted to ask you like what kind of quantum dust ferry is in Bristol on that university, because you are um at least fifth person that I'm meeting during um hosting this show who is coming out of the Bristol University.
00:15:06.399 --> 00:15:11.519
Recently I spoke with Reza Najpati, who's the head of quantum in Cisco.
00:15:11.759 --> 00:15:14.480
Um, but I know there's not only him, there's the small people.
00:15:14.639 --> 00:15:19.120
If I'm not wrong as well, Psych Quantum had their roots in Bristol as well, right?
00:15:19.360 --> 00:15:20.159
Yeah, yeah.
00:15:20.320 --> 00:15:22.080
So what's going on there?
00:15:22.639 --> 00:15:26.559
Yeah, um, effectively, Bristol is an amazing playground for quantum.
00:15:26.720 --> 00:15:35.519
Um, I joined arrived in uh 2014 in Bristol when the quantum engineering and technology labs were literally exploding, blooming.
00:15:35.600 --> 00:15:51.519
Yeah, uh it was uh it started off with uh um John Rarity, and then again the Jeremy Bryan and Mark Thompson, the founders of PsyQuent, were those that grew the team very significantly, um, up to 150 people at the time I was there.
00:15:51.759 --> 00:16:12.240
And these uh has uh produced several startups, and in fact, uh this uh the Bristol Hub, thanks to the Quantum U Technology and uh Engineering uh Centre, produced also um probably one-third of the quantum startups in the UK actually somehow passed through Bristol in their in their journey.
00:16:12.480 --> 00:16:32.879
Yes, Bristol uh and the amount of the amount of expertise, particularly at the intersection between quantum and integrated photonics, is really what has made uh Bristol a special place because it was the intuition of the founders of the group that quantum was cool, but to make it a thing, it needed to go through photonic integrated circuits.
00:16:33.120 --> 00:16:51.200
And literally they invented the field of uh that's what happened, effectively invented the the field of uh photonic integrated circuits for quantum, and uh then now there are maybe yeah, seven or eight startups that still live in Bristol, plus several others, such as like quantum, obviously uh they've grown massively.
00:16:51.600 --> 00:16:57.440
No, it sounds like a very magical place because I can understand the hype for quantum, right?
00:16:57.679 --> 00:17:22.720
And I can imagine that many academic organizations or institutions may have some specific part of hype on different stuff, but you know, a building a culture where this research isn't applied into real life and commercialized, including so many startups, including say quantum, um, that's a different story, yeah, and that requires much more work.
00:17:23.039 --> 00:17:32.160
Um, so that's a magical, and I think you are very lucky that you were that time um inside of it, and you could observe that from close distance.
00:17:32.400 --> 00:17:40.160
So, let me ask you what needs to happen for Cats to become another one billion dollar company then?
00:17:40.640 --> 00:17:41.599
That's a very good question.
00:17:41.680 --> 00:17:51.759
I think uh we can say that the difference between, say, quantum computing and quantum cryptography, such as uh QKD, is that quantum computing really uh lives in green fields, open fields.
00:17:51.920 --> 00:17:59.519
There is nothing compared that compares to quantum computing, and quantum computing is applications that are still to be discovered.
00:17:59.839 --> 00:18:01.440
It's very different for QKD.
00:18:01.599 --> 00:18:13.519
QKD lives uh in a very highly regulated market, which is cryptography, in a very highly uh controlled market because cryptography obviously is at the core of our digital safety.
00:18:13.839 --> 00:18:33.680
And therefore, really that what needs to change for quantum companies, quantum cryptography companies to grow is uh the understanding of certain um agencies, governments, and regulatory bodies uh of the role that quantum communications, quantum cryptography can have in the landscape of cryptography.
00:18:33.839 --> 00:19:01.359
Therefore, um certainly uh what we need is uh what we need to get uh to the scale for quantum QKD companies, for QKD companies is really the collaboration with the uh security agencies, with governments, and such that they promote and they promote QKD where QKD can be the right solution and they effectively explain QKD in a way that is fair, genuine, and correct.
00:19:01.440 --> 00:19:21.680
Effectively, I think this is uh uh what we are seeing at the moment is probably that uh because cryptography historically has been a computer science slash math industry, uh, cryptography is was invented by the mathematicians, brought forward by uh computer scientists.
00:19:22.000 --> 00:19:31.039
Now that physicists and telecom um engineers enter the the narrative, there is a bit of uh lack of uh understanding.
00:19:31.359 --> 00:19:33.200
There are miscommunications.
00:19:33.680 --> 00:19:53.839
It is not clear uh what uh QKD can do, what are the limits of QKD that exist, as for everything, and therefore, really the need the work that needs to happen is uh stronger and closer collaborations with the exist between the existing solutions and who promotes the existing solutions and the novel solutions such as uh QKD.
00:19:54.240 --> 00:20:13.119
This is uh the step, the big step that needs to happen for a company like that to I'm glad you said that because I have heard a lot of good things about the supporting environment in the UK in terms of the you know procurement um processes and the agencies supporting quantum startups.
00:20:13.279 --> 00:20:18.000
I think I heard that from Michael Cobart from National Uh Quantum Computing Centre.
00:20:18.480 --> 00:20:23.759
So, right now, do you have any cooperation with those agencies in the UK?
00:20:24.079 --> 00:20:27.920
So just it is very important here to make a clear distinction.
00:20:28.240 --> 00:20:33.039
Quantum computing uh yeah, is seeing uh the good support within the UK.
00:20:33.200 --> 00:20:45.119
Quantum key distribution is not being supported at the moment by the um uh security agencies, and this is really one of the of the steps that needs to happen.
00:20:45.200 --> 00:20:57.680
Uh, NCNC needs to understand to have an up-to-date view of what UKD, um which now probably it is fair to say that now probably they have a slightly out-of-date view of QKD.
00:20:57.759 --> 00:21:10.319
The effort is really to understand where QKD really sits at the moment in terms of technology development, and therefore um this kind of uh uh and therefore understand where can be a solution within the landscape of cryptographies.
00:21:10.400 --> 00:21:15.039
So there is a big distinction in the sense between quantum key distribution and quantum computing.
00:21:15.200 --> 00:21:20.000
That's uh so really this is uh so it's true that quantum computing is supported massively.
00:21:20.160 --> 00:21:24.559
It is not quite QKD is supported but not quite as strongly.
00:21:25.039 --> 00:21:41.200
Thank you so much, uh Francesco, for saying it because I was not aware, and definitely by saying it publicly, we send a strong signal uh to the agencies like, hey, it's time, it's time to step in and have a look and support in the same way that you support quantum computers.
00:21:41.359 --> 00:21:43.839
So thank you, thank you so much for saying it.
00:21:44.079 --> 00:21:56.400
Um speaking of, I noticed during my research that actually CAT has a lot of um different RD grants and utilizing pretty um efficiently.
00:21:56.559 --> 00:22:05.440
Uh and I don't know if that was your A weird decision to go that way to utilize the grants rather than private capital.
00:22:05.519 --> 00:22:06.960
I know you have both, right?
00:22:07.119 --> 00:22:22.799
But what I wonder is what's the common or maybe the biggest misconception that you heard from the investors who are potentially interested in investing in company like Cats about quantum cryptography.
00:22:22.960 --> 00:22:24.079
What do you usually care?
00:22:24.400 --> 00:22:41.680
I think that the main the biggest misconception is really while QKD is a very complex technology, basically what we are doing, we are sending one photo at a time across um commercial networks, we are detecting them and we are establishing secure keys by this process.
00:22:41.759 --> 00:22:44.319
So this is a very complex uh process.
00:22:44.799 --> 00:22:52.720
But what people maybe have not quite understood is the progress that uh has been achieved by CATS and others in the development of these systems.
00:22:52.880 --> 00:23:15.359
If you think that we have uh a single chip on the transmitting side and a single chip on the receiving side that do all of this in the optical domain and they do it to the right following all a set of mathematical proofs that are very sophisticated, and we do it with standard semiconductor foundries using standard electronics and using standard computing solutions.
00:23:15.599 --> 00:23:27.680
So, what people have probably uh not quite understood is that QKD is not anymore in an RD phase, QKD is very much in an engineering development phase.
00:23:27.839 --> 00:23:41.359
We have demonstrated the fundamentals, uh, we are ready with our solutions to scale, and what we need is really the right environmental conditions for the technology to be commercialized effectively.
00:23:41.519 --> 00:24:02.880
Therefore, hence the importance of working with the governments, with the security agencies, because the pull will come from banks, health sector once there is a good agreement, when there is the agreement, one there is the with the with these uh companies that are those with these institutions that are those that set effectively the regulations.
00:24:03.279 --> 00:24:06.559
So, really, but the bigger misconception I believe is in the technology.
00:24:06.640 --> 00:24:12.960
The technology is there, is ready to scale, and what is missing now is the commercial pool.
00:24:13.279 --> 00:24:14.799
That's that's great that you said that.
00:24:14.880 --> 00:24:27.759
Um, because again, we do not have that many investors skills on the uh quantum technologies market, and as you say, many of them um are not really aware that the technologies are here, so this is cool.
00:24:28.160 --> 00:24:37.759
Let's talk a little bit of what happened in April this year because I think CAT just announced the company to the security hardened and scalable uh QKD prototype.
00:24:37.920 --> 00:24:40.640
Um, can you tell us a little bit more about this?
00:24:40.880 --> 00:24:43.920
This was designed for critical infrastructure.
00:24:44.160 --> 00:24:51.119
And is that mean that you are already there to have the fully integrated telecom grid chip?
00:24:51.359 --> 00:24:52.640
Please tell us more about it.
00:24:53.200 --> 00:25:25.200
Yeah, so certainly what we have at the moment is we developed this system which have uh these chips, and the nice thing is that now all just not just the chip, but all the systems that that goes around it the electronics, the compute have gone through uh a further level of development to uh for where effectively also aspects such as the software, such as the hardware have have been pro have hardened, as we said, to be ready to be um used uh across across networks.
