Entangled Things
Entangled Things
Episode 153: Daniela Herrmann on End to End Quantum Driven Solutions
Use Left/Right to seek, Home/End to jump to start or end. Hold shift to jump forward or backward.
In Episode 153, Daniela Herrmann, CEO and Founder of Dynex, joins Patrick and Ciprian to talk about making quantum accessible to businesses that have no interest in understanding how it works. Dynex takes a problem driven approach — customers describe their computational challenge, Dynex identifies whether it’s a quantum problem, builds the solution, and delivers a working MVP in six to eight weeks, running on Dynex’s own proprietary soft and hardware platform. The conversation covers the breadth of use cases Dynex is pursuing across logistics, trade optimization, materials science, and healthcare, and why Daniela believes the market is ready even if the industry has been communicating badly. The team also discuss Dynex’s Zeus chip prototype, currently in benchmarking with ten thousand logical qubits, the coming pressure on cryptographic infrastructure, and why Daniela expects the quantum timeline to surprise the market again.
In episode 153, African Cyprians speak with Danielle Harmon, CEO and founder of Dynex. The team discussed Idex's end-to-end approach to quantum-driven solutions. This news chip and its 10,000 logical qubits, and why Quantum is already at the start of the next mega cycle of computing. Welcome to Entangled Things, your quantum computing podcast, hosted by Patrick and Cyprian.
SPEAKER_01Hey Cyprien, how are you doing? Hey Patrick. I'm doing good. Looking forward for another great episode of Entangled Things.
SPEAKER_03Yes, the guest today is someone I had the privilege of meeting at a conference in Boston. I'm very excited. Daniela, do you mind introducing yourself to our audience?
SPEAKER_00Yes, absolutely. My name is Daniela Herman. I'm the CEO and founder of Dynex. Dynex is an expert in developing quantum-driven solutions and delivering them end-to-end and powering them on its own technology, its proprietary technology, which is to be accessed on a platform infrastructure.
SPEAKER_03That's amazing. Now, we very rarely talk to people who are doing it end-to-end. Is there a reason that you've had that approach? Because it seems like you're taking on the whole burden.
SPEAKER_00Well, the whole burden, not. I wouldn't say that. We're taking the burden on to the minimal viable product. And then, you know, once that product, minimal viable product, has met its scope of work, it gets um handed over to those that do the integrant integration implementation. You know, so let's imagine it's a huge company and they have to then implement it throughout 300 departments or whatever. That's not our business, right? We do really minimal viable products. So end-to-end means that we offer the development because we learned very early that the industry needs support with this. Uh, and it's better that we do it than that we explain what quantum is. So uh that has led to a great business model. So we develop. And once we developed the quantum-driven solution and the MVP asset has matched its scope of work, it runs on our machines. So that means machines not in the sense of computer machines like the quantum computers. It runs on the soft and hardware, which we have developed as our own proprietary technology at Dynex.
SPEAKER_03It sounds like you're solving a problem we see with people who are just getting into AI, which is they just don't know what to do with it or how to leverage it. It sounds like you're taking that burden off your customers to get them started on the quantum path.
SPEAKER_00Yeah, well, um, so it's really we are actually doing a uh it's an interesting approach we chose, and we chosen them based on an experience we made because when we came 2024 and myself was on stages 2024, um, you can imagine people had projections that Dynex uh or that me uh um uh when I said when I said the quantum is here, people looked at me, the experts, and said, okay, uh keep on dreaming because the projection was 50 years from there. Jensen said 50, and others said 20, 25 years from now. And I said, no, it's here, it's at scale. But what we quickly realized when we confronted the industry is that they in conversations shied away because they thought they need to understand all the quantum, right? Right. And as they didn't, and also their resources and capacity, and also mainly in the industries right now, the CTOs are AI experts, but not quantum experts, right? Because that has been the trend. And the unexpected trend overlapping this trend, which is quantum, which is here, was unexpected. Hence the capacity on the CTO level is partially also not covered with quantum, right? But that the and we learned very quickly that we need to communicate and show them that that doesn't really matter because you don't need to be a physician and you need to understand quantum. You need to tell us your computational challenge. And when you have your challenge, then we identify if that challenge is a quantum challenge, right? Can be logistical, like in logistics routing, can be trade optimization, can be a weather prediction challenge, you know, just different challenges. And then when he uh we even made it a bit a process that we say, write that down on a one sheet of paper, two, three, four challenges, and describe your technological embedment. And that is enough for us to judge which one would be the best quantum challenge to start with.
