The Exchanges

Q Um, you mentioned that this is new and different. I think we should put a finer point on it. What, what is it that is distinct about this approach versus historical wave energy approaches? Like, what makes this unique?

A Yeah, great question. So we, the, the most distinctive thing first, I think, is that we are not doing it In coastal areas. And, and this was marine energy For the most part, for all of history before we decided to pursue this approach. Um, there were patents, scattered patents in the distant past of people thinking about doing something like this, but most coastal, most marine energy has been coastal because of course the idea that most people have had is you want to get the energy back on a cable. And if you're going to run a cable, you better be close to shore. Um, and also many of the, at least wave energy technologies, But also wind energy technologies have all relied on a seafloor connection of some kind, either just to moor the system and keep it in place, keep it from drifting off, or in many cases to actually push or pull against to create the reaction forces that you need to drive your generator or power takeoff or something like that. And so there's been this historical center of gravity to do it close to shore. We decided, um, we decided to cut the cable. And go to the middle of the ocean, and that's for many reasons. It's number one, because that's where all the energy is, frankly. Um, if, if you look at the entire globe, and you say, I'm gonna look at it for how much energy is there in the wind, in the waves, um, the coastal regions are very small in terms of both …

AI assessment note: “the most distinctive thing first, I think, is that we are not doing it In coastal areas.”

Q Um, okay. Okay, so onto then, I think, like, maybe the obvious really big question, which is O&M. Um, you know, most, both energy generation equipment and, uh, data centers require a fair bit of maintenance. How does one do maintenance in the middle of the ocean?

A Yeah. So we, let's talk about the nodes and then the servers as well. So the nodes, this has been our philosophy from the beginning. We wanted to design something that really doesn't require maintenance during operation. Um, and how do we get that? We get that by having the hull be just solid state, completely just steel, marine coatings, and that's it. And then you have your one water turbine. Which is fed by the, uh, fed by the reservoir, and that is also just a very simple rotary moving part, spins on bearings. You can design that for whatever lifetime you choose, you know, five years, 10 years, 15 years, and those are the only moving parts on the whole system. The hull is solid state, it moves up and down in the waves, and the turbine spins inside. None of that should require maintenance, at least not on the time frames that are relevant. So, Um, it remains to be seen whether we achieve that design goal, but we don't know what about that breaks, because we've run our systems at sea. They, of course, survive. We've run our turbines and endurance testing, and they survive. And those are the elements of the system that you need to survive in order for the systems to last a long time.

AI assessment note: “We wanted to design something that really doesn't require maintenance during operation.”

Q power plant design. Um, and so as a result, the folks Like yourselves and others who are saying, okay, we're going to get net facility gain in the next few years are generally not doing the same thing that INL did for that system. Why is that system not practical? And like, what is, what is the fundamental, what needs to be true to imagine that a design is practical?

A There are a lot of things that are important to consider here. So part one of practical might be what needs to be true to get more energy out of the entire fusion machine than was stored in the system. So what needs to be true to get in that facility gain? The second half of that question, which is equally important, I would argue, is what needs to be true to have a fusion power system that can scale, as in be maintainable, deployable, and affordable, such that it's a competitive source of power, right? Energy is a commodity market. At the end of the day, this whole thing is a race to the bottom on cost. So to start with the first, there are two things to consider. The first for lots of different fusion approaches is do you build on an established path to ignition and high gain, right? Is there experimental evidence that says the physics of what you're doing is going to work? Um, and then the second piece is, is your system efficient enough to do that in such a way that you get more energy out of the whole machine than, than was stored in the system? And so for example, The national ignition facility at Livermore was not designed to be a power system. It was designed for a different mission, and so it wasn't designed to get more fusion energy out of the whole system than was required to drive it. It has a relatively inefficient driver technology. What we do at Pacific Fusion is…

AI assessment note: “The national ignition facility at Livermore was not designed to be a power system.”

Q is believed that the path to one point oh one, uh, is way, way, way harder than the path from one to five. Is that your general view on it? Like, you get over this mountain that is net facility gain, and then after that you've just got, like, some version of engineering tweaks to make to get to five, or is that the wrong way to think about it?

A Yeah, I think we have to be careful, right? Nothing in fusion is ever easy, and I think trivializing the work required to make those improvements is not a good idea, but one way to think about, one way to think about what's often described as the ignition cliff is that to get your fusion fuel to generate more energy than is driven into the fuel, you have to get it to start self-propagating burn. And once you do that, it's easier to add a little more energy and get a lot more energy out. So there's lots of published literature on this. You can see different curves of the ignition cliff where you have for a very significant amount of current, for instance, in the case of a pulse system, you add a little more current, you don't get much more energy out, a little more current, you don't get much more energy out, a little more current, you don't get much more energy out. Then you hit this ignition cliff and suddenly a little bit more energy onto the target starts getting you way more energy out of the target. Um, so that, that is why, right, there's a lot of There's been a lot of focus in the field on finding that ignition cliff, and now on building systems that can drive targets up that ignition cliff, and obviously design all the supporting systems to be able to work in those kinds of yield regimes.