00:25:25.359 --> 00:25:31.839
So this is uh another another point that people sometimes is sometimes people maybe don't quite understand.
00:25:32.079 --> 00:25:39.599
That again, QKD is an advanced technology, it's very difficult, but it's also advanced and is really to scale.
00:25:39.759 --> 00:25:48.319
The nice thing about our solution now is that really um again, we are at that one step away from deploying at scale.
00:25:48.559 --> 00:26:08.960
So we what if once the opportunity comes, our chips are designed uh in a particular uh silicon photonic foundry that already can could produce this system in the millions, and uh all the electronics as well, obviously, can be produced at a very high scale.
00:26:09.039 --> 00:26:23.200
So that's really the important thing of our solution is that in with this version that we recently completed, we have removed all the little imperfections, and now really we are one step away to make this the solution scalable really.
00:26:23.680 --> 00:26:26.079
Well, first of all, Francesco, congratulations.
00:26:26.319 --> 00:26:36.559
I don't think there's many um quantum technology startup that can say, hey, we are one step almost from massive production that our solution can be you know um manufacturing millions.
00:26:36.720 --> 00:26:44.319
That's huge, and that itself should be a really good trigger for the investor to cue the front of your head headquarter.
00:26:44.720 --> 00:26:54.240
Um, but I wanted to talk a little bit about the solution as well because I know you are um testing in live right now with BT, with the British Telecom.
00:26:54.480 --> 00:26:55.680
How's that going?
00:26:56.960 --> 00:27:04.640
No, uh I would say that the team at BT has been incredibly uh supportive throughout our journey.
00:27:04.799 --> 00:27:11.759
Uh, the research labs at BT has been supporting really from day one since back in the day, I think it was 2018.
00:27:12.240 --> 00:27:39.599
And uh it took uh several years of interaction, of patience from their side for us to get through all the challenges of developing this technology, and it's been uh um a very good opportunity uh for us to understand the telecom infrastructure, to understand the standards required uh for our solutions, and now uh this uh this system uh the the most uh um yeah, and now uh the things are going very well.
00:27:39.759 --> 00:28:07.119
The best example of that is a project that was called uh recently was called um QRSure, which also saw the participation of uh the team of uh the National Physical Laboratories and other uh entities across uh the UK, where basically uh as part of our objective of making a genuine um commercial solution, we put our uh QKD systems under uh third-party scrutiny.
00:28:07.279 --> 00:28:26.720
So, what we did, we actually put our uh our own hands off the system, we shipped our systems to MPL, we let they we let them play with the systems, and they validated that effectively our solution does what it says in the tin, which is really the important thing of cryptography.
00:28:26.960 --> 00:28:37.039
And uh BT indeed has been incredibly supportive uh in setting the requirements, supporting, steering the direction towards something that is actually what the telecom uh really needs.
00:28:37.200 --> 00:28:40.960
I mean, uh, incredibly grateful to the to the teams uh BT.
00:28:42.000 --> 00:28:44.960
This is fantastic, and congratulations uh for that.
00:28:45.119 --> 00:28:47.039
I think you were somewhere announced online.
00:28:47.200 --> 00:28:55.920
I I saw that you were announced the only one commercially um ready or very close to being commercially ready, uh, QQD solution in the world.
00:28:56.079 --> 00:28:57.680
So that's a big thing as well.
00:28:57.839 --> 00:29:08.400
Um, I wanted to ask you what you mentioned that makes me think as well about the production in millions um because of the standard on the standard on the semiconductor market.
00:29:08.799 --> 00:29:29.279
Do you think that there is a risk for you as a company that um in some time in the next couple of years, if solution you know be very successful, that the large semiconductor player can replicate what you're doing and then try to sell it as, I don't know, white label solution?
00:29:29.680 --> 00:29:32.480
Yeah, I mean that this is risk in industries.
00:29:32.640 --> 00:29:39.839
Uh we believe that uh uh there are two things uh um that uh make us confident that this won't happen.
00:29:40.079 --> 00:29:54.480
The first thing is that uh we have very good IP protection of some of the critical aspects of our technology, and this makes us confident uh that uh this provides sufficient uh barriers to entry.
00:29:54.720 --> 00:30:18.640
Also, the other interesting aspect is that beyond the IP, which is only part the protected IP, there is the tens, hundreds of secrets that go from designing the chip to integrate it with the right electronics to uh developing the software and everything because Cats is a full-stack solution from the chip to the software that controls the electronics that controls the chip.
00:30:19.440 --> 00:30:35.599
We believe that uh while it is possible for a huge company uh to replicate this, the huge company won't will still need to go through a very significant amount of development, and that's where um which will take will take years, anyways.
00:30:35.759 --> 00:30:42.319
Therefore, there is a first uh started advantage in this sense, yeah, and first mover advantage as well.
00:30:43.119 --> 00:30:46.400
Okay, and um let's talk and just a little bit about the cost.
00:30:46.640 --> 00:30:48.640
So, can you just tell us a little bit more?
00:30:48.720 --> 00:30:55.519
Because I think that may be a question of a lot of enterprise executives right now or listening to our discussion.
00:30:55.759 --> 00:31:02.000
How much does it cost uh to deploy your system um versus classical encryption?
00:31:02.160 --> 00:31:15.039
And if you can please give us some estimation how much it costs now and how much it will you will expect this to cost in the next couple of years, because as you mentioned, the more the production, the cost going down, the unit, right?
00:31:15.440 --> 00:31:17.039
Sure, it is important, uh yeah.
00:31:17.200 --> 00:31:29.440
So the uh one thing we have uh um it is not very easy to compare the costs of of post quantum versus QKD, but I can talk about mostly about uh how the deployment of QKD will work.
00:31:29.599 --> 00:31:47.759
So if we expect to sell QKD boxes one at a time and we expect that someone in the telecom infrastructure will open the door to us and will let us put a standalone box inside their infrastructure to use it, that's it's never going to happen.
00:31:48.000 --> 00:31:56.960
So the deployment of QKD system is not won't scale as now QKD test pads are deployed.
00:31:57.039 --> 00:32:04.799
At the moment, QKD test pads are deployed with standalone systems, one in each node, and then someone that connects the fiber.
00:32:05.200 --> 00:32:06.640
This is clearly not scalable.
00:32:06.799 --> 00:32:10.000
Everyone is aware that this is not a scalable solution.
00:32:10.319 --> 00:32:12.799
The way we are approaching this problem is different.
00:32:12.960 --> 00:32:19.920
We are developing solutions that effectively seamlessly integrate into existing telecom infrastructures.
00:32:20.000 --> 00:32:36.799
So rather than going into a system and plugging a standalone solution, we effectively our aim is to insert single cards, cards of the size of your laptop, if not smallers, inside existing telecom infrastructure.
00:32:37.359 --> 00:32:47.599
This means and also this means that the QKT system will share the power supplies, the fibers, everything that is already there.
00:32:47.759 --> 00:32:52.559
And therefore, it will be a plug-and-play solution onto something that already exists.
00:32:52.799 --> 00:33:00.160
Therefore, there is a cost of a card, which is the same as any other card used in the telecom networks.
00:33:00.240 --> 00:33:05.759
We're talking about 10 to 20 to 30 uh thousand dollars, depending on the specific card.
00:33:05.920 --> 00:33:11.200
We won't have uh um, and these will be effectively seamlessly uh utilized across the network.
00:33:11.359 --> 00:33:20.240
Therefore, where the cost is per unit is not to be counted as the the cost of a box, but the cost of the use of QKD.
00:33:20.319 --> 00:33:27.039
So depending how much it will be used, uh the cost uh probably the cost will vary significantly.
00:33:27.119 --> 00:33:44.400
But the main point is that for those applications where QKD is critical, banks, uh health sector, and so on, they could rely on hardware that they don't need to buy, is already there, is already part of the telecom infrastructure, I would say.
00:33:44.559 --> 00:33:50.960
That's really the thing that uh needs uh uh to develop in the narrative of QKD uh suppliers.
00:33:51.119 --> 00:34:04.640
We are not selling boxes that go on top of the rack in the in the telecom infrastructure, but in the longer term, QKD solutions will be uh invisible in the sense, will be already there as part of the solution.
00:34:04.880 --> 00:34:17.119
This is the same that happens for a lot of technologies that in the early days they are independent boxes, and then as they evolve, they get more and more integrated onto your solution.
00:34:17.199 --> 00:34:27.519
For example, you know, I remember when I was a teenager, 15-20 years ago, that I used to have the camera on my desktop computer, which was an external solution.
00:34:27.760 --> 00:34:32.800
And now it's like this happens to every piece of technology, uh including QKD.
00:34:32.880 --> 00:34:36.719
QKD will be a feature of uh the future telecom networks.
00:34:36.960 --> 00:34:37.920
That's so cool, actually.
00:34:38.079 --> 00:34:45.440
And that's if you if done properly, can make you know create a monopoly market, even if you'll be the first in doing it properly.
00:34:45.840 --> 00:34:47.039
Sounds really interesting.
00:34:47.199 --> 00:34:50.800
Sounds like a big, big opportunity that you're sitting right now on.
00:34:50.960 --> 00:34:56.000
Um and are there any like um hidden costs of deploying QQD?
00:34:56.639 --> 00:34:57.360
Yeah, exactly.
00:34:57.440 --> 00:35:02.159
So uh this uh what I just described really tries to solve those hidden costs.
00:35:02.320 --> 00:35:18.400
What people tend to forget is that if you go into a data center and it's trying to store one box at a time, you have the cost of the engineers, you have the cost of uh the engineers that your own engineer and the engineer that hosts you, you have an incredible amount of uh um operational costs.
00:35:18.559 --> 00:35:36.000
And really um the idea of developing cars that naturally integrates into the existing networks really aims at removing, fully removing those operational costs because they are um they are combined together with the cost of delivering um the standard telecom equipment effects.
