Quantum Computing Master Class & Quantum Day Insights

In this special Silicon Futures podcast episode, Karl and Jim provide a master class on the modalities of quantum computing and the state of their development and commercialization. The trio then provide their insights and analysis of NVIDIA's Quantum Day 2026.

The following topics are explored in this episode.

➡️ Quantum computing landscape and modalities described. (5:15)
➡️ The pros and cons of quantum modalities. (15:57)
➡️ Comparative maturity and capabilities of quantum modalities. (17:42)
➡️ The quantum continuum and the race to 1 million Qubits. (19:16)
➡️ What is a quantum application? (25:38)
➡️ Thoughts on how the quantum industry will evolve and the game play out. (26:50)
➡️ What is the role of accelerated computing and AI in quantum computing? (30:54)
➡️ What is the AI opportunity and business model for quantum computing? (35:44)
➡️ The quantum computing ecosystems and emerging business models. (38:56)

Hit Leonard, Karl, and Jim up on LinkedIn and take part in their industry and tech insights.

Check out Jim and his research at Tirias Research at www.tiriasresearch.com.
Check out Karl and his research at Cambrian AI Research LLC at www.cambrian-ai.com. Check out Karl's Substack at: https://substack.com/@karlfreund429026

Please subscribe to our podcast which will be featured on the neXt Curve YouTube Channel. Check out the audio version on BuzzSprout or find us on your favorite Podcast platform.

Also, subscribe to the neXt Curve research portal at www.next-curve.com and our Substack (https://substack.com/@nextcurve) for the tech and industry insights that matter.

NOTE: The transcript is AI-generated and will contain errors.

DISCLAIMER: This podcast is for informational purposes only.

Transcript

Karl Freund 0:06

Next curve.

Leonard Lee 0:10

Hey everyone. Welcome to this episode of the Nextcar Rethink Podcast. And of course, this is our special Silicon Future Series, which I do in partnership. In collaboration, wonderful collaboration with the illustrious Carl Fre of. Cambrian AI and soon to be quantum research. Right. It became even quantum,

Karl Freund 0:35

I think.

Leonard Lee 0:36

Ah, yeah.

0:37

The

Karl Freund 0:37

next big thing.

Leonard Lee 0:38

Yeah. And then of course. The, quantum dotie.

Karl Freund 0:46

Are you ever gonna

Leonard Lee 0:46

run Jim McGregor for

Karl Freund 0:47

me?

Leonard Lee 0:50

Jim McGregor of serious research Clan mcg. Yeah. Okay. Clan McGregor. Yeah. Who has, some amazing, amazing kids? who are just remarkable programmers. I I saw Damien, by the way, at, Katie Cadence live, 2026 and, yeah, you have really cool kids, man.

Jim McGregor 1:14

Thank you.

Leonard Lee 1:14

anyway, welcome everyone. And of course we cover, all the exciting stuff that you need to know in the world, the semiconductors, ai, and apparently now quantum.

Jim McGregor 1:28

Quantum.

Leonard Lee 1:29

Jim and Carl have been going deep into all this quantum stuff that everyone is starting to get pretty excited about, and, they're gonna basically break down the landscape. and so this is gonna be a great education for me. and, these two gentlemen are going to basically lay it out. This is what the landscape looks right now. And this looks, this is how you need to structure your thinking around what the market will look like and how the technol, the competing technologies actually, right. There are many different approaches, how those will, likely, and may be. Play out in the future. And so it is a really interesting area. It's formative, I believe. I don't know if you both agree, but I'm really excited, to have this, presentation, by the both of you on this, emerging. Well, do you wanna call it emerging topic or is this just something that is hot but people need

Jim McGregor 2:30

to It's pretty one, it's one of the most important topics for the next decade.

Karl Freund 2:34

Yeah, absolutely. Totally agree. Totally agree.

Jim McGregor 2:38

And then we get into neuromorphic computing. So we can do even more there.

Karl Freund 2:41

Yeah. More. Okay.

Leonard Lee 2:42

Yeah. We'll have another, talk on that. Right?

Jim McGregor 2:45

we're gonna make his head explode, Carl, if we keep doing this.

Leonard Lee 2:47

But you know what? For those of you who are listening on audio, you really want to go to YouTube and watch this episode.'cause we're gonna have some graphics and, if you follow us on LinkedIn, when we post, on LinkedIn or, other, social. Outlets will, provide the presentation material. But, for those of you who are just listening in, you really want to, check out the YouTube, episode in the video so that you can see, the charts that these two gentlemen are prepared and they're quite pretty. You guys did a very good job. Pretty job. Yeah. Well, thank you. You are pretty too. You guys are very gorgeous in that regard. So,

Jim McGregor 3:27

well, when you have a face for radio, you have to make your slides look good.