SPEAKER_04Right.
SPEAKER_00And then we develop the conceptual frame, we develop the solution, and we give him or her an API to access it. Then they do a testing with us, and after six to eight weeks, latest, they have a fully usable minimal viber product.
SPEAKER_03That's amazing. Um, yeah, I the the analogy comes to mind of I don't need to understand a carburetor to push on a gas pedal.
SPEAKER_00Yeah, no, excellent. Or if you, I mean, you can take, I always took also the the example of using your mobile phone. I mean, if your grandmother is on your phone screen and you're walking around in the kitchen talking to her while looking at her, and she looks at you, I don't think you want to waste one second to understand why this is possible.
SPEAKER_02Right, exactly.
SPEAKER_00And that's exactly where I see quantum going as well. Quantum is logically, we are at the next mega cycle of computing and it has begun already. It's not just coming. So that's like when the entire computational industry uprose, right? We still all remember that. It's exactly we're at that kind of intersection again. So it's not that we have to now, you know, kind of see this as a continuation of what was now 20 years. It's really a new episode, and that episode partially is complementary to the old compute, I say, and partially it is not complementary and replacing because it doesn't make sense to do a simulation on an AI if you have, for example, suddenly different simulation tools that are just more effective, right? And efficient.
SPEAKER_03Cool. Is there a is there a technology you guys use predominantly, a language, um, a platform uh modality, or is it dealer's choice? How does how does that get chosen?
SPEAKER_00No, I mean, you know, the platform architecture which we've built, okay, is qubit agnostic. It means that once you have your API running, you could also choose different providers, right? That have quantum solutions for you to compute on.
SPEAKER_04Flexibility. Yeah.
SPEAKER_00There is a flexibility, and it shows you also the different pricing levels. Just one thing is really a fact at that scale, which we are doing quantum-driven solutions right now with the quantum uh with the quantum-inspired technologies we have, which are the algorithmic qubits, 1 million algorithmic qubits and 10,000 probabilistic qubits with the Apollo, which is a chip at room temperature. Those uh capacities are uh are most probably strong first mover advantage capacities.
SPEAKER_03When you say probabilistic qubit, do you mean logical? Is that I I haven't heard that term. No. You mean you mean physical qubits?
SPEAKER_00So yes, so so it is not logical qubits, no. So the probability no, that is uh that is still a probabilistic model.
SPEAKER_03Okay, so there's still uh there's still uh error prone qubits is the count that you're talking about. Well I what I one of the things that we noted recently was the it about two years ago when you were doing your speech, uh no one qualified the types of qubits because there were very few or no logical qubits at the time. Now we really do have to watch the that terminology, and we're seeing it more and more in articles where they specifically call out that they're physical qubits or they're probabilistic qubits versus logical qubits because we're getting a lot more of the logical. And um, and and that's uh kind of a break point that you're describing, I think. We're we're getting into the point where we have working systems with a with a number of qubits that is not, you know, three or five.
SPEAKER_00No, no, no, absolutely that are yeah, we're well past that. Yeah, so that's well past, and the logical qubits we have brought to market, not to market yet, so for market use, but we have a prototype that is called Zeus, and that Zeus chip has 10,000 logical qubits, and we are doing different testings on benchmarks currently, so there is nothing out yet, right? And the so it's really a project like others have it as well right now. But uh once that one is uh in part of the platform infrastructure we have, uh we are well complete for the moment with regards to technological innovation in quantum and quantum-driven technologies.
SPEAKER_03Cool. I had to smile because uh my 11-year-old Weimar Honor is named Zeus.
SPEAKER_00That's good. It's a good name.
SPEAKER_03Good name. Um, so is there a is there a date when you think that will be done? I I it's probably too early to name that. But uh I know a lot of people are eyeing 2028, 2029 for things.