AI assessment note: “once you do that, it's easier to add a little more energy and get”

Q but is it true to a first order that, um, basically every additional point of, of gain that you can get translates to, like, Pretty linearly higher, essentially what would be higher efficiency in another type of system, which is, which is lower LCOE. Like, uh, should a, an equal amount of focus be placed on every additional point of gain as it is on every additional reduction of capex?

A Yes, with some footnotes. So one way to think about it is if you treat, imagine a system and you treat it as fixed, like a fixed capital cost of the system, the nameplate capacity of that system, so the amount of power that it produces, is going to be determined in a pulsed inertial system like what we're building by the gain, so the amount of energy you get per shot, and the rep rate, the number of fusion shots that you take per unit time. And so, if you imagine building infrastructure, right, building fusion power systems, you can imagine building a fleet that starts with the ability to produce some amount of power per unit system at some nameplate capacity, and then with improvements over time, so upgrades to the fusion chamber to accommodate a higher rep rate, and upgrades to those fusion targets as we iterate on them on the demonstration system, that that same physical thing can produce more power and as a result have a lower effective LCOE.

AI assessment note: “Yes, with some footnotes. So one way to think about it is”

Q We just still have to do a first of a kind that has the net gain of five X and right. There's like another version of a, I guess there is a first of a kind. The demonstration is a demo. The next is first of a kind. Is that the way to think about it?

A It's a good way to think about it. I think people tend to, if you talk to people and ask them, define a demo system, define a pilot system, define a first-of-a-kind system, everybody needs something different. Um, but what I can talk about is sort of what our demo system can do from a capability perspective and how we think about what will be new on a first power system and then what it looks like to scale from there. So the demo system is designed in many ways to demonstrate a facility gain and also importantly to iterate on these fusion targets. So we have a lot of diagnostics on the demo system. We've designed it such that we can test prototype components of commercial systems on it, a commercial chamber on it. So take shorter runs at higher rep rates, for example. And as a result, it serves as a really important de-risking platform for a lot of the core technologies, including the targets and a platform upon which we can continue to iterate on targets that are relevant for commercial systems. On a power system, there's a really big difference between a power system and the demo system from a technical perspective. There are many, but one important one to understand is that the power system will be rep rated at about a hertz, so once a second, whereas the demo system will be rep rated at a shot a day. And when you think about a lot of the capital components like the Pulsar a…

AI assessment note: “It's a good way to think about it.”

Q when you think about the, the supply chain, obviously this is going to vary depending on which of the approaches are We talk about some of the approaches require high temperature superconducting magnets, which has its own pretty limited supply chain. Um, but I'm curious what you think of as like the potential, let's say we do scale up a fusion power generation market, where might supply chain bottlenecks emerge?

A I think you said it well when you said it's really dependent on the technology. So something like high temperature superconductors are a relevant, a relevant piece of the puzzle for steady state approaches like tokamaks instillerators. They're not relevant for inertial approaches like what we do. Our objective in designing our systems and one of the reasons that we founded this company is because we see a path to avoid reliance on what I would call broadly speaking specialized materials. So as I mentioned, a lot of the system is built from oil, plastic, metal, and water. We'll have the same procurement work to deal with as any other energy system. We think about balance of plant and other such components, but there's nothing inherently rare or expensive, meaning that the bottleneck to scale up ends up being effectively a manufacturing bottleneck. So for a lot of these core components, you can procure them in small volumes today, maybe not quite at the performance and lifetime requirements you need, but there's not, there's, you know, a lot of the vendors, Do these in small volumes. As a result, the multiple on the bomb is relatively high, and so the challenge is in building capacity, both in the supplier base and internally, to deliver those components at the scales and at the performance requirements needed.

AI assessment note: “the bottleneck to scale up ends up being effectively a manufacturing bottleneck”

Q What about fuel? Talk to me about the fuel supply chain.

A Yeah, fusion fuel obviously is very different from fission fuel. There's no uranium, no plutonium, nothing that you can use to make a weapon. For fusion systems, the fuel is deuterium, so it's an isotope of hydrogen found commonly in seawater and tritium, um, and what that means in practice is that it's deuterium and lithium because every fusion approach out there needs to breed its own tritium to be economical, so make tritium from fusion events using deuterium and lithium. That stuff is readily available. Tritium is important to consider, though, because every fusion approach to be commercially viable needs a sustainable tritium economy. So that means you need your startup volume of tritium to be relatively small, and then you need your operating fusion power system to produce more tritium than it consumes, such that you can start additional systems and cold start your system when needed. And there are a number of people working on this in the field, it's a challenge, it's common to the whole field, but there are differences in, across different fusion approaches around the volume of tritium required to start up, And then how rapidly that tritium can be bred, or how efficiently that tritium can be bred with the system that exists.

AI assessment note: “every fusion approach out there needs to breed its own tritium to be economical”

Q Okay, so how much of, I mean, I don't know, give me a sense of the magnitude of the shock that has created. How much of global supply then is just, has, you know, vanished while the Strait of Hormuz is closed?