00:35:36.239 --> 00:35:36.480
Amazing.
00:35:36.719 --> 00:35:50.000
So the so the hidden costs really of existing QKD are those of deployment, and in fact, probably they are bigger than the actually uh the cost of the materials, and this is really what needs to change.
00:35:50.159 --> 00:35:56.000
QKD needs to become a solution where the operational costs are effectively uh removed.
00:35:56.559 --> 00:35:58.719
Fantastic and reduced to zero.
00:35:58.880 --> 00:36:00.400
That sounds super exciting.
00:36:01.360 --> 00:36:13.119
In the big volumes, the idea is that they will be reduced uh to um very close to zero and will be uh made in such a way that can be automated and controlled and yeah, at the network level.
00:36:13.199 --> 00:36:14.320
Yeah, super cool.
00:36:14.480 --> 00:36:20.960
Uh Francesca, so can you tell us right now where are you with CADS in terms of the deployments?
00:36:21.199 --> 00:36:24.079
Um pilots, testing that you do.
00:36:24.239 --> 00:36:26.800
How many of such deployments do you have right now?
00:36:27.280 --> 00:36:29.360
Uh at the moment we have several deployments.
00:36:29.519 --> 00:36:39.920
Uh uh, I cannot share the numbers, but we have uh uh the the nice thing is that now QKD uh is really globally um uh explored as a technology.
00:36:40.159 --> 00:36:48.239
We have uh systems uh uh in Europe, in the UK, in um in France, um in the south of Europe.
00:36:48.639 --> 00:37:01.599
We have a solution in North America as well, and we've had uh and then we've had interest really globally for our systems that uh for for other technical reasons we're not able to always fulfill.
00:37:01.920 --> 00:37:28.480
But we have uh really um yeah, the interest for QKD now is global, and now really the steps is uh continue with these uh deployments for people to understand as an education tool, for people in the industries to understand how QKD QKD can be useful to their own solutions, and this will be a process that will grow and grow and uh until hopefully we will get the genuine commercial um deployments.
00:37:28.960 --> 00:37:33.760
Well, you mentioned several markets, uh that sounds super exciting.
00:37:33.920 --> 00:37:38.559
Can you tell us a little bit more, if it's not a secret, like who are those buyers or testers?
00:37:38.880 --> 00:37:44.159
Are they mostly telecom or are the government uh agencies as well, enterprises?
00:37:44.559 --> 00:37:45.840
Who is like the profile?
00:37:46.159 --> 00:38:07.679
At the moment, at the moment, uh uh it is mostly the telecom that uh use uh want to test our solutions as well as uh telecom equipment vendors that want to understand how uh such as, for example, um we recently made public the fact that our system is traveling with the Nokia's uh kits across the globe.
00:38:08.000 --> 00:38:21.119
Um and therefore we now have a system with the um fiber providers um and with the telecom equipment vendors uh of the likes of uh like Nokia.
00:38:21.599 --> 00:38:46.880
Um these are the two levels, and the end users uh at the moment uh it is those people in the supply chain that uh interact with the users, with the end users that could be banks and general finance sector finance and and the fans seem to be the sectors that seem to have the most the biggest urgency in understanding QQT as a solution.
00:38:47.119 --> 00:38:47.920
Following as well.
00:38:48.079 --> 00:38:53.920
Okay, and then may I ask you like, do you know what is driving the decision to start testing?
00:38:54.079 --> 00:39:08.480
It's like they want to get ready for the future in terms of telecoms, I can imagine, but but is this because they feel the real risk, or there is some regulatory pressure on some markets, for example, that even push them to do that?
00:39:08.800 --> 00:39:22.400
No, interestingly, uh good or bad, we don't know, but uh these tests are going uh mostly without uh uh a direct regulatory push on QKD.
00:39:22.559 --> 00:39:41.280
The push is always on is mostly on PQC at the moment, but people that work in the telecom sectors they understood the value of QKD more than maybe the governments at the moment or certain governments, and therefore that's really where the traction is coming from.
00:39:41.440 --> 00:40:00.159
They understand that uh in three to five years' time uh QKD might be a solution, and uh they understand that the risk of being late is a bit too big not to start understanding QKD to a deeper level, exactly, and the competitiveness, right, uh amongst them as well.
00:40:00.400 --> 00:40:00.639
Exactly.
00:40:01.360 --> 00:40:06.000
Okay, and then so can you tell us a little bit more about the sales cycle?
00:40:06.159 --> 00:40:08.159
Like how are you selling it?
00:40:08.400 --> 00:40:16.559
Um, how long is the cycle usually, and what maybe what kills the deal the most often?
00:40:17.199 --> 00:40:29.039
Well, the cycle at the moment, the way because uh generally speaking, there have been uh small batches of sold systems, and therefore so it's very uh very much a request and production.
00:40:29.199 --> 00:40:39.199
I mean, we we have a batch, we have a stock, a small stock was a request, and then when the stock was available, uh sale is a very because of the volumes.
00:40:39.440 --> 00:41:01.519
Uh, walk years, the deals um at the moment, these very days really is the cost sometimes, the cost of the system still, and uh is really part of our next uh phase of bringing down the cost of the really the bill of material um down to make it more more valuable.
00:41:01.679 --> 00:41:14.880
The cost it is important really to mention here that we've pushed a lot on the chips, and really the cost now is in everything else but but the chips is in the electronics, is on the electronics in those kind of uh aspects.
00:41:14.960 --> 00:41:24.079
But the chip, in fact, we got to the nice place where the chip, the photonic integrated circuits are not the most expensive parts in our system.
00:41:24.239 --> 00:41:30.400
There are digital electronics pieces that are more expensive than our chips, which is a very nice place to be.
00:41:30.639 --> 00:41:31.599
Yeah, yeah, definitely.
00:41:31.679 --> 00:41:32.400
That's so cool.
00:41:32.559 --> 00:41:41.360
And what do you think the price needs to be to be like you know, no-brainer for those telecoms to just say like, okay, 100% you on that?
00:41:41.840 --> 00:41:51.599
Uh the 100%, I won't be I cannot say the 100%, but certainly the price points need to be that of uh um existing um telecom equipment.
00:41:51.840 --> 00:42:05.840
So when uh basically how the traffic is moved uh in the fibers across the telecom networks uh is through uh this uh technology that lives inside the telecom telecom infrastructure.
00:42:06.000 --> 00:42:18.239
And depending on the specific uh application that uh um uh a piece of hardware needs to needs to do, the costs go from a few hundred pounds to ten thousand pounds to fifty thousand pounds.
00:42:18.400 --> 00:42:27.760
Depend where we believe QKD will be a viable solution is uh in the between around twenty thousand dollars, fifteen thousand dollars.
00:42:28.000 --> 00:42:32.320
That's the price point where we believe uh there could be an inflection point.
00:42:32.719 --> 00:42:33.360
Amazing.
00:42:33.599 --> 00:42:43.920
Okay, and you mentioned as well that the new uh potential use cases or usage of the chips are also in the defense industry, and that is very interesting for me.
00:42:44.000 --> 00:42:51.199
I would love you to um tell us a little bit more how the defense industrial defense agency can use your chips.
00:42:52.320 --> 00:43:18.159
Well, uh obviously the the nice uh um thing of uh of our solution is that uh um can be applied to, in particular, because of the integrated approach, uh can be applied, for example, nice applications are satellite-based solutions where uh the signals can travel um uh in the in the free space from the ground to the satellite back on the ground and therefore cover long distances.
00:43:18.400 --> 00:43:30.639
That's one way that uh uh and this solution is intrinsically harder to hack and therefore is um is a is a good way of establishing very secure connections across the globe.
00:43:30.880 --> 00:43:37.199
And uh in the past we also did a demonstration with uh um uh drone-based uh QKD.
00:43:37.840 --> 00:43:51.199
Effectively, uh, we had a transmitter QKD chip on a drone, and this was used uh and this uh was exchanging keys between the flying drone and uh and a ground station.
00:43:51.440 --> 00:43:59.760
These are two examples of application, and these are two examples of applications where uh effectively which mostly involve free space, so not the use of.
00:44:00.159 --> 00:47:26.719
fibers because the fiber will be constraining the the direction of the sign the the positioning in uh in uh in uh in uh let's say um what and and and the real deployment uh and the role as well this is very question what you said because i think we we don't think about it i think because of the war with ukraine everyone gets to use to that drones are like you know integrated part of the war in um self-defense and also um the uh on the attack side and they obviously have no human involved but nobody thought about it that actually this connection can be overtaken as any other connection with with technology so you can protect that communication with the machine yeah and also and also the the direct uh the direct optical communication in itself has the advantage that it's much harder to hack anyways compared to radar signals and therefore much harder to jam which is also another another um yeah parallel feature of of optical communications in general exactly because if if hacked it could be very destructive I mean very yeah instead of attacking your enemies can attack you uh back but wow that's very interesting okay so um knowing all of these potential use cases especially for the defense um defense sector um is this it is is your roadmap kind of driven by the government and defense demand or is something that um you're just doing on site anyway at the moment uh at the moment uh um we yeah there is interest in the government and there is uh interest in the defense we we believe yeah the the telecom in general infrastructure as such is probably that that will see the the the the the more direct uh the the the beach market yeah and uh we're talking about uh probably the banks is still probably where we will see the um the biggest big the the earliest uh utilization of of uh of qkd systems rather than defense is certainly valuable but obviously um there is there are other levels of complexity that uh might stop the the understanding so the user telecom would be the the beach hat market for you and uh defense or or banking as well could be the secondary market in the future but it may change in the future right we don't know yeah yeah yeah we don't know uh there's a lot of tensions going on right now globally and the political geopolitical situation is pretty tense so you never know how this may accelerate and and maybe potentially change the demand right and as a startup we usually try to adjust okay uh Francesco we're going slowly to the end of our discussion this this is the time I think for for for asking for for me asking you on the predictions so what do you think is gonna happen in the next five years if you can imagine um qkd and also um post quantum cryptography in the next five years where do you think we will be yeah so the way uh I was asked this question recently and the way I replied is that I was asked where I would see QKD where I want to see QKD in the next five to ten years.