SPEAKER_00Yeah, so I would say in the next one, one and a half years for sure. And if and if you look back to our track record, we normally surprise the market much quicker than we said we will.
SPEAKER_03Yeah, things are things are happening. You can't promise it, but it's happening very, very quickly.
SPEAKER_00Exactly, very quickly.
SPEAKER_03So Cyprian and I are both developers from you know from way back. I'm I'm floated into develop uh away from development and into um management more, and he manages teams too, but he's still more hands-on keys. Um, so the development side of things, are you working with the and educating developers on the client side, or is it really they just want just give us a function, give us an MVP?
SPEAKER_00Yeah, exactly. So um if you really want to keep it tight and people excited and you want to work along the KPIs that the people are confronted with in the companies, right? You need to find a solution that takes six to eight weeks and turn this around and make it a success and a confirmation of their valuable use. Because if you start going, I mean, others are doing that, going into educational and education developers and said, A, it's not our expertise. We do have, we developed with our independent family office uh foundation, we developed a uh QQ Edu called Quantum Education Online, because it was for us important that people have easier access to quantum education. Right. Which, so that's out there. Yeah, but it's not it's not our expertise to start having workshops and education.
SPEAKER_03It's an ongoing, but not the primary purpose.
SPEAKER_00No, no, no. It's really we focus on extraordinary powerful technologies, solutions, software, hardware, and we focus on quantum-driven solutions. It's actually very simple, right? You tell us a channel challenge, we do the solution, we power it on our soft hardware, and that's it. And it's a subscription model, a kind of a script subscription model. It goes after, so once you compute, it goes after quantum calls. That is a little bit per definition. So each client customer, let's say if it's a, let's say if it would be a logistical, high, high frequently using uh uh logistical company, right? Like a delivery, for example, company, then you would do um per call very short calls, right? And then the movement of the call would be constantly, let's say per second, and hence the pricing would be a different one to the pricing if somebody bundles once a month a call.
SPEAKER_03Right.
SPEAKER_00Right? So depending on the calls and the complexity of calls, we normally identify that in the first month of customer relation and then build out a framework for standardized uh subscription. Let's call it for a standardized subscription framework. But you vary uh at a minimum, if you do little calls at a minimum of 5,000 Swiss francs, and you range if it's a solid bundle once a month or two, uh, you know, and um and and you let's say you're um you want to do 100 calls to 50 to 50 francs, we're at that 5,000.
SPEAKER_03Right. That makes sense.
SPEAKER_00And also the also the developments, I mean, I think it's important to state because also the developments, if you think about it, we're ranging between you know, smaller developments, 50,000, right-ish, and then logically it goes up to million, million five hundred if it's higher complexity, or then customize, right? Yeah, but at the end of the day, I mean, these are nowhere in relation to the costs that you have when you start, you know, needing to build quantum infrastructure in your company.
SPEAKER_03Yeah, it sounds like you're bringing this to small to medium businesses where every every news item I see is a billion-dollar oil company is doing this or Microsoft is doing this. It it and I think that's that's very important.
SPEAKER_00It's very important. I mean, any technological breakthrough in the past over the last century, if you look, if it can't go mainstream, it won't fly.
SPEAKER_03That's true.
SPEAKER_00Right? And AI was the same thing, it went mainstream, and that's why it started to fly. And it's also besides that, there is a philosophical touch to it. It's the right for the commercial, smaller, and mid-businesses as well to start having quantum as a tool because there are huge advantages and huge arbitrage you can reach with that, right?
SPEAKER_03Absolutely.
SPEAKER_01So a question that I always have um to folks who are working directly with with uh lots of customer organizations is how where do you think the market is in terms of being open to using quantum, understanding the need, and maybe even further down a little bit the the pipeline of being able to understand how to address their problems or where to think of quantum. Um do you think the market is is has been evolving for the past five years or so, or we're still at the very early kind of stages of that?