A With that body of water closed, just in itself, I think we're talking about a third of the world's tradable urea, gone. And a good chunk of the global anhydrous, I mean, the same thing. These same plants that produce urea also produce a tremendous amount of anhydrous. But that's not even the worst part of it, right? Then all of a sudden you got to think about, well, what else comes out of that body of water? You've got LNG shipments that are stuck behind there, and that feeds major manufacturers like India. I mean, to give you a scale of it, Europe, we're talking about, you know, their 75% production rate, that's three and a half million tons missing. India is very reliant on those LNG exports from the Persian Gulf. That's how they feed their nitrogen plants. And just to give you a scope of size, they produce over thirty million tons of urea every single year. And at one point at their worst spot, their production rates were down to 50 or 60% of normal because of high prices, because of the lack of the input. So that's on an annualized basis, that's fifteen million tons per year that was missing. That's the equivalent of all of Europe. So that's where this thing is just completely blown out, right? It's, it's no longer just the individual products. Now we're talking about the inputs and for phosphate, it's the lack of sulfur and the lack of anhydrous and the high price of those…

AI assessment note: “I think we're talking about a third of the world's tradable urea, gone.”

Q little bit more resilient because we're a net exporter in that space than you see in other countries. So give me a sense, I mean, you said we were, we've experienced a version of supply tightness that we have never seen before in the history of these markets. How has that manifested? Like, give me, give me an example of pricing or Shortages. Like, what has happened as a result?

A The best thing to show as far as talk about that's actually living proof of it is India. Now, India is a little bit of a different beast. A lot of the world farmers have to ebb and flow their demand based on what prices are because their price is a direct correlation reflection of what the world market is. Indian farmers, they don't care what's going on around the world. Their price stays low and steady, and that's dictated by the government, and what the government does is they subsidize the price difference between that farmer and And the rest of the world. So their demand doesn't move because they just don't care. They paid one of the highest prices we've seen in a very, very long time here recently to buy two and a half million tons. Over 900 dollars a ton. I think the West Coast prices were 935 dollars a ton. East Coast prices were nine 50 plus. An incredibly high price, but they were still able to find the product. What we've been watching very, very closely is countries like Australia. Now, again, I'm not going to say it I was about ready to say fortunately for their drought, because that's not the right way to say it, right? Their drought has been terrible, but from a strict fertilizer standpoint, their demand has been much lower because the farmer is saying, I'm not getting rain, I'm not going to invest in the fertilizer, right? If that rain were to come, they are goin…

AI assessment note: “They paid one of the highest prices we've seen... Over 900 dollars a ton.”

Q So just to clarify, historically, China was the number one phosphate exporter, the number two urea exporter, and then they basically stopped exports. And as of now, they've stated they're going to start exports again this summer, but TBD if that actually happened. So that's thing one that was already contributing to global supply tightness. Then there's Europe and the Ukraine-Russia situation. What impact did that have?

A This goes all the way back to 2021, and during that time, that is when you saw Europe basically sit there and tell Russia, we're not interested in any flows from your Nordstrom II pipeline that you just spent all this money building. And they got into a TIF, and eventually Russia shut off the flows, and then somebody, I still don't think we actually know who did it, plenty of theories, but we don't know, actually blew up the pipelines under the water. Well, from Russia's standpoint, why would I go repair them? Europe doesn't want the gas. I'm not going to go spend the time and the money and the effort to repair those pipelines. European gas values skyrocketed, right? That Dutch GTF went from whatever the normal was, three, four, five dollars in MMBTU to, I think it was August, 20, 21, a 103 dollars in MMBTU. Now, since then, we've normalized the market, and they're finally back to, like, 1015 is kind of their normal range, but that means their production, their nitrogen production in Europe is only sitting about 75% of normal, and that's for all of Europe, and when you start to break that down, That's about three and a half million tons of urea that's not being produced a year. That's a couple million tons of uan not being produced a year. And that hurts. Their farmers aren't going to just sit there and say, well, darn, I guess if we're not producing it, I just won't use it. Th…

AI assessment note: “That's about three and a half million tons of urea that's not being produced”

Q And so where are we today in terms of autonomous trucking? I mean, you said Otto and Starsky were the first two. There have been a bunch of other, I guess, what I would call serious attempts since then. Like, how far have we gotten?

A Yeah, many serious attempts, I would say. And by serious, I mean real technology development, real capital, real, um, real, real efforts, uh, They've all kind of followed a similar blueprint, I would say. They take existing tractors that are, that, that have been manufactured by OEM, sometimes in partnership with an OEM like Volvo, uh, sometimes, um, just sort of buying from a lot and, and, and doing a quick retrofit. Um, and, uh, you know, the attempts so far, you know, there have been a few driverless runs where, and by driverless, I mean, nobody in the front seat. There's a lot of, like, sort of debate about what driverless even is. Right. Uh, for example, I think Aurora, which is one of the largest players in this space, you know, they do what they call driverless runs, but they, they sometimes have a safety observer. This is from their public, public writings. Um, somebody sort of, sort of watching the system, but they, they call it driverless, as in maybe they don't touch the wheel or engage in any way. Um, so there have been driverless runs, even going back to that 2016, uh, auto. Um, there was a company, there's a company in Bot Auto that just, I think yesterday did a, Commercial, what they call the first commercial driverless run. Um, I saw that on LinkedIn. Um, but I would say there's not a regular driverless service for trucking on highway that exists today. Um, and …

AI assessment note: “there's not a regular driverless service for trucking on highway that exists today.”

