theCUBE + NYSE Wired: Powering Tomorrow | Yasir Arafat, Aalo Atomics

Clips
More from theCUBE + NYSE Wired: Powering Tomorrow

play_circle_outline High AI data center power demand: 80–100 gigawatts needed over next five years.

play_circle_outline Design Once, Build Many: Factory Manufacturing to Cut First-of-a-Kind Costs for Hyperscalers

play_circle_outline Austin Pilot Factory: Modular Container Reactors Assembled Like LEGO, Built in 36 Days, Order-to-Electrons Under 12 Months

play_circle_outline Purpose-built 50 MW Small Modular Reactor Plant with Five Reactors and Three Turbines Powering Co-Located AI Data Centers

play_circle_outline Partnership with Crusoe to demonstrate co-located nuclear-powered AI data center deployments.

Info

Yasir Arafat, Aalo Atomics

search

Gemma Allen

>> Palo Alto Studio Connection, Silicon Valley and Wall Street.

Yasir Arafat

>> I'm John Furrier, co-host here with David Vellante, my co-host.

Gemma Allen

>> Welcome back to theCUBE Studio here at the New York Stock Exchange. This is Gemma Allen with NYSE Wired, powering tomorrow. And one thing we know for sure is that AI has a power problem and nuclear may be the answer. Aalo Atomics is building a new generation of small, factory-made nuclear reactors designed for reliable power right next to data centers consuming it. And this summer, they hit a huge milestone. Its reactor reached criticality, meaning it achieved a sustained nuclear chain reaction. Joining me now to understand exactly what that means is Aalo co-founder, CTO, and president Yasir Arafat. Welcome, Yasir.

Yasir Arafat

>> Thanks, Gemma. Thanks for having me here.

Gemma Allen

>> So Nuclear. what a fascinating world. It's a world that we all know of for many reasons. But you guys have really done something and achieved something very interesting this summer. As I said in my opening, for those of us not in the nuclear space, maybe unpack that for us. What does criticality mean?

Yasir Arafat

>> Yeah. So just literally 8 weeks ago, we built our first nuclear reactor of our design. And took the reactor to criticality, which basically means, for your reactor, you were able to sustain a fission chain reaction inside the machine. So in a nuclear reactor, you have always a bunch of neutrons flying around. They go and hit a uranium atom and then make 3 more neutrons, and they can kind of propagate There's a lot of neutrons being born and a lot of neutrons being lost. When you can balance those two things right in a controlled manner, that's what's called criticality. Essentially, it's very similar to hopping into a car and turning the ignition on, but you haven't yet pressed the gas to go from point A to point B. Everything works. The engine is running, all the lights are on, your AC is working. So that's essentially very equivalent to what a criticality event looks like. Wow.

Gemma Allen

>> And let's talk for a second about nuclear broadly. I know we spoke when we were chatting off screen there about The Simpsons, right? And the whole world of the nuclear power plant and the PR issue it's had for a long time. But it's certainly on somewhat of a PR reimagination. We think about the world of AI and power and how much power will need to be consumed to make things work in the way this industry is headed. Talk to me a little bit about, the decision to really— and what you're seeing too, to really commercialize this, bring that to life for us. And also tell me about the decision to start in this space with data centers as opposed to working with maybe some of the traditional telecom or, you know, power suppliers. Like, give me the backstory here.

Yasir Arafat

>> Yeah, absolutely. So a lot of folks don't know this, but 20% of the United States electricity comes from nuclear power plants. And these power plants were built from '50s to mid-'70s, '80s, and then we stopped. For nearly 50 years, we haven't built any new nuclear power plants. Those old ones have been continuously running, and they work plenty reliable, plenty safe, unlike the public perception and whatnot. But because we stopped for a few decades, the country essentially has lost its muscle to build new nuclear power plants. Can we do it? Yes, but it takes a very long time. I think after many years, the country tried to build several gigawatt-scale nuclear power plants, and instead of taking 9 years, it took us about 18 to get those built. And that gives you a scale of how long each of these projects really takes. So when we started Aalo, I've been in the industry for many years and trying to kind of solve this challenge on how do we really fundamentally come up with a framework of a nuclear technology where we can deploy faster? It doesn't take a decade or two. Can we deploy 10x faster potentially? So that whole idea about factory manufacturing, mass manufacturing is at the center of all of this. Now, factory mass production of nuclear power plants is not a new idea. It's been around since the '50s. But back then, there was a missing link. nobody could really pull it off because we didn't have a big demand the way we are seeing today. And that big demand is coming from AI data centers. Just in the next 5 years, the country wants 80 to 100 gigawatts worth of power that we currently don't have. And the grid is too old to pull all that power so rapidly.