SPEAKER_00So I I think it's it's not the people not ready, I think it's the mistake of the industry that has happened, and it's a bit similar to blockchain, you know, they came so complicated and so um also academic into the the main crowd that everybody thinks, oh my god, until I understand that, oh my god, you know? And I think that's so I would like to divert from your question and actually say, I mean, almost everybody's ready. You just need to guide them differently. Because if you ask them what's your challenge, ask a logistic company what's your challenge, ask a shipping company what's your challenge, they all know what's their challenge, right? With regards to coordination of routing and multi-vehicle use and all of these things. As soon as the operation becomes more complex across nations and different vehicles and different points and whatever all is there, as soon as that happens, you have optimization challenges and you can't handle these with your computers. Think of the reinsurance sector or the insurance sector as well, right? I mean, they need to break down their data because they're they have too heavy data to pass along from one step to the other. And so I think it's not really um that they're not ready, it's how you approach people that they understand that what they need to understand is do they have a computational challenge that will, when they solve it, improve their competitive edge in which they operate. And almost everybody's when you help them. Yeah. And that's like if you I mean, think of the Queens, I mean, it just as an example, cosmetics. I mean, today, uh, you know, the number of billions of people on this planet all have three types either you have a greasy skin, a dry skin, or a mid middle skin, right? But that's not true, everybody is different. If you would start using quantum, you would have individualized, customized creams for every type of skin type. Right? But you can't do that today, high production, because that's much too complex.
SPEAKER_03Right. Yeah, we we we envision getting to just in time production, you know, where I get exactly the sneaker that I need, yeah. Which is different than the one that Cyprian would need.
SPEAKER_00Exactly. That's the thing, and then so I do think it's it's yes, you're right, it's in the young shoes, the whole industry. It's the beginning of the mega cycle, which is also a tremendous opportunity for everyone, right?
SPEAKER_03Now and I I was gonna say, I are you I and I I don't want to oversimplify this, but I I'm trying to wrap my head around it from the outside. Are you looking for local minimums in this? Is it the traveling salesman problem for in many cases, or is there is it a machine learning aided by what's the what's the most common, and maybe there's there is no most common, it's it's always a different approach. But I'm trying to make sense of like how how does quantum help with these?
SPEAKER_00Yeah, these use cases. I mean, traveling salesman for sure confirmed, right? Was also was also calculated by us. And then what you have as well is um in trade, you know, the different trade optimizations with higher complex positions, numbers of positions, right? Becomes very quickly complex. Then you have the medical field where you can say multimolecular analysis becomes super complex. You have, but you have also, for example, you know, like in the car industry, the dummy testing, right? Of the of the dummy puppets, etc. That's also a very, very complex computational challenge. So then you have um, I mean, classical is also logistics, classical is um material science, you know, combination of different materials that you maybe didn't think of because you weren't computationally able to simulate, right?
SPEAKER_04Yeah.
SPEAKER_00Um, so there's there's really a lot of use cases out there.
SPEAKER_03Yeah, and I I again I I it goes back to I think that even for people who are exposed to this, if they're not in the trenches like you are all the time, they don't know where this hammer goes, what you can do with this electricity when it was first put in the homes, people didn't envision, you know, the machines that we would use it for, televisions and things like that that eventually came. So I think we're still in that early phase where we're still using quantum mostly for light, like we did in the electricity days.
SPEAKER_00Yeah, yeah. No, I mean it's still you, as you say, it's still at the beginnings, right? And and um and uh and evolving. But I well, I think one entry point is very interesting, for example, is the correction overlay of with quantum over AI. Because as soon as you address a pain point, everybody understands that there is too much electricity used, that there are too much GPUs needed, that there is capacity constraints, that there is whatever all, right? I mean, precision, for example, also still with AI, right? Uh security, I mean, that we all read as well. Um, so um, if you do a quantum um correction overlay, you will experience as an AI infrastructure provider, right? That you will have an improvement of costs, you will have precision, you will be able to create and establish certain security concepts, and and and right. So that's for example, a I'm I'm promising actually, I think that's a promising um tool because it's it's already in need, in a very specific need, and it's it's it's it's it's on top of the trend understanding AI, right?
SPEAKER_03Yeah. And you you pointed out security, which is my my area. I run a uh consulting company that does uh pen testing. Um Uh Shore's algorithm is hanging over us like a sword of Democles to some extent. And but it's also an opportunity to make things better to improve our encryption. Does that touch on your work or you mostly stick to the other sides of things? Or do you get involved in post-quantum cryptography?