Q camera for autonomy are more around, like, it can get obscured in certain conditions and things like that, and that's where you are Want your LIDAR or whatever, which, which doesn't have the same set of issues. Is your view, like, you can go camera only as a result of the VLMs, or is it you can lean more heavily on camera, but you still want this sensor fusion approach?

A I think, especially for trucks and, you know, given what I was talking about earlier with this 8000 pound vehicle that you might be moving on the road, you want it to be as safe as it can possibly be. And so, you know, in my mind, it's not a It's not a dogmatic debate about camera or lidar. It's what's the best technology for the moment that makes it the safest. And for, from my perspective, for a vehicle, a truck, for example, you put on the road, you would want it to have camera, lidar, and radar to do a level four, level four being, you know, driverless, to do a level four truck today. Um, and the reason, and the reason I say that is because you want to be able to see the world in multiple ways. Um, it's, but There's always a sensor doing the most heavy lifting. There's not sort of this, like, co-equal, like, democracy between the three sensors. It's usually, like, some, some sort of priority is put on some algorithms, depending on what you're doing. Um, LiDAR doesn't see traffic lights, right, very well, or it doesn't see the red-green, so you would use camera primarily for that. Uh, LiDAR can see better at night. LiDAR can see through weather in some cases. Um, and so if you want to make a safe product, And trucking in particular, you want to really take advantage of all those today. A human, and when you think about the end state, you know, a human is, is, is effectively …

AI assessment note: “you would want it to have camera, lidar, and radar to do a level four”

Q Right, because part of the market is, like, the tractor and the trailer are owned by different entities, as it stands today, right? Some, some, it's not the entire market. Sometimes it's the same player, but, you know, they get separated sometimes, and sent off in different directions.

A They get separated, and sometimes for very good reasons, like, you know, uh, trailers are relatively inexpensive, you know, tractors are relatively expensive, so there, there's a lot of, like, there's a lot of, kind of, interesting thinking around, like, trying to combine this concept, um, But just from the technology side, for example, if you've made the tractor smart, like, like, like the other autonomous truck players, and you've left the trailer, uh, conventional or like a dumb trailer, quote unquote, right? Um, you can't, for example, put sensors on the back of the trailer. Like, you have a smart tractor, but you're just taking trailers from everywhere, so you can't see behind you, ok? Uh, you can't see directly behind you. If you can't see directly behind you, you cannot Handle, for example, an accident where somebody just drives right into you. And I think that's, there's maybe some clever ways to do that, but that's, that's tricky. You cannot back into a dock, right? Because that requires some amount of understanding of what's going on behind you. Uh, one, one interesting, this is just like a kind of like an inside baseball thing, but one interesting challenge we've had in the industry is that trucks today, oh, if they're, if they're pulled over, they're required by law to deploy warning triangles, and, uh, those warning triangles go behind the vehicle. If you only have…

AI assessment note: “They get separated, and sometimes for very good reasons”

Q two, is there any argument for that changing in the future? Because that reliability requirement causes so much challenge and capex, right? Why is it such a problem that we have lead times on, on gas generators, all this kind of stuff? It is because of the reliability requirement. So is it intrinsic to something about what you're doing, or is it just a function of how the Businesses evolved.

A Yeah, a fantastic question. And I think that if I were to probably, um, send one message here is no, it is not intrinsic and we should be thinking about lower reliability, uh, power delivery, uh, overall. I'll tell you, uh, why it has been, but I think that, um, I'll also get to why it has changed substantially. So for most modern software services, the compute is actually a relatively small fraction of your costs. And so now it makes sense to, um, over-provision it. You want to have 99 point actually nine nine nine percent, five nines reliability for your software services. You don't need quite that, but many of our data centers aim for four nines of, I mean, minutes of downtime a year maximum, which as you said, uh, has a large amount of costs associated with it. Now, if you think about it though, as of now, given how constrained resources are, And how costly they are, a much larger fraction of your overall service cost is in the compute. So if you went to your internal customers, if I were to go to my internal customers and said, would you rather have four nines of availability and half the capacity, or two nines of availability and twice the capacity, which do you pick? Very often, not always, very often they'll say, oh my gosh, give me two extra capacity. And if I need to have a nine, nine percent, a nine, nine percent sounds good. You all know the math. That's 3.65 days o…

AI assessment note: “no, it is not intrinsic and we should be thinking about lower reliability”

Q you enroll in a program, we, and as that enrollment, you know, we are going to send you a signal, or we're going to call you, as it's been historically, uh, a couple times a year at peak hours, and we're going to ask you to ramp down. This is kind of an extension of Demand response, right? Like, what do you think of as being the same and different?