Leonard Lee 3:36

That's a good one. Right. Okay. So on that note, before we get started, please remember to like, share and react and comment on this episode and also subscribe here on YouTube and Buzzsprout. Listen to us on your favorite podcast platform. Opinions and statements by my guests are their own and don't reflect mine. Are those the next curve? Although I think I'm, we're gonna be in very good alignment since you guys are really. Going to be giving a masterclass here. So I'm all ears and I think the audience will be as well. And, opinions and statements, like I said, are their own, but we're doing this for informational purposes and in this case it's educational, it's gonna be great. we do this to provide an open forum for, discussion and debate on all things, ai. Semiconductor industry and now Quantum. And so with that, gentlemen, Where should we start this conversation about Quantum?

Karl Freund 4:32

let me start, start with my, if that's okay. They're a little higher level and Jim gets into more detail. It's actually been a big week this week, or last week, I guess since we posted next week. because, this week was, quantum World, world Quantum Computing Day. On, Tuesday, right, Jim?

Jim McGregor 4:49

Yeah. Supported by Nvidia. It was NVIDIA's Quantum computing

Karl Freund 4:52

Nvidia had their second quantum day, not associated with GTC this time. It was a standalone event with Jensen Wong speaking, and some of the leaders in Quantum, that are, at least those that are friendly with Nvidia, they were there. Those that are not as friendly weren't there, but, it was really great. Nvidia made some important announcements.

Jim McGregor 5:12

Yeah.

Karl Freund 5:12

With their new, quantum model for basically improving the error detection and error mitigation correction. The problem when an error happens, where did it happen? And that's where GPUs can help quantum a lot. So there's a really emerging view that GPUs and Quantum are not gonna be competing technologies. They're really gonna be, collaborative, technologies that help bring quantum to the useful quantum computing era. So what I've done here is just created a slide,'cause this gets very confusing for all of us, of what is going on in this marketplace and how can you have this many billions of dollars pouring into a market that's currently producing. Call it zero revenue. but, as Jensen Wong says, the most interesting markets of this, the markets of$0 that will grow into billions of dollars. And many of us are firmly of the belief that is exactly what will happen with quantum computing. It's not gonna replace ai. It will facilitate ai, and AI will facilitate quantum. It's a reinforcing, cycle here, it's virtuous cycle. So on the left you have the modalities. By modality, you have basically mean what kind of physics are these guys using to create these quantum states? And the next, column is basically in plain English. How's it work? what's this stuff all about? Next column is key companies. And then finally, the maturity level. This is saying how close were they to producing quantum advantage. By quantum advantage, we mean that you could solve meaningful problems in a cost-effective way that you cannot solve, in, digital computing technologies regardless of how big your supercomputer is. so let's start this. I'll try to do this quickly. You have superconducting qubits. This is the one that's been around the longest. This is where IBM and Google ti, as well as IQM and OQC. this is where you see those huge chandeliers of gold that are used as. Final stage of cooling that creates a, sta a a temperature of millikelvin. So this is colder than outer space. And that is where the actual supercomputing chip, superconducting chip, excuse me, can operate. Mm-hmm. And so you have these back rooms full of these huge refrigerators that eventually chill down through these chandeliers. To the bottom where this chip is gonna operate. The advantage of this technology is it's really fast. the disadvantages. It has to operate ultra low temperatures, which is very expensive in space, consumes a lot of space. but, it's also, unfortunately very noisy. so if you think about a normal semiconductor chip, it's okay that you're, you know, 10 10 to the minus seven. But here we're talking about 10 to minus two or three, so that means one, one in a hundred or one in a thousand results are bad. And considering you're producing hundreds of millions of results, that's not good. So you have to identify these errors and you have to correct mitigate these errors, which means reduce their impact. But you also eventually have to correct these errors. This is often done by teaming together physical cubits to create a logical cubit. The ratio of the, of which is super important to both speed and economics. if you can do. Two or three physical cubit to produce one logical cubit. You're gonna be three times more powerful than one. an architect that maybe takes 10 cu logical qubits or a hundred logical cubits to produce one, excuse me, physical cubits. To produce one logical cubit. Logical cubits are what matter. that's the, if you can try to distill it down and across all these modalities, logical cubits and their performance and costs are really what matter matters here.