00:47:26.880 --> 00:47:33.920
My answer was that I would like not to see it because I would like it to be really seamlessly integrated into the existing telecom networks.
00:47:34.239 --> 00:48:51.519
So we don't have it yeah and really this is uh where I believe uh QKD will be in the next uh within the next five years qkd will be a feature of existing uh of standard infrastructure will be uh widely deployed and will be up to the um and based on the uh on the needs it will be offered as a sys as a solution to to the um to the to this to the sector that uh require them so really where I want to see where I dream to see QKD is really as a standard uh well highly integrated component onto the existing uh infrastructure as it is anything in in the the in the networks and in comparison uh with uh PQC PQC will will uh cover those uh the the a wider possibly a range of of solutions particularly in the mobile um in the mobile landscapes but our QKD will be used uh in all those critical um places that where security is absolutely paramount where there is no there is no optionality between um choice A and B but where you need all the layers of security for for your application.
00:48:52.480 --> 00:49:04.880
That perspective of the QKD being so common so obvious and um standard um it looks very attractive and I hope I wish that will happen.
00:49:05.280 --> 00:49:24.639
When you were speaking about the the security of the data one of the things that came to my mind straight away was the medical data because right now there's a lot of you know institutions that store our medical data and it's happened all the time the leaks.
00:49:25.280 --> 00:51:04.480
Yeah right and that should not happen definitely so do you think QKD can also have a um absolutely 100% and in fact is uh is uh yes 100% and this is very true the reason why qkd in in this particular instance is uh is incredibly uh useful is also because as opposed as to many other applications usually the health sector actually owns fibers they own networks and therefore qkd is even more easily integrated in this kind of infrastructure because basically the solution is already there and what you need to add is just the terminals and therefore really is um yeah the the certainly the the health data of of patients across across um countries is certainly um one one very useful one very one uh very good uh use case yeah so in summary if the telecoms are properly integrated um qqd in the future we all will be safe yeah that sounds good and very optimistic francesco thank you so much for the discussion today um it was a really pleasure um and thank you so much for being with us thank you thank you very much that was blonde and quantum thank you for joining me on this journey through the quantum business frontier if you like the episode please review another Spotify or Apple Podcast and help more people discover the quantum world without needing to untangle the theoretical physics see you next time unless the cat change the time
00:00:00.880 --> 00:00:09.679
Hey, my name is Eva, and this is Abelong Quantum, the podcast that breaks down quantum technology into real-world business impact.
00:00:10.320 --> 00:00:12.320
Here we make the complex simple.
00:00:12.560 --> 00:00:14.320
No equation, no overthinking.
00:00:14.720 --> 00:00:18.239
Just insight, innovation, and a bit of cute.
00:00:18.879 --> 00:00:22.079
So don't worry, you don't need a PhD in physics to follow.
00:00:22.640 --> 00:00:29.839
In each episode, I talk to funders, scientists, and investors about how quantum is reshaping the industry today.
00:00:30.000 --> 00:00:31.839
Not in some distance future.
00:00:32.479 --> 00:00:42.320
Oh, and if you hear a black cat pouring in the background, that's my co-host, very alive shorting as cat joining the conversation.
00:00:42.560 --> 00:00:43.759
Her name is Moon.
00:00:44.479 --> 00:00:46.079
Let's get started.
00:00:47.679 --> 00:00:52.560
Welcome everyone to the next episode of Blond and Quantum Startup Series.
00:00:52.719 --> 00:00:58.240
Today I have a pleasure to host Francesco Raffaelli, a CTO of Kats Quantum.
00:00:58.799 --> 00:01:01.039
Hello Francesco, how are you?
00:01:01.439 --> 00:01:03.280
Hi, nice to meet you and thank you very much.
00:01:03.359 --> 00:01:04.239
I'm very good.
00:01:04.640 --> 00:01:05.040
Great.
00:01:05.439 --> 00:01:07.599
It's a really pleasure to have you here today.
00:01:07.760 --> 00:01:15.040
And we're gonna talk a lot about cats, what you guys are building, and also a little bit more about the industry.
00:01:15.200 --> 00:01:25.519
And on the end, I usually ask, I have this tradition to ask my um my guest about some prediction, like how you see the industry in the next, I don't know, 5, 10, 15 years.
00:01:25.760 --> 00:01:27.359
So please be ready for that.
00:01:27.439 --> 00:01:29.280
Um, but let's start from the kats.
00:01:29.519 --> 00:01:36.640
So you guys building a chip-based quantum K distribution, which we in the industry called QKD.
00:01:37.120 --> 00:01:41.920
Um, and also, as I read, quantum random number generation.
00:01:42.079 --> 00:01:47.680
Um, and this is all embedded in the hardware integrated photonics.
00:01:48.000 --> 00:01:50.959
So, again, silicon chips, right?
00:01:51.040 --> 00:01:59.760
So, this is about the theory, what I read, what I learned, but as you know, our audience are usually enterprise executive and business people.
00:02:00.000 --> 00:02:06.319
So, from that two-liner definition, they probably would not understand anything.
00:02:06.560 --> 00:02:14.639
So, if you can please kick it off from explaining in a very plain English what you guys are doing.
00:02:15.120 --> 00:02:15.680
Sure.
00:02:16.080 --> 00:02:22.080
Basically, yes, at Cats, we are building, as you said, the chip-based uh quantum key distribution systems.
00:02:22.319 --> 00:02:42.000
What we mean by that means uh we have a quantum key distribution system, so we have these systems that are used to generate and send and distribute um uh optical signals that then are used um to distribute uh keys uh to be used for cryptography applications.
00:02:42.400 --> 00:03:00.800
Uh the main differentiator of CATS is that from day one, so since our time at the University of Bristol, we knew that uh the only way to make uh quantum key distribution scalable was to build it upon technologies that are scalable.
00:03:01.039 --> 00:03:13.439
And this is why uh um we focused uh it is we spent the last uh 10 years effectively developing systems where the core technology is a photonic integrated circuit.
00:03:13.520 --> 00:03:19.840
So most technologies, most QKD systems will have uh would have uh fiber optics, effectively.
00:03:19.919 --> 00:03:36.639
Uh these are um components that are uh assembled together, they are connected by optical fibers that are good for prototypes, but they are not scalable because it would take a day to set up one, and it would take a thousand days to set up a thousand days to set up one thousand.
00:03:36.960 --> 00:03:37.439
Exactly.
00:03:37.599 --> 00:03:39.120
And they're also big, right?
00:03:39.680 --> 00:03:52.719
Also big and they are also um quite expensive um uh because every single component cost a few thousand pounds, and these sum up to become a very, very nice bill of materials.
00:03:52.960 --> 00:04:04.080
Instead, what we have done at CATS, we have uh uh spent a lot of time and effort in uh translating this uh um generation and detection of photons uh into photonic integrated circuits.
00:04:04.240 --> 00:04:15.840
Basically, these are uh these photonic integrated circuits use the same technology that is used for um computers, is the semiconductor technology used for computers, telephones, and everything.
00:04:16.079 --> 00:04:38.079
And by doing so, we have created a technology that uh um effectively enable to um have miniaturized solutions, but most importantly, is a technology that once you have demonstrated the first batch of chips, you effectively have already uh a direct route towards volume production.
00:04:38.240 --> 00:04:47.519
So once we have a first batch of say 30-50 units, then it's just a matter of repeating the order and then you suddenly scale to the thousands of units.
00:04:47.680 --> 00:04:57.040
And the nice feature of that is that the cost is as opposed to with fiber solution, the cost is not linear and the cost decreases as the volume increases.
00:04:57.279 --> 00:05:00.480
And this is uh really the nice aspect of our exploration.
00:05:00.959 --> 00:05:08.319
So you're basically saying that in the next couple of years you may have a manufacturing similar to semiconductors.
00:05:08.399 --> 00:05:23.759
So, again, for summary, we have a small chip which is much more scalable because of the production and will be cheaper in the future comparing to those bigger rocks that right now need to be installed um somewhere in the hardware and they are pretty expensive, right?
00:05:23.920 --> 00:05:25.759
Yeah, so that's what I understood.
00:05:25.920 --> 00:05:28.959
Now tell us, Francesco, why do we need it?
00:05:29.279 --> 00:05:30.879
Why this is needed for?
00:05:31.360 --> 00:05:32.240
Yeah, sure.
00:05:32.399 --> 00:05:43.839
Uh the the main the really the the the the reason of existence of QKD is the fact that um uh computing power is evolving at exponential speed, as we know.
00:05:44.000 --> 00:05:49.600
Everyone knows about uh AI, but there is also a lot of people also know about quantum computing.
00:05:49.839 --> 00:06:00.800
Quantum computing is a technology that, amongst its big uh promises, has also uh the potential of uh breaking the existing um cryptography.
00:06:01.040 --> 00:06:15.120
Existing current uh cryptography is based on effectively mathematical problems that are very hard to solve by uh standard computers, but for which quantum computers unfortunately seem to be tailor-made.
00:06:15.199 --> 00:06:27.360
So quantum computers seem to be incredibly efficient, will be incredibly efficient in five to ten years' time in breaking these uh existing um uh cryptography.
00:06:27.600 --> 00:06:36.079
Quantum key distribution is one of those technologies that uh uh is understood to be secure against quantum computing attacks.
00:06:36.319 --> 00:06:38.319
And the reasons are several.
00:06:38.560 --> 00:06:50.560
Two main reasons are the first one is that QKD has this nice uh feature of being um able to detect any attempt of ease dropping.
00:06:50.639 --> 00:07:01.759
So if someone tries during the exchange of a key tries to gather information about the keys, a QKD system can detect this type of hacking.