A I, I completely agree that, first of all, there are, ah, there are natural pricing tiers. So if you're in ERCOT in Texas, you can pay more money for power right now, or you can ramp down and pay less money because power prices are high. Then there are demand response programs, as you mentioned. A utility might say, you can make some money if you agree to curtail during this period. You might own a Nest thermostat and be enrolled in a smart thermostat program, and they will pay you if you are willing to reduce your consumption. There are even mandatory programs where a prerequisite of your enrollment is your promise to curtail and a hefty penalty if you do not curtail. But what's missing from all of these is the ability to offer the service of curtailment at the large scale that data centers could theoretically provide it and get a real benefit out of it. And that benefit is not just a cheaper cost of power or a flexibility payment. That benefit is a larger power connection or faster access To the power grid, whether you're an existing data center seeking to increase your capacity or brand new data center seeking to connect, you should be allowed to harvest the existing stranded capacity on the grid that can be served to you if you are willing to curtail every so often. And that, that product doesn't exist today. And with good reason, you know, the electric utility industry for …

AI assessment note: “But what's missing from all of these is the ability to offer the service”

Q Somebody else might be running the workloads or actually being the customer. So can you walk me through, like, how you think about the stack of who needs to do what? If we're going to deliver on this promise and we're going to take advantage of the hundred gigawatts of latent capacity we've got on the grid by making data centers flexible, like who needs to sign off on what?

A Yeah, it is a wickedly complicated multi-party problem. I'm, I'm delighted, Shail. You got comfortable getting your hands around this, and now we can work together. Look, I go back to an earlier question you asked, which is, what's the most critical thing that has to happen? The most critical thing that has to happen is, Power utilities and system operators and regulators and governors saying, if you are willing to be power flexible, oh data center, we want you in our state, you get to skip the line, you get to connect faster as a flexible load fast track, you get a bigger data center, etc. If that happens, I believe everything else quickly falls in line. Now, you're right. The data center is not one monolithic entity. It comprises a lot of players. You might have, for example, a data center developer, Owner and Operator. I'll make one up. Digital Realty, a terrific one that we partner with, that is operating a data center within which they have a tenant, right? That tenant might be one of the many folks we've partnered with, like Nebius, for example, or Oracle, or Lambda. And within that cloud provider, by the way, it could be a hyperscaler as well. Within that cloud provider, you might then have a customer, And that customer, by the way, may not be the end customer. You might have together.ai or fireworks.ai, which is an inference serving service, which is then serving tokens…

AI assessment note: “You might have, for example, a data center developer, Owner and Operator.”

Q Okay, so back to the supply chain then. So we, we mine our uranium, maybe in the U.S., but probably in Canada or in Kazakhstan. We convert it, again, maybe in the U.S. through one facility, but more likely again in Canada or in Europe. Where then today does the enrichment typically take place?

A Yeah, so this is, this is what really put us onto this, onto this problem and, and deciding, hey, we need to really start a company to address enrichment in the U.S., was if you look at enrichment today and what the U.S. consumes, It's about 75% Europe, European producers, um, and it's about 25% Russia. And so we can talk about the history of, of how we got here, but there's no commercial at scale, uh, U.S. producer operating, um, anywhere. There is one facility in the U.S. that's run by a European firm called Urenco down in New Mexico, and that produces about 20% of U.S. demand. But the other 80% is coming from overseas, and a full 20, 25% is Russia, depending on the year.

AI assessment note: “It's about 75% Europe, European producers, um, and it's about 25% Russia.”

Q to. My understanding from chatting with you is it's not actually quite that simple, and, and it is kind of a different process, or at least you want different equipment if you're going to be producing LEU versus HALU. So at the high level, can you just walk through, like, are those the same process run at different frequencies or for different lengths, or is it actually a different process?

A Um, I think the, the thing to remember is this is all really just a separation distillation process, and so Producing HALU, you typically ingest LEU, and then you'll enrich that up to HALU. And so it's, it's really a repeated process. Now, the things that are different, the important things that are different are, um, around criticality and licensing. And, you know, the licensing is different to reflect the criticality difference and a few other differences. Um, but fundamentally the process does not have to change from a physics standpoint. Um, what does have to change is things like you mentioned, um, things that hold uranium, um, a certain volume of uranium may need to be smaller in the case of HALU to make sure that you can't have accidental criticality and that you're ensuring safety. But I would say that that's the primary, primary difference between the two is, is criticality considerations, which is, which is why you see, um, you know, in the past couple of years, the DOE putting out Awards for, uh, HALU enrichment specifically, um, and LEU enrichment capability, and then also HALU deconversion. And so that last step that you asked about, those last two steps of deconversion and fuel fabrication, as you bring, uh, HALU UF-Six down into solid form, uh, you're now getting even more density of, of uranium. And so as it's at a HALU level, um, You know, up to 19.75% U-Tur-th…

AI assessment note: “fundamentally the process does not have to change from a physics standpoint.”