Jim McGregor 9:05

It's what you're using. It's really what you're using this form, the gates, the circuits, the those, the circuits, random circuits that you're actually developing. And by the way, if you've never seen pictures of the Poughkeepsie Supercomputing. center in New York that IBM has, it kinda looks like an invasion of the body snatchers. But all these hanging refrigerators.

Karl Freund 9:24

All these hanging refrigerators, they're huge. They're huge as well. As well as at the TJ Watson Research Center where they have a major demo center now. Mm-hmm. A demon demonstrating the, IBM Quantum system two. now if you go down the list, trap s kind of nest, these are kind of. They're not random orders here. These are in decreasing state of maturity, potentially increasing potential to be disruptive. Tract ions are much more, high, much higher fidelity, which means that they're not gonna be as error prone, but it's hard to scale that's in progress and it's hard to get the performance. They don't have the same performance. As a supercomputing super, conducting circuit does, they, these ions, they're exotic ions. I can't even remember their names. They're not normal elements you could think of. they're trapped in control with lasers and walking around the computer show floor at, let's say TTC last month, you could actually see models of what these look like. Very high fidelity. There's gotta work on performance. And you can see the companies that are listed there, neutral Atoms, kind of a dark horse here, is late, late in development, but is showing tremendous potential. Tremendous. Promise because they operate at room temperature. You don't need these massive refrigeration systems to get them down to sub, sub one degree Kelvin. so that means they're a lot more portable, they're more affordable. They're more flexible. and they're fast. They're not as fast as superconducting qubits or even as trapped ion cu qubits, but they can be as fast when you combine them, because of the, cause. You combine several physical, neutral atom qubits to create a, logical, qubit and get pretty much the same performance. bot is, using light particles and you can see some of the companies there that have been quite vocal about their ability, or again, kind of like neutral atoms, early stage, early days, but very promising technology. The next one is spin qubits or quantum dots. These are electron spin in a semiconductor structure, and you can see the companies they're the most notable are probably Intel. again, this is early stage. It is not something you're gonna, you, or maybe even your drill from will be using, but it does have a lot of potential.

Jim McGregor 11:50

Intel actually started on Supercomputing. Yeah. And they switched and then switched to the, the spin bits. So That's right. It, it did. So a lot of the, you have to realize that this has all been in development stage. So some of these companies have actually evaluated or worked on multiple different, multiple of these different, qubit solutions or modality.

Karl Freund 12:09

Yeah. Quantium is like that. Quantium used to be known as, Honeywell.

Jim McGregor 12:13

They used to be part of

Karl Freund 12:14

They used to be part of Honeywell and they spun'em out and they really decided to focus at that time on trapped ions. But before they were more of an application layer. And, really trying to develop, independent of the modality, they decided to just focus on trapped ions. They think that's the sweet spot.

Leonard Lee 12:29

So, Carl, some of the audience might be wondering, as you go down this list, at what point are you. Departing from superconducting and super cool, like this, near mm-hmm. zero Kelvin cooling. so what does that look like here? Because you mentioned neutral atom is room temperature, right? And so, yeah. what are we looking at in terms of the environment that these different modalities, outside of the supercomputing, operate in?

Jim McGregor 12:58

Well, it's actually, supercomputing and the quantum and kneeling that use the, ultra low temperatures. everything else is either a higher temperature or, at room temperature, especially when you start talking about using photons or ions and stuff like that.

Leonard Lee 13:12

Right, right. Okay.

Karl Freund 13:13

Yeah. Cool. Exactly.

Leonard Lee 13:15

All,

Karl Freund 13:16

quantum and kneeling is kind of a niche. It has commercial users today, and there's really two players that are Dwa and Fujitsu. and then there's topological qubits, which created a tremendous roar in the market when mi, when Microsoft disclosed it. They're the only ones doing this. If they can make it work, which is a big, if, if they can make it work, it could be the winner, but it's very far from production, probably the least developed research platform.

Leonard Lee 13:42

Yeah, but it was pretty amazing at the time, how much, how much buzz it created. But, earlier you mentioned, you said super computing. You meant superconducting,

Karl Freund 13:51

I meant superconducting conducting.

Leonard Lee 13:53

Just wanted to clarify that.

Karl Freund 13:55

Thank you, Leonard.

Leonard Lee 13:55

Yeah.