SPEAKER_00Um Well, I mean, we we're not like we're not like in groups participating, etc. But you can you can you can with a little bit of research, um I mean, you know, you can see that what what was what was uh over the last few years thought that you needed, you know, a huge load of logical qubits to ever get in that region came down to very few.
SPEAKER_03Yeah, 10,000 logical would probably do it.
SPEAKER_00Well, I I I wouldn't want to kind of uh state that as such, but you know, you're i it could, right?
SPEAKER_03Maybe there's been papers that say that you need like uh if you get it down to a very high level efficiency depending on circuit depth. Yeah, a few thousand.
SPEAKER_00You're totally right, yes.
SPEAKER_03But we're still not efficient.
SPEAKER_00And I think, and I think, you know, the prognosis, I mean, what I can say, and where I always push back, is when I hear somebody saying, yes, that's coming in 2035. I mean, that is really that is just that's just not true. That is gonna come much earlier.
SPEAKER_03I agree.
SPEAKER_00And I it is a fact that people are not ready. It is a fact that even the quantum keys, which are already, you know, you can have them from IOMQ, you can have them from us as well. We have quantum keys in our web shop. I mean, quantum keys you can have already today, and even that is not being put into place, right? And that's a quite a simple fix, right?
SPEAKER_04Yeah.
SPEAKER_00Now, that whole thing, what you're saying with regards to um the logical qubits, I mean, logic as much as it would be groundbreaking in the sense of breaking, really groundbreaking, um, in the sense of um, let's say cautiousness that would need to operate with it, right? As much as it's a it's in a tremendous opportunity, right?
SPEAKER_04Yeah.
SPEAKER_00Because as said, it becomes the definition of new materials.
SPEAKER_03Yeah. What I think is gonna happen with the with the shore is that it's going to happen a little bit at a time and then all at once. There's already talk about getting to the point where we have almost enough logical qubits to break the elliptical curve that protects some crypto. And that's gonna cause a bit of a panic. And that's gonna cause a rush of education, which I'm kind of looking forward to the rush of education. I'm not looking forward to the panic of the.
SPEAKER_00Yeah, yeah, no, and and and and as you say, you know, it's like uh, you know, we've I I think, I mean, we've had this all in our history, right? And especially when when certain cycles get hit by a technological surprise, for example, then you have this uh rush of panic, then you have a breakdown in a market, then everybody recovers, and then there's a new paradigm in that market, right?
SPEAKER_04Right.
SPEAKER_00And and hence I think that will that will be a bit the shaker, you know? Yeah, I think I don't think we will I don't think we will avoid the shaker because it it's just not enough separation that has been done.
SPEAKER_03Yeah, I I it's scary that we're waiting so long, especially with some of the threats. And I don't mean a term.
SPEAKER_00You know, and that is and that is really something where I must say, where the communication, the PR on quantum, was led by groups wanting to dominate the market, which is clear, right? Yeah, but but I mean, you should never only listen to like let's say, yeah, one angle, right?
SPEAKER_03Right. Yeah, you need multiple perspectives. I would think that once a company, once an organization has dealt with you and gotten results from quantum, they're more open to see what else it might.
SPEAKER_00Yeah, no, it's actually it's it's funny what you're pointing to because when you get into when they relax and know that A, we're not gonna absorb, I don't know, 50 people resources for six months losing KPIs because we're doing some kind of weird project, right?
SPEAKER_04Yeah.
SPEAKER_00Um, once they understand that and they actually can only express their challenge, they normally, anyways, come with two to three to four challenges, right? Yeah, and then you have actually already a project pipeline because if the first project runs great, you can be sure that they're fascinated and then they want to do a second project.
SPEAKER_04Right.
SPEAKER_00Right. We tested that all intensively over the over the last one and a half years, right? With since 24, with you know, use cases that we also presented with video material to at least at for the moment show the people, hey, this is for real, right? Right. So so yes, so you will automatically once they get a feel, and and I can see that already in conversations we have always because we then have weekly short conversations, but people suddenly it it clicks very quickly. Yeah, because it's it's an additional creative um mind drill that you're giving a situation, right? Where you suddenly understand you can, it's actually like if you're understanding you can play more, and why? Because you can see more, right? I'm always saying at the end, at the end, what quantum is doing, everything has been quantum. Your body is quantum, right? Right. I mean, that you're getting up in the morning is quantum, yeah. You're a quantum computer, right?