Q my favorite questions to ask. If you could, um, if you could wave a magic wand and solve some problem in the nuclear supply chain that isn't the one you're currently solving, something other than enrichment capacity in the U.S., what would you solve? It's another way of asking the question, like, what do you view as the biggest bottleneck besides the one you're going to try to go tackle?

A Yeah, I think we already talked about conversion. I think, I think as enrichment in the U.S. gets scaled up, You know, five to 10 times. You're gonna need more conversion capacity. I think people are working on that. I think that will get solved. Then you look at the next bottleneck, um, of US mining. And if you look at, you know, ideally we have that in the US too at scale, at a scale that meets all of our needs. Um, that would be, that would be really where I would wave the magic wand. Um, you know, does the US have as good of deposits as some other countries? No, it doesn't. Um, but today For us product to be mined, you know, it's shipped all the way out of the country to be converted in, in a lot of cases, and then shipped all the way back. I think we should have a full domestic supply chain. And so, um, you know, you talk to us mining companies and, and a lot of the challenges are just around things like mining permits and, uh, how long that takes. And so I think if we can see, uh, rationalization of those processes, Especially given, you know, things are moving to ISR, uh, and are much more, uh, you know, lower impact to the environment. I think if, if regulations can begin to reflect the reality of what mining is today, uh, and, and make that more streamlined and allow for U.S. mining to come back, I think that'll be a great thing for the U.S. supply chain, because at th…

AI assessment note: “Then you look at the next bottleneck, um, of US mining.”

Q Yeah, so let's get straight to the, the geopolitics, I guess, or at least the geography of it. So as it stands today, what's a typical supply chain look like, starting from mining through to ultimately usage in a, in a reactor?

A Yeah, so with mining, um, U.S. gets mined product from a bunch of sources, including from mines in the U.S., but Canada is a, is a very large producer, and Kazakhstan's a large producer. And, um, Australia also has great deposits, but if, if you look at, at today, it's really Kazakhstan, Canada are going to drive it for the U.S. That's where we get most of the U-Thia weight that we, we consume. Um, conversion is, is also international. We have one facility doing conversion in the U.S. Um, that's Honeywell, uh, sold under Converdine in Southern Illinois. It's actually, um, five miles from where our facility is. So, uh, that's in Metropolis, Illinois. You also have conversion, uh, in Canada, outside Toronto, uh, done by Cameco. And then you've got, uh, the Europeans who also do conversion. And so.

AI assessment note: “with mining, um, U.S. gets mined product from a bunch of sources”

Q Partially because there aren't, those advanced reactors don't mostly exist yet, or at least they're not commercial in the market, but Halo, there is, uh, zero current capacity, or essentially zero. I mean, walk me through, like, if I wanted to go buy Halo tomorrow, what would that look like?

A Uh, you would have to purchase it from Russia, and so that's what actually triggered me looking into the space. So, if we rewind to, like, late At Founders Fund, I was, you know, looking at all the advanced reactors companies, deciding to invest in one. Um, I asked them what the hardest thing about building their company was going to be, and it was purchasing fuel, it was obtaining HALU. And they said, the only place we can get it is actually Russia, and we have to import it. And so I said, well, why don't, why don't you just get the U.S. companies to, to make HALU? Is it that much harder to go to a higher level? And they said, there, there really is no U.S. owned production. Um, and so that kicked this all off, and I, I realized pretty quickly that, um, that Russia was the only source, and unless there was a new source that came online very soon, really by end of decade, um, all the advanced reactor companies would, would have a hard time scaling up. And so, you know, fast forward to today, the DOE has actually stepped up and made some HALO available to advanced reactors. Um, but that's, that's really only going to take them through first Demonstrations first deployments to really scale up will need a new supply. And so, um, you have Europe saying that they're going to bring capability online in Europe, um, in the early 20 thirties. And then, uh, the other two companies saying…

AI assessment note: “you would have to purchase it from Russia”

Q Russia, for that matter, um, then, yeah, like, great for the government to step in. So, okay, so you got this big DOE award to, to go build that. Like, just talk to me about what that, first of all, talk to me about the Paducah site, because it's interesting. Um, And then what do you, what's it going to look like? What are you actually going to build there?

A Yeah, so Paducah is actually the last place the U.S. did commercial scale enrichment. Um, it's where we did enrichment that fueled all the U.S. reactors, and that facility was, was shut down in 2013, and so the Paducah community, uh, Paducah western Kentucky, the very western tip of Kentucky, that community remembers when the enrichment plant was operating, and they're very comfortable with enrichment, they understand it, they're comfortable with nuclear, um, and so as we looked around for basically a year, In over 10 other states, something like a thousand different pieces of land, um, we found Paducah to be the most supportive, the most excited about bringing enrichment capacity back. So our site in Paducah is on the DOE site. That's the site where the enrichment was performed previously. It was called the Paducah gaseous diffusion plant, and we have about a hundred acres at the south end of that site. That we've leased, um, for a long period of time that we will build our facility on. And so a hundred acres and, and, you know, again, building enough capacity there to satisfy HALU through the next decade, and then enough LEU capacity to displace, um, adversarial imports into the US. And so that's the scale of it. That's where we're doing it. Timeline is, you know, Before end of decade. Um, and then, and then, yeah, that's the rationale on why Paducah, but it's an incredible, …

AI assessment note: “we have about a hundred acres at the south end of that site”

Q Let's talk about nuclear waste. First of all, what is it, actually? Like, what, what is the waste that we care about?