Karl Freund 13:57

Some pros and cons here. I've probably covered quite a bit of this already.

Leonard Lee 14:01

Yeah,

Karl Freund 14:01

Pros, superconducting, fast gates. Very strong ecosystem, by the way, because they've been around a long time. And IBM's willing to spend a lot of money to, to, facilitate the collaboration of research institutions around the world. it's the near term. Later. Right. big, big, strong, force in the industry. Trapped ions, they're really the contender right now have fault tolerance systems, neutral atoms. They're the dark horse. But there is one company, for example, which is inflection, did a SPAC in February. And so they're now publicly traded and, they're seeing a lot of interest from investors in neutral atoms. and then they also have an existing business, which is called quantum sensing. We're not talking quantum sensing here, we're talking quantum computing. so they can use that to fund and finance the development of quantum computing. FO photonic, very strong quantum networks and distributed quantum computing. quantum dots promising long-term pathway for building these on silicon. On CM OS compatible, chips, topological, potentially a breakthrough. very high risk, very high reward. There's quantum and kneeling. It's commercially viable today. DD Wade's doing quite well, in their niche. but it is a niche. It's not gonna be a general purpose. Quantum computing. With that, I'll turn it over to Jim and you can, dive into your analytics. Jim,

Jim McGregor 15:26

I took a, another, a similar view to what, Carl did and I have to kind of preface this and say that, all these numbers are relative. In other words, things that, they're how I kind of, positioned the different factors of these different technologies. And quite honestly, you have to remember that all these technologies are still developing. So they are changing and they will change. So even looking at the spider radar graph that you're looking at, all these different technologies will change over time. And if you. Click down one slide here. we'll see. When you start looking at the numbers and you start looking at how they're positioned, supercomputing provides that best balance. Trapped on trapped ion provides that, high performance solution or the best performing solution. The photonic and quantum dots, which are very similar, have the best manufacturability capabilities today. although the topological, that was the one that, Microsoft is working on is. Theoretically the best from an error correction standpoint or error mitigation standpoint.

Karl Freund 16:26

Mitigation.

Jim McGregor 16:26

So it will be very interesting. And then once again that a kneeling is really a specialized niche. I don't ever see that necessarily being a general market capability. and, one more slide., I outlined some of those, the same thing that, Carl did, but you have to remember one thing you have to really remember, and let's go to the next slide. Leon. Leonard

Leonard Lee 16:49

d Neil is, is that, is that

Jim McGregor 16:52

my basic Leon? But I, you know, I, I, is that my Scott team? Yeah, exactly. I, no, you have to remember where we are and the fact that, a lot of people said, well, think about where AI was in 2015. Before, everything exploded. We're kind of there in what we call the Nisc era. It's kind of noisy, and unstable qubits where we're really still learning about it. But you have to understand that the industry has actually been working on this since the mid nineties. I mean, this is a technology that's taken. So when, when everyone says, oh, it's gonna be decades away, it's been decades. this has been in development for a long time. we went through a long period of just. Understanding is this even feasible? And then kind of proof of concept era through the early two thousands. And now we're in an era where we understand it a little bit better. we're having to do a lot of error mitigation and suppression just. To be able to get rid of some of that noise that's in these different types of qubits. And obviously there are many different ways to create a qubit, and read a qubit. but we've created quantum utility, and this is probably the most important thing. Why are there billions and billions of dollars going in this, not just in terms of the technology, but also in terms of the use of it? I mean, you have banks, you have materials companies, you have. Scientific research organizations all pouring billions into this because they see this as a very significant leap in terms of computing. This is an accelerator technology. It's not going to displace classical computing, and this is why AI and standard computing and basically the accelerated computing environment we're in now is gonna be so beneficial to this because basically it's gonna help us on the front end. Basically create, the quantum circuits to organize the data and everything else that feeds into this quantum accelerator, and then understand that information and do any error correction as it comes out of that quantum computer. So basically this quantum computer or QPU quantum computing unit, it's gonna be another. Type of rack that sits in the data center or even maybe remotely that's tied to that standard supercomputer or ai, powerhouse. and right now we're finding that, you know, even when you get to like a hundred qubits and that's. physical qubits. You're getting into an area where it's starting to be interesting, where you can understand, how to use the technology, and even run some, interesting simulations. And when you get to a thousand physical qubits, then you're getting into an area where you're creating, a significant number of logical com, qubits, maybe a hundred or 200 logical compute qubits. Creating gates, creating, circuits and actually exceeding that capability, as Carl said, creating quantum advantage where you're exceeding that capability of even what a supercomputer or any supercomputer on the planet can do to do today. Now you have to remember that there's always a probability, risk here in terms of, you know, how close to an actual. Answer it is, but you have to remember that what it's trying to do is beyond what we've been able to do before, analyzing the human genome, modeling the, the earth and the weather patterns, actually discovering new material technologies. it's very, very significant and we're already getting to that quantum utility, being able to use this quantum technology and going forward, we're really looking at this here that we're calling fault tolerant era, where basically we understand the qubits a little bit more. We can control the fidelity of the qubits a little bit more. And once again, this could be in any of those different modalities, but we get to a point where the accuracy is fairly high and we're getting to a point where we can scale. And that's probably the most important thing. So when you see these rapid things, we're basically doubling to quadrupling the number of physical qubits that we can do every year. depending on the different modalities, that's significant because we're gonna be able to go from, thousands of cubits today, or hundreds or to thousands of qubits today to millions of cubits, shortly after the, into the next decade. someplace in between, 20. 32 to 2035. we place it come somewhere around 2034, but that's very significant and once again, as Carl said, once you get into that range and you're only doing anywhere from, it could be a couple physical cubits to a hundred cubits per logical qubit, you're still getting into a point with a million physical qubits to where you can do some very, significant things that you couldn't do anywhere else or any place else.