00:07:02.079 --> 00:07:12.800
And the other aspect of QKD is that it fundamentally lives in a different dimension to the one that quantum computers can attack.
00:07:13.040 --> 00:07:18.480
Quantum computers attack software and QKD lives in optical fibers.
00:07:18.639 --> 00:07:31.360
So essentially, there are two substantial reasons for why QKD is important, and these are all rooted to the fact that quantum computers will be soon able to break existing um uh cryptography.
00:07:31.600 --> 00:07:52.160
There is also another technology which is called uh post-quantum cryptography, which is effectively, which is effectively um different new kind of algorithms that have been uh developed and they will um replace the existing ones for which there are not known quantum attacks.
00:07:52.399 --> 00:08:00.160
And therefore they are effectively a complementary, there will be a complementary um solution to QKD.
00:08:01.600 --> 00:08:06.959
These will have different um type of application, different scale of of deployment.
00:08:07.519 --> 00:08:30.079
Post-quantum cryptography will cover the majority of the cases, but it won't be sufficient for uh those situations where security is paramount, such as banking, critical infrastructure, such as um power plants and similar applications is where uh QKD uh will find its role to support the security.
00:08:30.480 --> 00:08:31.839
So the reason, yeah.
00:08:32.080 --> 00:08:45.759
Yeah, so basically, if I just got you saying is that for most of standard enterprise post-quantum cryptography, it can be enough because it's working on the software layer, which um normally quantum computers attack.
00:08:46.000 --> 00:09:06.240
But for crucial infrastructure, so we're talking here, I don't know, finances, banking, uh, a lot of people asking about crypto these days, maybe uh pharmaceutical, maybe defense, um, those companies and system will need on top of that um also the hardware protection, which is um QKD.
00:09:06.639 --> 00:09:11.120
Yes, the reason being that would they have both or only the hardware one?
00:09:11.440 --> 00:09:12.639
They would probably have both.
00:09:12.799 --> 00:09:28.960
The reason uh for why QKD is necessary because in a similar way as uh for which the current RSA protocols are being replaced, no one can guarantee that this post-quantum algorithm will be secure for how long?
00:09:29.200 --> 00:10:00.559
We know that the exponential development uh of computing power, which is not necessary, just the direct uh attack on the algorithm, as quantum computers will do, but also the detection of faults in how an algorithm has been written, are developing so fast that relying on a single uh software approach, which is not proven to be secure but deemed to be secure, is a too risky approach for the for certain critical industries.
00:10:00.879 --> 00:10:01.039
Okay.
00:10:01.759 --> 00:10:03.919
And last question on that topic.
00:10:04.000 --> 00:10:14.240
So, do you think um the QQD and PQC will coexist, or maybe one will dominate in 10 years?
00:10:14.399 --> 00:10:16.080
For example, they will coexist.
00:10:16.240 --> 00:10:45.919
They will coexist because there will be situations because the um what people maybe don't it's difficult to quite to picture, but the complexity of the networks is huge, and therefore there are several situations where uh one will be sufficient, there are several situations where the other would be more practical to deploy and faster to deploy, and therefore we expect that the future security landscape would be very heterogeneous with different solutions for different uh uh specific uh cases, so they will coexist.
00:10:46.240 --> 00:10:56.799
And and you mentioned um telephones as one of the examples, and I know that you have some very fruitful cooperation with some of the big telecoms.
00:10:56.879 --> 00:11:00.000
Maybe you can I will leave it to you to talk about it.
00:11:00.080 --> 00:11:09.919
Um, but I wanted to ask you in general, why should telecoms operator invest in QKD instead of the software-based encryption?
00:11:11.279 --> 00:11:12.399
This is a good question.
00:11:12.559 --> 00:11:23.840
So, the reason why um so telecom operators effectively offer are those that offer the service to the banks, to the health sectors, to the uh energy grid.
00:11:24.399 --> 00:11:36.240
And uh the reason is because as providing the service, they want to provide the most uh secure service, and this is a combination of uh QKD and MPQC.
00:11:36.399 --> 00:11:48.240
So uh again, they provide the service to the to the banks, say, and they will make and based on the specific case, they will uh implement one or the other or both solutions.
00:11:48.559 --> 00:11:56.240
I think that's and if you think about your place in the Spark cooperating with telecoms, for example.
00:11:56.320 --> 00:11:57.840
So, how do you place yourself?
00:11:58.000 --> 00:12:04.799
You're more like a telecom vendor or more like a semiconductor company, or maybe something else, maybe security company?
00:12:05.279 --> 00:12:06.639
Yeah, it's a good question.
00:12:06.799 --> 00:12:10.080
I think uh we are uh a bit of the these three.
00:12:10.320 --> 00:12:18.480
We did what we sell, we sell um systems, solutions that fit within the existing telecom equipment.
00:12:18.720 --> 00:12:28.960
So, what we sell, we sell our QGD system that might fit uh into uh the equipment that already exists in the telecom uh network and therefore that can be integrated.
00:12:29.039 --> 00:12:35.759
So, in that sense, we provide components for uh the telecom equipment um industry.
00:12:36.240 --> 00:12:40.639
We also uh sell uh a lower level down into the supply chain.
00:12:40.720 --> 00:12:56.080
We also sell uh simple components that provide a different level of integration where people might explore different applications that is not necessary, the standard ones, where they can take our own chips, maybe packaged, and develop their own uh solution.
00:12:57.120 --> 00:13:04.639
And these are it's like a development kit, yeah, kind of like a develop development kit with the core component, exactly.
00:13:04.720 --> 00:13:16.080
We can imagine our us selling the chip, but also a development kit, and therefore that can use can be used for different applications that is beyond what we have imagined, for example.
00:13:16.240 --> 00:13:16.720
Yes.
00:13:16.960 --> 00:13:20.240
And these are two different layers in the supply chain.
00:13:20.399 --> 00:13:34.799
These are all technologies that uh entail uh security, and therefore uh the kind of approaches that we use in developing our hardware and software stack are those of uh we try to follow those of HSMs and security hardware companies, yes.
00:13:35.039 --> 00:13:40.799
So there's basically two different products, depends on the need on the company that is a client.
00:13:40.960 --> 00:13:57.200
Uh, what's about the uh random numbers generators that yeah, and we have uh um as part of our journey we have developed uh uh random number generators, which was in fact my uh started off with my PhD thesis, which is uh from Bristol, again from Bristol.
00:13:57.440 --> 00:13:59.759
From Bristol, from Bristol under uh Jonathan Matches.
00:14:00.399 --> 00:14:06.799
And yes, the quantum random number generator is a very interesting device that again we developed on on chip.
00:14:07.039 --> 00:14:14.720
So we have a very nice little component that produces very high quality random numbers, and this can be used in a variety of solutions.
00:14:14.960 --> 00:14:39.120
Originally, a catch was developed to be to provide the the randomness, the entropy for our quantum key distribution systems, but also the quantum random number generators can find applications in several uh other solutions such as classical cryptography, but also Monte Carlo stochastic simulations, anything, effectively any solution that requires very high quality and provable randomness.
00:14:39.200 --> 00:14:42.559
Uh, that's what the quantum random number generator um can offer.
00:14:42.720 --> 00:14:50.320
And you mentioned your PhD and your ten spend in the University of Bristol, and we talked about this a little bit before um recording.
00:14:50.399 --> 00:15:06.159
I wanted to ask you like what kind of quantum dust ferry is in Bristol on that university, because you are um at least fifth person that I'm meeting during um hosting this show who is coming out of the Bristol University.
00:15:06.399 --> 00:15:11.519
Recently I spoke with Reza Najpati, who's the head of quantum in Cisco.
00:15:11.759 --> 00:15:14.480
Um, but I know there's not only him, there's the small people.
00:15:14.639 --> 00:15:19.120
If I'm not wrong as well, Psych Quantum had their roots in Bristol as well, right?
00:15:19.360 --> 00:15:20.159
Yeah, yeah.
00:15:20.320 --> 00:15:22.080
So what's going on there?
00:15:22.639 --> 00:15:26.559
Yeah, um, effectively, Bristol is an amazing playground for quantum.
00:15:26.720 --> 00:15:35.519
Um, I joined arrived in uh 2014 in Bristol when the quantum engineering and technology labs were literally exploding, blooming.
00:15:35.600 --> 00:15:51.519
Yeah, uh it was uh it started off with uh um John Rarity, and then again the Jeremy Bryan and Mark Thompson, the founders of PsyQuent, were those that grew the team very significantly, um, up to 150 people at the time I was there.
00:15:51.759 --> 00:16:12.240
And these uh has uh produced several startups, and in fact, uh this uh the Bristol Hub, thanks to the Quantum U Technology and uh Engineering uh Centre, produced also um probably one-third of the quantum startups in the UK actually somehow passed through Bristol in their in their journey.
00:16:12.480 --> 00:16:32.879
Yes, Bristol uh and the amount of the amount of expertise, particularly at the intersection between quantum and integrated photonics, is really what has made uh Bristol a special place because it was the intuition of the founders of the group that quantum was cool, but to make it a thing, it needed to go through photonic integrated circuits.
00:16:33.120 --> 00:16:51.200
And literally they invented the field of uh that's what happened, effectively invented the the field of uh photonic integrated circuits for quantum, and uh then now there are maybe yeah, seven or eight startups that still live in Bristol, plus several others, such as like quantum, obviously uh they've grown massively.
00:16:51.600 --> 00:16:57.440
No, it sounds like a very magical place because I can understand the hype for quantum, right?
00:16:57.679 --> 00:17:22.720
And I can imagine that many academic organizations or institutions may have some specific part of hype on different stuff, but you know, a building a culture where this research isn't applied into real life and commercialized, including so many startups, including say quantum, um, that's a different story, yeah, and that requires much more work.
00:17:23.039 --> 00:17:32.160
Um, so that's a magical, and I think you are very lucky that you were that time um inside of it, and you could observe that from close distance.