SPEAKER_03Yeah, every the whole whole universe. I think MIT has a pendulum in one of their um their main halls that is actually quant that it measures as a quantum system.
SPEAKER_00Yeah, and and it's like 17 kilograms. Yeah, and it's nature, you know? Yeah, and so I think what what goes what is so funny to see, what goes so fast suddenly is that everybody in its origin understands it.
SPEAKER_04Yeah.
SPEAKER_00And actually, actually the forcing us to think linear for so many years was more difficult actually in our nature because it's not our nature.
SPEAKER_03Yeah, probably I mean, we live it's funny, we we've come to deter expect the world to be deterministic, but in our daily lives we know it's probabilistic. I might win the lottery, I might get hit by a car. Yeah. We we it's funny how we've kind of trained ourselves to think deterministically when when our experience in most things is probabilistic.
SPEAKER_00And and the interesting thing is with a deterministic uh view is also that you are already so in your brain fix on that that when it comes to probabilistic, you're almost not wanting it to be true.
SPEAKER_03Right. Yeah, yeah. Unless it's unless unless you like the outcome like the lottery.
SPEAKER_00Exactly.
SPEAKER_03So I I I I want to turn real quick to like I know you guys are based in Switzerland. Do you do work everywhere? Because I did see you, I did initially meet you in Boston. Do you do work in the United States? Do you do work around Europe, or you focus more locally?
SPEAKER_00Yeah, so um, first of all, I love the United States, and um I am just now next week going again. And um we are um we are building relationships in the United States as well, building relationships currently, right? How this manifests structurally in the future, we'll see. We are today a Swiss group, fully, right? And how this will be structured in the future, let's see, right? We do have some also capital market fantasy that we're thinking of, right? But this is all this is all right now in the mix and how it how we work. Um, we've been always like a typical tech company, you know, people are um where at least where they are allowed to be, but um they're all around all around.
SPEAKER_03Excellent. Well, like we hope we'll we'll definitely be seeing more of you on the podcast, we hope, and uh, and the company. We're we're coming up on time. We're not we still have a little bit of time left. Is there anything you want people to know? Any any events or or something you guys have published that you want to talk about? Uh we we got a little bit of time left. I want to make sure we we hear everything you carefully.
SPEAKER_00I mean, I mean, we're certainly very proud that we were just mentioned uh in a peer review uh in a health journal, Jamir, and it's just published. Cool. And and yeah, that's very cool. We're very excited because that's not so easy to get in. And we're also on the National Health Library now um linked, so that's great. And um, other than that, I just uh would love to have more people come with their challenges, also mid-sized companies. The situation is not just completely unaffordable, so I think it should be everyone should be looking into this because as a very small company, you can today create a competitive edge you've never thought about using quantum.
SPEAKER_03You're either going to be the disruptor or the disrupted. So if you don't get ahead of this stuff, then you might find a competitor doing it, and uh and that's not a good place to be.
SPEAKER_00Exactly. And then embrace it. Quantum is here.
SPEAKER_03Excellent. Well, Daniela, I'm very happy I ran into you into that conference in Boston. And uh I'm very glad that you joined us. Hopefully, you know, we'll see you again in in 2027. Uh, and and thanks again for giving us this time.
SPEAKER_00Thank you so much, everyone.
unknownOkay.
SPEAKER_03Thank you. It's been a pleasure. Bye, everybody. Bye, everyone.
SPEAKER_00Bye, everyone. Thank you.
SPEAKER_03See you next time.
SPEAKER_02Cybercrime is one of the biggest threats to businesses of all sizes and industries. With almost half a million open cyber positions, the problem is compounded by the lack of available talent in the marketplace. At Pulsar Security, our elite team of highly credentialed experts collaborate with you to assess your current defenses and develop solutions tailored to your specific needs. With services ranging from cybersecurity education to advanced penetration testing and red teaming, you can start reducing your risks today. Visit pulsarsecurity.com and let's secure your digital future together.