Karl Freund 21:26

And in fact the, all the companies we've mentioned all have roadmaps that explain how they will get to a million, qubits, which is sort of the milestone for quantum advantage. Probably some areas that can actually get to quantum advantage, far lower numbers of cubits. But in terms of a more general. Platform for doing scientific work and, and research in areas that we don't touch today because today's computer simply can't do it. So it's gonna open up all kinds of areas of computational, applications that do not exist at all today. So it is funny, some people talk about how. You know, there's an article about three or four months ago, some guy wrote, I think it was in Wall Journal, that, computing's all gonna come down to the sides of the shoebox. Right. And it's all gonna be quantum and stuff like this. And I'm like, well, I don't know what this guy's smoking, but I'd like some. It's crazy. This is not gonna replace the traditional.

Leonard Lee 22:24

Yeah. And, and

Jim McGregor 22:26

Carl's address is no.

Leonard Lee 22:30

Yeah, and you know what? I think that is one of the biggest clarifications that needs to be made at the moment because I think there's, so many misconceptions about what a quantum application is and one of the things that was very clear from Quantum Day and I managed to get. You know, listening on some of it, and I'd love to give both of your takes on this, but for the most part, the best way of, thinking about what these applications are the ones that are traditionally, you know, for supercomputing, right? Whether it's for, scientific, very much

Jim McGregor 23:06

so.

Leonard Lee 23:07

Right. Certain,

Jim McGregor 23:08

Doing that analysis, doing the scientific research, doing those types of applications. And quite honestly, I don't know how you feel, Carl, but I think that, it's very much like AI modeling. I think we're gonna end up with different types of quantum modalities for different types of applications. I think we're gonna find as we get farther down the road that, you know, and some of this is definitely AI enabled. On the front end and backend, that we're gonna be able to do, cer, certain modalities are gonna fit certain applications better than others, and.

Leonard Lee 23:36

Yeah.

Jim McGregor 23:37

in looking at this chart specifically, that growth chart, it gets to a point, you have to remember that, this is very much like the AI was, over a decade ago. we're gonna see changes in the industry. So we're probably gonna see a consolidation. Some of these companies will get bought for their IP or bought all out. We may have, larger companies doing multiple modalities by the time we get to this fault tolerant era. So there's going to be a lot of change and that's why everyone really needs to pay attention because, we don't know what the end solution is going to be yet.

Leonard Lee 24:09

Yeah, that's,

Karl Freund 24:09

and are you saying there's gonna be kind of a Cambrian explosion?

Jim McGregor 24:12

Couldn't resist.

Karl Freund 24:13

couldn't resist my bad.

Leonard Lee 24:16

Lots of explosion,

Karl Freund 24:17

self-promotion, my apologies. No, I agree with Jim. I think we'll see specialization start to occur. Right now we're just trying to get to the point where we can demonstrate quantum advantage across a reasonable number of, universal spaces. but as that becomes more general purpose, you'll then see it specialized and you'll see specific quantum modalities. Being used for certain problems and others being used for other problems. So I think there's room for all here to succeed. And, quite honestly, there's plenty of money to be made by all of the players, Most of the players that we're seeing today have a significant research and results that will turn into competitive advantage in specific areas.