A Right, so, so nuclear waste is basically any commercial waste that the U.S. has is uranium dioxide, basically, uh, ceramic is the way that I describe it, what's made out of your, your coffee, coffee mug there, and it has many things from the periodic table, and it's a largely second-row transition metals, early second-row transition metals, As well as lanthanides, as well as basically some actinides heavier than uranium. And what this means is that the uranium composition, as far as the actual irradiated meat, ends up being about 95% of the material, and then the other five percent are these fission products that grow in over the course of the irradiation, as well as the, what I call the, the transmutation products, basically, when the uranium gets struck by By a neutron. Captures that neutron instead of fissioning, and that grows into your plutonium, your neptunium, your americium, all these sorts of pieces, right? So that's largely what it's comprised of. It stays in the fuel. It's a solid. Some people are surprised to learn that it's not a liquid goo. I've just been watching Teenage Mutant Ninja Turtles with the kids recently, so that's not what it is.

AI assessment note: “nuclear waste is basically any commercial waste that the U.S. has is uranium dioxide”

Q so, so 95% of it is, as I understand it, 95% of it is the stuff we don't worry so much about, but we worry a lot about that five percent, right? And that, that five percent is comprised of, at least in part, a bunch of these isotopes that have various degrees of radioactivity and various half-lives. Can you just kind of, like, break it down a little bit?

A Yeah, so we can get into that a little bit more. So your, your actinides and your minor actinides, so your plutonium, your neptunium, your americium, so all those sorts of things are, grow in, and they basically constitute the long-term waste management burden for the nuclear material, and this is something that's very different than what you have, say, in, in fusion nuclear waste, right, because fusion nuclear waste is just activation products. It's the, Uh, basically your, your transition metals and other things that were rated over, radiated over the course of that. And so, that maybe comprises about one percent of it by, by volume with respect to that. And then you have this four percent of other material that are, are your fission products. And so, those are things that if you're able to take them out, for the most part, could generate a waste management timeline of, on the order of more like several hundred years, not talking about something on the order of millions of years. There are Interceptions to that. For example, technetium 99 has a very long, uh, waste management burden. Things like iodine, uh, one, I believe it, I forget if it's one 29, one 31, that also can have a very long waste management burden. So there are a couple of standouts from that perspective, but it's largely your actinides that drive that waste management challenge over the long haul.

AI assessment note: “your actinides and your minor actinides, so your plutonium, your neptunium, your americium”

Q those intermediate pieces? Let me ask you this, though. Why, if you look at that sort of historic barbell of, we had some polysilicon, we had a fair amount of module assembly, why are the ones that, why are those the ones that got domesticated first? And people would assume it's low labor, basically. Are those the things that require the least labor per watt, or is it something else?

A On, you know, on the poly side, uh, the predominant cost for polysilicon manufacturing is, is the cost of energy, electricity to input, and so, you know, you have, uh, you look at REC Silicon, it's in Washington State, they have some of the cheapest hydroelectricity in the world, right? Same with Michigan, a lot of nuclear, a lot of coal, a lot of natural gas, like very low, relatively low electricity prices, and that sort of drove and enabled a lot of poly, which is also supporting the semiconductor industry, right? So, you know, there's a matter of which solar-grade Poly manufacturing is critical to semiconductors. On the module side, you know, part of it was just that it's, uh, the economics can make sense. It's not necessarily labor. It's the fact that it's bulky manufacturing, shipping costs matter, and it's relatively quick and simple. Not to say that they're not really, uh, incredible factories. We've had module manufacturing in Georgia for the last eight years. Anytime anyone visits it, they're, they're blown away. It used to be, you know, it's like, oh, we don't need these factories here. They're, they're highly automated. They're not a lot of jobs, but Even a, you know, two gigawatt module assembly plan is going to be 800 people, uh, well-paid manufacturers, engineers, technicians, all sorts of stuff, and it's a really impressive, complicated process, but it's not, it…

AI assessment note: “On the module side... It's not necessarily labor. It's the fact that it's bulky”

Q but we should talk about the whole supply chain, because I think it is important to do so and to recognize what that looks like. So, In the context of crystal and silicon solar, which is basically the entire market X for solar, um, the supply chain is polysilicon wafer cell module. So can you walk me through how much domestic manufacturing we have of each of those four steps?