00:17:32.400 --> 00:17:40.160
So, let me ask you what needs to happen for Cats to become another one billion dollar company then?
00:17:40.640 --> 00:17:41.599
That's a very good question.
00:17:41.680 --> 00:17:51.759
I think uh we can say that the difference between, say, quantum computing and quantum cryptography, such as uh QKD, is that quantum computing really uh lives in green fields, open fields.
00:17:51.920 --> 00:17:59.519
There is nothing compared that compares to quantum computing, and quantum computing is applications that are still to be discovered.
00:17:59.839 --> 00:18:01.440
It's very different for QKD.
00:18:01.599 --> 00:18:13.519
QKD lives uh in a very highly regulated market, which is cryptography, in a very highly uh controlled market because cryptography obviously is at the core of our digital safety.
00:18:13.839 --> 00:18:33.680
And therefore, really that what needs to change for quantum companies, quantum cryptography companies to grow is uh the understanding of certain um agencies, governments, and regulatory bodies uh of the role that quantum communications, quantum cryptography can have in the landscape of cryptography.
00:18:33.839 --> 00:19:01.359
Therefore, um certainly uh what we need is uh what we need to get uh to the scale for quantum QKD companies, for QKD companies is really the collaboration with the uh security agencies, with governments, and such that they promote and they promote QKD where QKD can be the right solution and they effectively explain QKD in a way that is fair, genuine, and correct.
00:19:01.440 --> 00:19:21.680
Effectively, I think this is uh uh what we are seeing at the moment is probably that uh because cryptography historically has been a computer science slash math industry, uh, cryptography is was invented by the mathematicians, brought forward by uh computer scientists.
00:19:22.000 --> 00:19:31.039
Now that physicists and telecom um engineers enter the the narrative, there is a bit of uh lack of uh understanding.
00:19:31.359 --> 00:19:33.200
There are miscommunications.
00:19:33.680 --> 00:19:53.839
It is not clear uh what uh QKD can do, what are the limits of QKD that exist, as for everything, and therefore, really the need the work that needs to happen is uh stronger and closer collaborations with the exist between the existing solutions and who promotes the existing solutions and the novel solutions such as uh QKD.
00:19:54.240 --> 00:20:13.119
This is uh the step, the big step that needs to happen for a company like that to I'm glad you said that because I have heard a lot of good things about the supporting environment in the UK in terms of the you know procurement um processes and the agencies supporting quantum startups.
00:20:13.279 --> 00:20:18.000
I think I heard that from Michael Cobart from National Uh Quantum Computing Centre.
00:20:18.480 --> 00:20:23.759
So, right now, do you have any cooperation with those agencies in the UK?
00:20:24.079 --> 00:20:27.920
So just it is very important here to make a clear distinction.
00:20:28.240 --> 00:20:33.039
Quantum computing uh yeah, is seeing uh the good support within the UK.
00:20:33.200 --> 00:20:45.119
Quantum key distribution is not being supported at the moment by the um uh security agencies, and this is really one of the of the steps that needs to happen.
00:20:45.200 --> 00:20:57.680
Uh, NCNC needs to understand to have an up-to-date view of what UKD, um which now probably it is fair to say that now probably they have a slightly out-of-date view of QKD.
00:20:57.759 --> 00:21:10.319
The effort is really to understand where QKD really sits at the moment in terms of technology development, and therefore um this kind of uh uh and therefore understand where can be a solution within the landscape of cryptographies.
00:21:10.400 --> 00:21:15.039
So there is a big distinction in the sense between quantum key distribution and quantum computing.
00:21:15.200 --> 00:21:20.000
That's uh so really this is uh so it's true that quantum computing is supported massively.
00:21:20.160 --> 00:21:24.559
It is not quite QKD is supported but not quite as strongly.
00:21:25.039 --> 00:21:41.200
Thank you so much, uh Francesco, for saying it because I was not aware, and definitely by saying it publicly, we send a strong signal uh to the agencies like, hey, it's time, it's time to step in and have a look and support in the same way that you support quantum computers.
00:21:41.359 --> 00:21:43.839
So thank you, thank you so much for saying it.
00:21:44.079 --> 00:21:56.400
Um speaking of, I noticed during my research that actually CAT has a lot of um different RD grants and utilizing pretty um efficiently.
00:21:56.559 --> 00:22:05.440
Uh and I don't know if that was your A weird decision to go that way to utilize the grants rather than private capital.
00:22:05.519 --> 00:22:06.960
I know you have both, right?
00:22:07.119 --> 00:22:22.799
But what I wonder is what's the common or maybe the biggest misconception that you heard from the investors who are potentially interested in investing in company like Cats about quantum cryptography.
00:22:22.960 --> 00:22:24.079
What do you usually care?
00:22:24.400 --> 00:22:41.680
I think that the main the biggest misconception is really while QKD is a very complex technology, basically what we are doing, we are sending one photo at a time across um commercial networks, we are detecting them and we are establishing secure keys by this process.
00:22:41.759 --> 00:22:44.319
So this is a very complex uh process.
00:22:44.799 --> 00:22:52.720
But what people maybe have not quite understood is the progress that uh has been achieved by CATS and others in the development of these systems.
00:22:52.880 --> 00:23:15.359
If you think that we have uh a single chip on the transmitting side and a single chip on the receiving side that do all of this in the optical domain and they do it to the right following all a set of mathematical proofs that are very sophisticated, and we do it with standard semiconductor foundries using standard electronics and using standard computing solutions.
00:23:15.599 --> 00:23:27.680
So, what people have probably uh not quite understood is that QKD is not anymore in an RD phase, QKD is very much in an engineering development phase.
00:23:27.839 --> 00:23:41.359
We have demonstrated the fundamentals, uh, we are ready with our solutions to scale, and what we need is really the right environmental conditions for the technology to be commercialized effectively.
00:23:41.519 --> 00:24:02.880
Therefore, hence the importance of working with the governments, with the security agencies, because the pull will come from banks, health sector once there is a good agreement, when there is the agreement, one there is the with the with these uh companies that are those with these institutions that are those that set effectively the regulations.
00:24:03.279 --> 00:24:06.559
So, really, but the bigger misconception I believe is in the technology.
00:24:06.640 --> 00:24:12.960
The technology is there, is ready to scale, and what is missing now is the commercial pool.
00:24:13.279 --> 00:24:14.799
That's that's great that you said that.
00:24:14.880 --> 00:24:27.759
Um, because again, we do not have that many investors skills on the uh quantum technologies market, and as you say, many of them um are not really aware that the technologies are here, so this is cool.
00:24:28.160 --> 00:24:37.759
Let's talk a little bit of what happened in April this year because I think CAT just announced the company to the security hardened and scalable uh QKD prototype.
00:24:37.920 --> 00:24:40.640
Um, can you tell us a little bit more about this?
00:24:40.880 --> 00:24:43.920
This was designed for critical infrastructure.
00:24:44.160 --> 00:24:51.119
And is that mean that you are already there to have the fully integrated telecom grid chip?
00:24:51.359 --> 00:24:52.640
Please tell us more about it.
00:24:53.200 --> 00:25:25.200
Yeah, so certainly what we have at the moment is we developed this system which have uh these chips, and the nice thing is that now all just not just the chip, but all the systems that that goes around it the electronics, the compute have gone through uh a further level of development to uh for where effectively also aspects such as the software, such as the hardware have have been pro have hardened, as we said, to be ready to be um used uh across across networks.
00:25:25.359 --> 00:25:31.839
So this is uh another another point that people sometimes is sometimes people maybe don't quite understand.
00:25:32.079 --> 00:25:39.599
That again, QKD is an advanced technology, it's very difficult, but it's also advanced and is really to scale.
00:25:39.759 --> 00:25:48.319
The nice thing about our solution now is that really um again, we are at that one step away from deploying at scale.
00:25:48.559 --> 00:26:08.960
So we what if once the opportunity comes, our chips are designed uh in a particular uh silicon photonic foundry that already can could produce this system in the millions, and uh all the electronics as well, obviously, can be produced at a very high scale.
00:26:09.039 --> 00:26:23.200
So that's really the important thing of our solution is that in with this version that we recently completed, we have removed all the little imperfections, and now really we are one step away to make this the solution scalable really.
00:26:23.680 --> 00:26:26.079
Well, first of all, Francesco, congratulations.
00:26:26.319 --> 00:26:36.559
I don't think there's many um quantum technology startup that can say, hey, we are one step almost from massive production that our solution can be you know um manufacturing millions.
00:26:36.720 --> 00:26:44.319
That's huge, and that itself should be a really good trigger for the investor to cue the front of your head headquarter.
00:26:44.720 --> 00:26:54.240
Um, but I wanted to talk a little bit about the solution as well because I know you are um testing in live right now with BT, with the British Telecom.
00:26:54.480 --> 00:26:55.680
How's that going?
00:26:56.960 --> 00:27:04.640
No, uh I would say that the team at BT has been incredibly uh supportive throughout our journey.
00:27:04.799 --> 00:27:11.759
Uh, the research labs at BT has been supporting really from day one since back in the day, I think it was 2018.
00:27:12.240 --> 00:27:39.599
And uh it took uh several years of interaction, of patience from their side for us to get through all the challenges of developing this technology, and it's been uh um a very good opportunity uh for us to understand the telecom infrastructure, to understand the standards required uh for our solutions, and now uh this uh this system uh the the most uh um yeah, and now uh the things are going very well.
00:27:39.759 --> 00:28:07.119
The best example of that is a project that was called uh recently was called um QRSure, which also saw the participation of uh the team of uh the National Physical Laboratories and other uh entities across uh the UK, where basically uh as part of our objective of making a genuine um commercial solution, we put our uh QKD systems under uh third-party scrutiny.