Leonard Lee 25:02

Yeah, and

Jim McGregor 25:03

I,

Leonard Lee 25:03

and

Jim McGregor 25:03

there's one other area that, since Leonard, since you brought up the misconceptions, the other misconception we have to kind of address is that quantum is gonna break everything.

Leonard Lee 25:14

Yeah.

Jim McGregor 25:14

It is a, it is definitely a threat because,, traditional security algorithms can be broken by a quantum computer in record times. However, you have to remember that, IBM especially, that's been list, yeah. Kind of leading the charge on these nist. Algorithms that basically to actually, you know, create algorithms that quantum computers can't solve for new security algorithms. They've been working on this for over a decade and they continue, and these new NIST algorithms keep coming out every year, every couple of years. So, quantum computing is not designed to be an encryption, to be a hacker's gold mine. It is something that every company today has to be aware of and has to start thinking about, integrating these NIST algorithms because they need to be quantum safe today. Because there is the theory that, and I completely believe this, you can go and scrape, information today and possibly use quantum computers tomorrow to actually decipher the security algorithms. So that is. A part of the quantum discussion. It's not the part of the quantum discussion.

Leonard Lee 26:17

And I, I think that maybe that's a topic that we talk about, separately because, one of the risks going forward is that, especially around this, quantum era security problem, that. Some of the thinking it is based on, fundamental misconceptions about quantum. And I think, you guys are doing a lot of, good work that will be a basis of clarifying a lot of those misconceptions because I think quantum computing is in general not well understood. They think that, for instance. Quantum computing will displace and replace accelerated computing, right? Traditional computing as you both, cited before. One of the things that I think might be, a good clarification. Education for the audience might be helping them understand what is the role of accelerated computing vis-a-vis quantum computing. Jim, you mentioned, that AI will be important in enabling and, scaling, helping, quantum computing to scale. There was a lot of talk at Quantum Day actually last year about how. AI and accelerated computing, could, enable or be a foundation for these control planes for these different modalities of, quantum, computing runtime. So, how does this all kind of interrelate with each other? Because, you do have Nvidia obviously making big investments. In the quantum ecosystem, but also positioning their technologies for quantum computing enablement. Maybe both of you can take some time to stitch that picture together for the audience.

Jim McGregor 28:08

Well, you have to remember that we're dealing with quantum mechanics, so we're actually dealing with physics and physical systems here in understanding, quantum computing. You have to remember that classical computing has always been a part of that, but also has some manual tasks, like actually trying to calibrate the qubits, actually trying to make sure that they're going. Perform the way you hope they're going to perform, because they're always in a state of flux, and that's a huge, huge challenge. Nvidia and I credit Nvidia for this. Instead of trying to jump into the quantum computing fray, they've said, listen, we're going to enable it. We're gonna be the key enabler here. So they've created their icing, their Nvidia icing models for to help with that calibration on the front end. Basically, hopefully automate that to where you don't have, it's not a manual process, it's an automated process and a reiterative process for each run that you're running, but also on the backend for doing the error correction, which, as we've mentioned before, just getting the right information, ensuring that you have accurate information is a challenge. Today we're using a combination of suppression, mitigation, and error correction today. We hope we get to, and that error correction really isn't. That great at this point in time. So we're hoping we're gonna get to a point where, A, we have more steady states of the quantum, of the qubits, and B so we have higher fidelity, and B, we get to a point where we can actually do quantum error correction. With a very good understanding and very high speeds to be able to do that. So I mean, Nvidia, and this is a perfect application for it, trying, taking these mult multi parallel solutions, basically their GPUs and saying, listen, just like we power AI with this parallelism, we can power these tasks in the front end and back end of quantum computing. I credit them because I think it's going to be, what they're doing is not important for one specific modality or one specific partner. It's gonna be important for the entire industry in any modality that's out there.

Karl Freund 30:05

Yeah, I hats off to Nvidia for really. Putting themselves at the center, the conversation of quantum computing without having a quantum computer. that, that's pretty magical, accomplishment. I do wonder how they're gonna monetize that going forward. but they'll sell enough GPUs. it will definitely help them. But more importantly. It puts them at the center of the ecosystem for most quantum computing because everybody has to air correction. Everybody has to do calibration, q calibration. And so, that's an important first step. I don't think it's the last step.