A Yes, so even, again, even before the IRA, we had a fair amount of module assembly, which was supported by trade policy, and we've had, you know, a historically decent amount of polysilicon for You know, two decades now. And, you know, at the moment, we have enough module capacity to supply U.S. demand. That's, you know, 40 to 50 gigawatts a year, and there's probably even more than that. On the poly side, we're probably 10 to 20 gigawatts. It's a moving target. Obviously, there's three big poly manufacturers in the U.S., or at least there have been historically. You know, one of one, one of which is REC silicon, which has switched to silane gas. You know, something that we, my company, Qcells, really participated in. But then there's Hemlock and there's Valker. Valker mostly gets their semi-grade, or solar-grade polysilicon from Germany. So it's really just a couple of players for probably 10 to 20 gigawatts or so of poly capacity. And then in the middle, cells and wafers, you know, you didn't, you had, up until the IRA, you had zero of those factories. None. You know, and in fact, wafers, pretty much 99% of them were coming from China. They largely shifted to Southeast Asia over the last five years as a result of another trade case. But you've seen investments in that sector, finally. There's more cell manufacturing than there is wafer manufacturing. At the moment, there's onl…

AI assessment note: “we have enough module capacity to supply U.S. demand. That's, you know, 40 to 50 gigawatts”

Q and it's resilient supply chain, and so on. And so if you're trying to stand up a supply chain from, from virtually nothing, particularly if we're talking wafers or cells, which you guys are doing at Qcells in the US, um, you don't benefit from that. So like, does that, is that a meaningful disadvantage? Just that You know, we don't have enough suppliers for basically everything that you need.

A It's definitely a challenge to reshoring. It's like one of the critical first steps of it, right? You know, I think we face two challenges when you're trying to build a factory, which is the first of its kind in the United States. One is that, you know, you have to find a contractor to build, say, a wafer factory that's never built a wafer factory before. And then you have to work with, you know, local permitting jurisdictions that have definitely never permitted a solar cell or wafer factory before. So, There's a lot of education work that has to get done to overcome that. It is overcomable. It's not something that we can't do. Um, you know, we build things that are new in the United States all the time. It's just a matter of, um, getting that infrastructure in place and doing the education work. The supply chains, yeah, you're buying a lot of things off global markets, and, you know, I, you're doing work to not just reshore your own part of the market, but other parts of the market. First Solar's been good at this in Ohio. They have, you know, they've got a local glass float manufacturer who they've, you know, supported and helped invest over the years, and You know, I think a good analog is the auto industry. You know, in the South in particular, where I live, you've had, since like the late eighties, you've had a bunch of global car manufacturers move down here. You've had …

AI assessment note: “It's definitely a challenge to reshoring. It's like one of the critical first steps”

Q compare and contrast, right? I'm going to talk about the, the constraints on each of them, but why don't we start with the incumbent thing, the thing we are doing right now, where all the data centers are, which is like large hyperscale data centers connected to the grid. What do you think of as being like the core constraint to just delivering 10 x the compute in that way?

A Yeah, no, great, great question. I think it's going to make for a great, great conversation as we look across the different options here. I think the constraints on the grid side, right, are fairly well known at this point. Um, it's a speed issue in particular on the transmission side. How, how long can we, how much time will it take to build out the transmission capacity necessary to interconnect these mega sites, gigawatt scale sites to new power supply, ideally, you know, carbon free power supply. And in many markets, right, that's running five to seven years now, which is, is a pretty massive Timeline for data centers given the speed of, uh, power and speed of deployment on the AI build out that we're, we're trying to drive. Um, maybe the other couple issues that we should at least be mindful of here are power quality, right? These large data centers, especially as they cluster in certain locations, can have bigger impacts on the grid writ large, and the extent to which society, regulators, and utilities are willing to serve those customers if they have bigger grid impacts, I think is still a bit to be determined, and A space I'm watching pretty closely. And then, of course, maybe the third vector from my side would be, let's call it, you know, social license to operate. And we're seeing in many states, right, just blanket bans on new data center developments. We're seeing …

AI assessment note: “it's a speed issue in particular on the transmission side.”

Q but they're smaller and you put them at the edge. So I've talked a little bit about edge compute on the podcast before. You and I have spent a lot of time thinking about it separately. First of all, define what you think of as edge compute, um, because it is sort of malleable, and then, like, what is your latest thinking on what role that plays in the market?

A Yeah, so, so this is a really tricky question, right? Edge computing has been around for a while. Um, historically, it evolved to serve certain use cases like telecommunications, and more recently, video streaming, for example, is something that happens much closer to the edge than other hyperscale data center. Um, but moving forward, I think there's a school of thought that says that AI inference in particular might move to the edge, and I think the, you know, first principles argument that folks tend to make is that latency is going to matter more, um, and so siting compute infrastructure closer to demand just has a performance benefit that can't be met via, you know, large central sites in West Texas, for example. Um, as we've dug in a little more, I think that's a little bit of a red herring, and so let's come back to that in a second and talk about why you would actually pursue edge data centers and edge computing. Um, but latency has certainly been one of the reasons historically. Um, that said, edge computing can mean a few different things, to your point, right? Um, so in the extreme scenario, I think, you know, as you move out, you know, 10, 10 plus years, more and more is going to happen on device. We already know that, um, you know, like Waymo cars, for example, have a lot of their day-to-day, or all of their day-to-day navigational tools and driving decisions get ma…

AI assessment note: “edge computing can mean a few different things, to your point, right?”