00:28:07.279 --> 00:28:26.720
So, what we did, we actually put our uh our own hands off the system, we shipped our systems to MPL, we let they we let them play with the systems, and they validated that effectively our solution does what it says in the tin, which is really the important thing of cryptography.
00:28:26.960 --> 00:28:37.039
And uh BT indeed has been incredibly supportive uh in setting the requirements, supporting, steering the direction towards something that is actually what the telecom uh really needs.
00:28:37.200 --> 00:28:40.960
I mean, uh, incredibly grateful to the to the teams uh BT.
00:28:42.000 --> 00:28:44.960
This is fantastic, and congratulations uh for that.
00:28:45.119 --> 00:28:47.039
I think you were somewhere announced online.
00:28:47.200 --> 00:28:55.920
I I saw that you were announced the only one commercially um ready or very close to being commercially ready, uh, QQD solution in the world.
00:28:56.079 --> 00:28:57.680
So that's a big thing as well.
00:28:57.839 --> 00:29:08.400
Um, I wanted to ask you what you mentioned that makes me think as well about the production in millions um because of the standard on the standard on the semiconductor market.
00:29:08.799 --> 00:29:29.279
Do you think that there is a risk for you as a company that um in some time in the next couple of years, if solution you know be very successful, that the large semiconductor player can replicate what you're doing and then try to sell it as, I don't know, white label solution?
00:29:29.680 --> 00:29:32.480
Yeah, I mean that this is risk in industries.
00:29:32.640 --> 00:29:39.839
Uh we believe that uh uh there are two things uh um that uh make us confident that this won't happen.
00:29:40.079 --> 00:29:54.480
The first thing is that uh we have very good IP protection of some of the critical aspects of our technology, and this makes us confident uh that uh this provides sufficient uh barriers to entry.
00:29:54.720 --> 00:30:18.640
Also, the other interesting aspect is that beyond the IP, which is only part the protected IP, there is the tens, hundreds of secrets that go from designing the chip to integrate it with the right electronics to uh developing the software and everything because Cats is a full-stack solution from the chip to the software that controls the electronics that controls the chip.
00:30:19.440 --> 00:30:35.599
We believe that uh while it is possible for a huge company uh to replicate this, the huge company won't will still need to go through a very significant amount of development, and that's where um which will take will take years, anyways.
00:30:35.759 --> 00:30:42.319
Therefore, there is a first uh started advantage in this sense, yeah, and first mover advantage as well.
00:30:43.119 --> 00:30:46.400
Okay, and um let's talk and just a little bit about the cost.
00:30:46.640 --> 00:30:48.640
So, can you just tell us a little bit more?
00:30:48.720 --> 00:30:55.519
Because I think that may be a question of a lot of enterprise executives right now or listening to our discussion.
00:30:55.759 --> 00:31:02.000
How much does it cost uh to deploy your system um versus classical encryption?
00:31:02.160 --> 00:31:15.039
And if you can please give us some estimation how much it costs now and how much it will you will expect this to cost in the next couple of years, because as you mentioned, the more the production, the cost going down, the unit, right?
00:31:15.440 --> 00:31:17.039
Sure, it is important, uh yeah.
00:31:17.200 --> 00:31:29.440
So the uh one thing we have uh um it is not very easy to compare the costs of of post quantum versus QKD, but I can talk about mostly about uh how the deployment of QKD will work.
00:31:29.599 --> 00:31:47.759
So if we expect to sell QKD boxes one at a time and we expect that someone in the telecom infrastructure will open the door to us and will let us put a standalone box inside their infrastructure to use it, that's it's never going to happen.
00:31:48.000 --> 00:31:56.960
So the deployment of QKD system is not won't scale as now QKD test pads are deployed.
00:31:57.039 --> 00:32:04.799
At the moment, QKD test pads are deployed with standalone systems, one in each node, and then someone that connects the fiber.
00:32:05.200 --> 00:32:06.640
This is clearly not scalable.
00:32:06.799 --> 00:32:10.000
Everyone is aware that this is not a scalable solution.
00:32:10.319 --> 00:32:12.799
The way we are approaching this problem is different.
00:32:12.960 --> 00:32:19.920
We are developing solutions that effectively seamlessly integrate into existing telecom infrastructures.
00:32:20.000 --> 00:32:36.799
So rather than going into a system and plugging a standalone solution, we effectively our aim is to insert single cards, cards of the size of your laptop, if not smallers, inside existing telecom infrastructure.
00:32:37.359 --> 00:32:47.599
This means and also this means that the QKT system will share the power supplies, the fibers, everything that is already there.
00:32:47.759 --> 00:32:52.559
And therefore, it will be a plug-and-play solution onto something that already exists.
00:32:52.799 --> 00:33:00.160
Therefore, there is a cost of a card, which is the same as any other card used in the telecom networks.
00:33:00.240 --> 00:33:05.759
We're talking about 10 to 20 to 30 uh thousand dollars, depending on the specific card.
00:33:05.920 --> 00:33:11.200
We won't have uh um, and these will be effectively seamlessly uh utilized across the network.
00:33:11.359 --> 00:33:20.240
Therefore, where the cost is per unit is not to be counted as the the cost of a box, but the cost of the use of QKD.
00:33:20.319 --> 00:33:27.039
So depending how much it will be used, uh the cost uh probably the cost will vary significantly.
00:33:27.119 --> 00:33:44.400
But the main point is that for those applications where QKD is critical, banks, uh health sector, and so on, they could rely on hardware that they don't need to buy, is already there, is already part of the telecom infrastructure, I would say.
00:33:44.559 --> 00:33:50.960
That's really the thing that uh needs uh uh to develop in the narrative of QKD uh suppliers.
00:33:51.119 --> 00:34:04.640
We are not selling boxes that go on top of the rack in the in the telecom infrastructure, but in the longer term, QKD solutions will be uh invisible in the sense, will be already there as part of the solution.
00:34:04.880 --> 00:34:17.119
This is the same that happens for a lot of technologies that in the early days they are independent boxes, and then as they evolve, they get more and more integrated onto your solution.
00:34:17.199 --> 00:34:27.519
For example, you know, I remember when I was a teenager, 15-20 years ago, that I used to have the camera on my desktop computer, which was an external solution.
00:34:27.760 --> 00:34:32.800
And now it's like this happens to every piece of technology, uh including QKD.
00:34:32.880 --> 00:34:36.719
QKD will be a feature of uh the future telecom networks.
00:34:36.960 --> 00:34:37.920
That's so cool, actually.
00:34:38.079 --> 00:34:45.440
And that's if you if done properly, can make you know create a monopoly market, even if you'll be the first in doing it properly.
00:34:45.840 --> 00:34:47.039
Sounds really interesting.
00:34:47.199 --> 00:34:50.800
Sounds like a big, big opportunity that you're sitting right now on.
00:34:50.960 --> 00:34:56.000
Um and are there any like um hidden costs of deploying QQD?
00:34:56.639 --> 00:34:57.360
Yeah, exactly.
00:34:57.440 --> 00:35:02.159
So uh this uh what I just described really tries to solve those hidden costs.
00:35:02.320 --> 00:35:18.400
What people tend to forget is that if you go into a data center and it's trying to store one box at a time, you have the cost of the engineers, you have the cost of uh the engineers that your own engineer and the engineer that hosts you, you have an incredible amount of uh um operational costs.
00:35:18.559 --> 00:35:36.000
And really um the idea of developing cars that naturally integrates into the existing networks really aims at removing, fully removing those operational costs because they are um they are combined together with the cost of delivering um the standard telecom equipment effects.
00:35:36.239 --> 00:35:36.480
Amazing.
00:35:36.719 --> 00:35:50.000
So the so the hidden costs really of existing QKD are those of deployment, and in fact, probably they are bigger than the actually uh the cost of the materials, and this is really what needs to change.
00:35:50.159 --> 00:35:56.000
QKD needs to become a solution where the operational costs are effectively uh removed.
00:35:56.559 --> 00:35:58.719
Fantastic and reduced to zero.
00:35:58.880 --> 00:36:00.400
That sounds super exciting.
00:36:01.360 --> 00:36:13.119
In the big volumes, the idea is that they will be reduced uh to um very close to zero and will be uh made in such a way that can be automated and controlled and yeah, at the network level.
00:36:13.199 --> 00:36:14.320
Yeah, super cool.
00:36:14.480 --> 00:36:20.960
Uh Francesca, so can you tell us right now where are you with CADS in terms of the deployments?
00:36:21.199 --> 00:36:24.079
Um pilots, testing that you do.
00:36:24.239 --> 00:36:26.800
How many of such deployments do you have right now?
00:36:27.280 --> 00:36:29.360
Uh at the moment we have several deployments.
00:36:29.519 --> 00:36:39.920
Uh uh, I cannot share the numbers, but we have uh uh the the nice thing is that now QKD uh is really globally um uh explored as a technology.
00:36:40.159 --> 00:36:48.239
We have uh systems uh uh in Europe, in the UK, in um in France, um in the south of Europe.
00:36:48.639 --> 00:37:01.599
We have a solution in North America as well, and we've had uh and then we've had interest really globally for our systems that uh for for other technical reasons we're not able to always fulfill.
00:37:01.920 --> 00:37:28.480
But we have uh really um yeah, the interest for QKD now is global, and now really the steps is uh continue with these uh deployments for people to understand as an education tool, for people in the industries to understand how QKD QKD can be useful to their own solutions, and this will be a process that will grow and grow and uh until hopefully we will get the genuine commercial um deployments.
00:37:28.960 --> 00:37:33.760
Well, you mentioned several markets, uh that sounds super exciting.
00:37:33.920 --> 00:37:38.559
Can you tell us a little bit more, if it's not a secret, like who are those buyers or testers?
00:37:38.880 --> 00:37:44.159
Are they mostly telecom or are the government uh agencies as well, enterprises?
00:37:44.559 --> 00:37:45.840
Who is like the profile?