Leonard Lee 30:38

I'd like both of your reactions to the comment that I'm gonna make right now. the way I view it is, Nvidia is doing a really good job of positioning themselves in practical terms, right? Mm-hmm. Not just fluffy marketing terms as the picks and shovels for, quantum computing. So this hockey stick that you have here, Jim, there, the. The concept is that they will be able to, provide that, control plane, and enable these, execution environments, the design of these execution environments so that quantum computing can scale. The other thing that I also noticed, from conversations and presentations coming out of Quantum Day is that, accelerated computing is great for, discovering applications, right? Because some of these applications are not known. And it's not that you're going to use quantum computing to figure those out. You actually need to have, accelerated computing or AI to support that discovery process. much like what we see with, drug discovery, right? So there's a role there and, your question, Carl, how are the likes of Nvidia? Because I assume that they're gonna be more players than just Nvidia in this game. How will they monetize? Well, they would monetize by supporting these important activities in not only scaling. Quantum computing itself from in terms of progressing all these different modalities, but also the discovery of. the applications, but then also the design of the systems, right? And the circuits. That was the other thing that I took away as well.

Karl Freund 32:26

I think in general, you're right. it was really interesting article. I believe it's Wall Street Journal Suite. But it was talking about how Jenssen's investment strategy is not to pick winners. Don't pick winners. Invest in everybody who has a shot and eventually you'll hit one or two that really. Give you a good return on investment. and I think that's what he's doing right now with Quantum. He's saying, look, I don't know which one's gonna be best. In fact, I don't know which ones are gonna be best at which problems to Jim's point earlier. I'm gonna invest in all of them by giving them the tools they need to accelerate, fault tolerant computing era. Get there sooner, get there with mo with higher quality qubits and lower costs. right now cost is not in the equation, right? It is. Just can you get there? Yeah. But eventually, if you can get there, well then it'll, we'll determine the winner is well, how much it costs. And how manufacturable is it and how high quality is it? How dependable is it? All that stuff, right? all the Illities will come into play, but right now capability, is king. And so he's investing in everybody. Yeah. And expects to be able to make money along the way.

Leonard Lee 33:34

Carl, You're getting soft. I'm out in the open. You can take a shot. Come on.

Jim McGregor 33:39

one of the interesting things is that, with any new technology, there usually comes new business models. We don't know what those business models look like at this point in time. That's going to be, have to evolve over time. And, and like Nvidia, we saw them do an end-to-end solution for ai. We see. IBM doing the same thing for quantum computing. They're the only company I know of that has, well, D-Wave and others. But really, most of the modalities are relying on kind of a ecosystem partnerships to go through there. And I'm not knocking that. I think that's great too, but IBM puts so much in end to end from. The refrigeration solutions to the systems, to the wiring, to, they have invested billions and continue to in the entire thing, and now they've got a partnership with a MD. Yeah. Now, if you don't know, these two companies have worked together for decades. Yeah. On copper interconnects and silicon on insulator technology. They have helped pioneer, these two of these companies have helped pioneer some of the major breakthroughs we've had in semiconductor manufacturing and design over the past three or four decades. So having those two work together, I think is gonna be even, is gonna be a big opportunity. And trying to see what, because once again, now you've got another ai company that's. Fray there. And it'll be interesting to see how everything kind of meshes together as we get farther on down the road. and it's funny'cause some of the quantum company, quantum computer companies are promoting on-prem systems. I don't know if that's going to be the preferred business model. I still think that for the vast majority, cloud-based, kind of rental solutions are probably going to be the most likely scenario for quantum computing, especially for the next decade.

Karl Freund 35:21

I think the exception to that's probably in two areas. One is in financial services and the other is, government. you look at a lot of these companies right now, they're, they have two primary. Markets, customer segments. One, one is, is researchers and the other is national defense. Mm-hmm. and so national def both those areas, I think there is a role for on-prem quantum computing, even though it may be an internal self-hosted cloud. so kind of a hybrid model.

Jim McGregor 35:52

Yeah. And you, it's important to note that, there's obviously a lot of investment from the research community too, matter of fact. Mm-hmm. IBM has been putting some of their systems out there in the research community, like, uh, believe the University of Cleveland or something like that, or mm-hmm. They've gotta be universities and research organizations around the world now. So it's gonna be interesting to see how this all plays out. And matter of fact, two of the companies are spin outs from Oxford University. So, it's, I I think we're probably, maybe not to the point where we saw transformer models and we saw LLMs with, ai, but we're getting there very quickly and by the end of the decade it's gonna get exciting.

Leonard Lee 36:30

Yeah. Sounds a lot like super computing.

Jim McGregor 36:34

It's

Leonard Lee 36:34

how it's,

Jim McGregor 36:35

super computing on steroids.

Leonard Lee 36:36

Steroids, yeah. It's quantum computing. So, Carl, do you think we're gonna have quantum factories.

Karl Freund 36:43

Quantum factories.

Leonard Lee 36:44

It sounds like Jim described, quantum infrastructure. Right. The whole supply chain and the build out large, data centers and or quantum centers.

Karl Freund 36:52

yeah. I think there will be quantum factories, because the cost of these systems is still quite high.

Leonard Lee 36:59

Yeah.

Karl Freund 36:59

And so you'd amortize those costs over as many. Use cases and customers as as you can. And so building quantum factories where you have CPUs, GPUs, xus, and Q all basically providing services for applications we can barely imagine today.

Leonard Lee 37:20

This has been great. This has really been, informative. I know I've learned a lot from both of you in this 30 minutes or so that we've had this chat. And, it's great to have an opportunity also to talk about Quantum Day. I, I thought, that it was really well done program.

Jim McGregor 37:38

even some of the detail they went into Was incredible. I,

Leonard Lee 37:42

yeah,

Jim McGregor 37:42

I need to go back and watch it probably a dozen more times because some of the mathematics, especially in the last presentation,

Leonard Lee 37:48

yeah. I was in transit. I was trying to listen to this stuff as I was on a flight. So,

Jim McGregor 37:53

yeah, the professor from Israel was just, oh my god. He, his mathematics depth was incredible.

Leonard Lee 38:00

So, gentlemen, Thank you so much. It was really our pleasure. Really, really great, great stuff. And, I think the, audience will undoubtedly find it incredibly grounding and, a good orientation as to how they should, view the aqua, the Cambrian explosion. That is quantum explosion. The what?

Jim McGregor 38:21

The quantum explosion.

Leonard Lee 38:23

The quantum explosion. Nah, I gotta make it Cambrian.

Karl Freund 38:27

Cambrian.

Leonard Lee 38:29

Cambrian. Cambrian Quantum explosion.

Karl Freund 38:32

I've got

Leonard Lee 38:32

all these qu

Karl Freund 38:33

the qc,

Leonard Lee 38:34

I've got all

Karl Freund 38:35

these Cambrian fossils. I can't change'em all to quantum fossils.

Leonard Lee 38:43

So, thank you gentlemen and everyone, if you're interested in tapping into, Jim's research calls research, reach out to both of these gentlemen and their respective firms. Of course, Jim McGregor is with tedious research. How's that? It's

Jim McGregor 39:03

pretty good.

Leonard Lee 39:03

Pretty good. Great. Yeah. I mean, sad. Hey, I'm an honorary Scotsman now because my nickname, or at least my Scottish name is, Liam, right?

Jim McGregor 39:12

Yes, exactly.

Leonard Lee 39:14

I gotta live up to that honorary designation. So yes, reach out to him at cheer. research@www.teusresearch.com. He's also on Forbes and he's quite active He Times and other publications. And then of course we have Carl Fre of Cambrian Quantum AI Research.

Jim McGregor 39:37

Put Hy keep putting hyphens on the company name.

Karl Freund 39:39

Cambrian AI hyphen Quantum hyphen Research.

Leonard Lee 39:43

So catch him at www.cambrianaium.com. Did I get that right?

Karl Freund 39:51

That's right.

Leonard Lee 39:52

Yeah. And then, he has a substack. He has, a, he's on Forbes and he's on LinkedIn. He's also public, well published, across a number of publications and outlets. And of course, please subscribe to our podcast here. We do a, A, an episode every month to cover everything that's happening that you should know about, in the world of. AI, semiconductors and quantum. So, make sure to check us out on YouTube, subscribe, like, share, and then check out the audio version on Buzz Sprouts and listen to us on your favorite podcast platform and subscribe to the next curve research portal@wwwnextcurve.com for the tech and industry insights that matter. Gentlemen, thank you so much. This is awesome.

Jim McGregor 40:46

Thank you.

Leonard Lee 40:47

We will see you by the end of the month.

Karl Freund 40:49

talk to you soon.

Jim McGregor 40:49

Yep.

Leonard Lee 40:50

Yeah. Cheers.

Jim McGregor 40:52

Cheers.