00:37:46.159 --> 00:38:07.679
At the moment, at the moment, uh uh it is mostly the telecom that uh use uh want to test our solutions as well as uh telecom equipment vendors that want to understand how uh such as, for example, um we recently made public the fact that our system is traveling with the Nokia's uh kits across the globe.
00:38:08.000 --> 00:38:21.119
Um and therefore we now have a system with the um fiber providers um and with the telecom equipment vendors uh of the likes of uh like Nokia.
00:38:21.599 --> 00:38:46.880
Um these are the two levels, and the end users uh at the moment uh it is those people in the supply chain that uh interact with the users, with the end users that could be banks and general finance sector finance and and the fans seem to be the sectors that seem to have the most the biggest urgency in understanding QQT as a solution.
00:38:47.119 --> 00:38:47.920
Following as well.
00:38:48.079 --> 00:38:53.920
Okay, and then may I ask you like, do you know what is driving the decision to start testing?
00:38:54.079 --> 00:39:08.480
It's like they want to get ready for the future in terms of telecoms, I can imagine, but but is this because they feel the real risk, or there is some regulatory pressure on some markets, for example, that even push them to do that?
00:39:08.800 --> 00:39:22.400
No, interestingly, uh good or bad, we don't know, but uh these tests are going uh mostly without uh uh a direct regulatory push on QKD.
00:39:22.559 --> 00:39:41.280
The push is always on is mostly on PQC at the moment, but people that work in the telecom sectors they understood the value of QKD more than maybe the governments at the moment or certain governments, and therefore that's really where the traction is coming from.
00:39:41.440 --> 00:40:00.159
They understand that uh in three to five years' time uh QKD might be a solution, and uh they understand that the risk of being late is a bit too big not to start understanding QKD to a deeper level, exactly, and the competitiveness, right, uh amongst them as well.
00:40:00.400 --> 00:40:00.639
Exactly.
00:40:01.360 --> 00:40:06.000
Okay, and then so can you tell us a little bit more about the sales cycle?
00:40:06.159 --> 00:40:08.159
Like how are you selling it?
00:40:08.400 --> 00:40:16.559
Um, how long is the cycle usually, and what maybe what kills the deal the most often?
00:40:17.199 --> 00:40:29.039
Well, the cycle at the moment, the way because uh generally speaking, there have been uh small batches of sold systems, and therefore so it's very uh very much a request and production.
00:40:29.199 --> 00:40:39.199
I mean, we we have a batch, we have a stock, a small stock was a request, and then when the stock was available, uh sale is a very because of the volumes.
00:40:39.440 --> 00:41:01.519
Uh, walk years, the deals um at the moment, these very days really is the cost sometimes, the cost of the system still, and uh is really part of our next uh phase of bringing down the cost of the really the bill of material um down to make it more more valuable.
00:41:01.679 --> 00:41:14.880
The cost it is important really to mention here that we've pushed a lot on the chips, and really the cost now is in everything else but but the chips is in the electronics, is on the electronics in those kind of uh aspects.
00:41:14.960 --> 00:41:24.079
But the chip, in fact, we got to the nice place where the chip, the photonic integrated circuits are not the most expensive parts in our system.
00:41:24.239 --> 00:41:30.400
There are digital electronics pieces that are more expensive than our chips, which is a very nice place to be.
00:41:30.639 --> 00:41:31.599
Yeah, yeah, definitely.
00:41:31.679 --> 00:41:32.400
That's so cool.
00:41:32.559 --> 00:41:41.360
And what do you think the price needs to be to be like you know, no-brainer for those telecoms to just say like, okay, 100% you on that?
00:41:41.840 --> 00:41:51.599
Uh the 100%, I won't be I cannot say the 100%, but certainly the price points need to be that of uh um existing um telecom equipment.
00:41:51.840 --> 00:42:05.840
So when uh basically how the traffic is moved uh in the fibers across the telecom networks uh is through uh this uh technology that lives inside the telecom telecom infrastructure.
00:42:06.000 --> 00:42:18.239
And depending on the specific uh application that uh um uh a piece of hardware needs to needs to do, the costs go from a few hundred pounds to ten thousand pounds to fifty thousand pounds.
00:42:18.400 --> 00:42:27.760
Depend where we believe QKD will be a viable solution is uh in the between around twenty thousand dollars, fifteen thousand dollars.
00:42:28.000 --> 00:42:32.320
That's the price point where we believe uh there could be an inflection point.
00:42:32.719 --> 00:42:33.360
Amazing.
00:42:33.599 --> 00:42:43.920
Okay, and you mentioned as well that the new uh potential use cases or usage of the chips are also in the defense industry, and that is very interesting for me.
00:42:44.000 --> 00:42:51.199
I would love you to um tell us a little bit more how the defense industrial defense agency can use your chips.
00:42:52.320 --> 00:43:18.159
Well, uh obviously the the nice uh um thing of uh of our solution is that uh um can be applied to, in particular, because of the integrated approach, uh can be applied, for example, nice applications are satellite-based solutions where uh the signals can travel um uh in the in the free space from the ground to the satellite back on the ground and therefore cover long distances.
00:43:18.400 --> 00:43:30.639
That's one way that uh uh and this solution is intrinsically harder to hack and therefore is um is a is a good way of establishing very secure connections across the globe.
00:43:30.880 --> 00:43:37.199
And uh in the past we also did a demonstration with uh um uh drone-based uh QKD.
00:43:37.840 --> 00:43:51.199
Effectively, uh, we had a transmitter QKD chip on a drone, and this was used uh and this uh was exchanging keys between the flying drone and uh and a ground station.
00:43:51.440 --> 00:43:59.760
These are two examples of application, and these are two examples of applications where uh effectively which mostly involve free space, so not the use of.
00:44:00.159 --> 00:47:26.719
fibers because the fiber will be constraining the the direction of the sign the the positioning in uh in uh in uh in uh let's say um what and and and the real deployment uh and the role as well this is very question what you said because i think we we don't think about it i think because of the war with ukraine everyone gets to use to that drones are like you know integrated part of the war in um self-defense and also um the uh on the attack side and they obviously have no human involved but nobody thought about it that actually this connection can be overtaken as any other connection with with technology so you can protect that communication with the machine yeah and also and also the the direct uh the direct optical communication in itself has the advantage that it's much harder to hack anyways compared to radar signals and therefore much harder to jam which is also another another um yeah parallel feature of of optical communications in general exactly because if if hacked it could be very destructive I mean very yeah instead of attacking your enemies can attack you uh back but wow that's very interesting okay so um knowing all of these potential use cases especially for the defense um defense sector um is this it is is your roadmap kind of driven by the government and defense demand or is something that um you're just doing on site anyway at the moment uh at the moment uh um we yeah there is interest in the government and there is uh interest in the defense we we believe yeah the the telecom in general infrastructure as such is probably that that will see the the the the the more direct uh the the the beach market yeah and uh we're talking about uh probably the banks is still probably where we will see the um the biggest big the the earliest uh utilization of of uh of qkd systems rather than defense is certainly valuable but obviously um there is there are other levels of complexity that uh might stop the the understanding so the user telecom would be the the beach hat market for you and uh defense or or banking as well could be the secondary market in the future but it may change in the future right we don't know yeah yeah yeah we don't know uh there's a lot of tensions going on right now globally and the political geopolitical situation is pretty tense so you never know how this may accelerate and and maybe potentially change the demand right and as a startup we usually try to adjust okay uh Francesco we're going slowly to the end of our discussion this this is the time I think for for for asking for for me asking you on the predictions so what do you think is gonna happen in the next five years if you can imagine um qkd and also um post quantum cryptography in the next five years where do you think we will be yeah so the way uh I was asked this question recently and the way I replied is that I was asked where I would see QKD where I want to see QKD in the next five to ten years.
00:47:26.880 --> 00:47:33.920
My answer was that I would like not to see it because I would like it to be really seamlessly integrated into the existing telecom networks.
00:47:34.239 --> 00:48:51.519
So we don't have it yeah and really this is uh where I believe uh QKD will be in the next uh within the next five years qkd will be a feature of existing uh of standard infrastructure will be uh widely deployed and will be up to the um and based on the uh on the needs it will be offered as a sys as a solution to to the um to the to this to the sector that uh require them so really where I want to see where I dream to see QKD is really as a standard uh well highly integrated component onto the existing uh infrastructure as it is anything in in the the in the networks and in comparison uh with uh PQC PQC will will uh cover those uh the the a wider possibly a range of of solutions particularly in the mobile um in the mobile landscapes but our QKD will be used uh in all those critical um places that where security is absolutely paramount where there is no there is no optionality between um choice A and B but where you need all the layers of security for for your application.
00:48:52.480 --> 00:49:04.880
That perspective of the QKD being so common so obvious and um standard um it looks very attractive and I hope I wish that will happen.
00:49:05.280 --> 00:49:24.639
When you were speaking about the the security of the data one of the things that came to my mind straight away was the medical data because right now there's a lot of you know institutions that store our medical data and it's happened all the time the leaks.
00:49:25.280 --> 00:51:04.480
Yeah right and that should not happen definitely so do you think QKD can also have a um absolutely 100% and in fact is uh is uh yes 100% and this is very true the reason why qkd in in this particular instance is uh is incredibly uh useful is also because as opposed as to many other applications usually the health sector actually owns fibers they own networks and therefore qkd is even more easily integrated in this kind of infrastructure because basically the solution is already there and what you need to add is just the terminals and therefore really is um yeah the the certainly the the health data of of patients across across um countries is certainly um one one very useful one very one uh very good uh use case yeah so in summary if the telecoms are properly integrated um qqd in the future we all will be safe yeah that sounds good and very optimistic francesco thank you so much for the discussion today um it was a really pleasure um and thank you so much for being with us thank you thank you very much that was blonde and quantum thank you for joining me on this journey through the quantum business frontier if you like the episode please review another Spotify or Apple Podcast and help more people discover the quantum world without needing to untangle the theoretical physics see you next time unless the cat change the time