Gemma Allen

>> Okay. So on the topic of building power plants, a couple of years ago, we had a lot of press, a lot of headlines around the Microsoft investment in Three Mile Island, right? There was a lot of industry speculation on what this could mean. Are we going to see more of this, etc.? I think it's due to go live in 2027, if I'm not mistaken. But what interests me when I read about this was actually the economics of it, because at no point are they really saying that immediately this is a cheaper option, right? Like, I think they're saying it's about $115 per megawatt hour for nuclear, which is, what, maybe 2x on what we expect. Talk a little bit about, how you think about the economics of this market. Are you planning to create an option here that could sell into hyperscalers. Where do you see the numbers adding up kind of longer term and what is the overall economic viability? Yes, Yasir.

Yasir Arafat

>> Yeah, totally. And this is what I was talking about earlier is the model that we are taking is instead of building a gigawatt-scale plant that takes 10 years and not knowing how much it will actually take, we're shrinking that down to less than a year. So every year we want to generate a gigawatt-scale power plant. And the way we do that is instead of building a single 1,000-megawatt plant, we build 100 of these 10-megawatt nuclear reactors that we can mass manufacture the way we build cars or jet engines or smaller machines in factories. And this ties in essentially to our ultimate thesis in Aalo is why Aalo really even exists. It's not that nuclear is not safe. It's plenty safe if you look at the data right now. It's not that it's not reliable. It runs compared to any other power source, at very high availability. But it's unpredictable in how long it takes and how much it costs. To your question, though, the advantage of building, design once, build many, is that you can actually come down the cost curve pretty rapidly compared to just building one giant thing. And so our— to answer your question, instead of giving specific numbers, economics and economic viability have been at the very forefront of every major tech decision we have made in Aalo since the beginning. Right. It all boils down to what kind of supply chain is available, what kind of fuel are we using? Do we get the right performance? How long does construction take? What are the costs of materials that make up the overall power plant from a first principles perspective? So all that being said, now we have a product that we're very confident is going to have a very strong economic posture. Now, a lot of people say, hey, nuclear first of a kind will be expensive, and then over time it's going to get to a point where it's attractive enough. We did not wait for that to happen. We're saying What if we can make the first of a kind attractive for today, for AI data center hyperscalers? What if we can make it low cost enough that it's attractive from a first of a kind perspective? And over time, as we build more, then it becomes even more cost effective over time, right? So that's kind of like the goal that we are marching towards. And if you think about economics, there's two fundamental drivers. One is the fundamental cost of materials that build up the hardware that makes the energy, but also the speed of execution. If you can handle— if you have a handle over those two things, then certainly you have full, full grip over your overall cost structure.

Gemma Allen

>> I guess if I'm correct, what you're kind of saying, if I'm summarizing correctly, is that you're thinking about nuclear more like an airplane factory than like a massive construction bet, right? That there's opportunity to scale the economics of this in the traditional way we've scaled other parts of our infrastructure supply chain. Talk to me though about that supply chain, maybe just for Aalo alone. Talk to me about the footprint, about the CapEx. this must be a huge undertaking to create a factory that can create in-a-box nuclear reactors, right? Bring it to life for us. help us visualize this and what your life looks like day to day.

Yasir Arafat

>> Yeah, no, totally. And there's multiple segments of that question that I want to slowly unfold here. So if you think about what our day to day looks like, what our mission looks like, how much capital is needed to pull all of this together. Unlike software, deep tech is very capital intensive. And you have to figure out how to not only just build the right product that has a great product-market fit, but also create the capability, the infrastructure, the manufacturing muscle, to create the right supply chain. There's a whole host of things, licensing, there's like 9 major areas that you have to really excel in to bring a successful nuclear product to life. So a key central area in all of this is our factory mass production. And why do we really overemphasize this? Because if you think about any product that you build, the cost of a product is entirely dependent on the process that builds it. So we did not wait to build a factory until our product kind of nuclear design matures. When we started Aalo, within 6 months, we built a 40,000-square-foot pilot factory here in Austin, Texas to figure out how do we learn and how can we build such hardware as we're maturing the design are there learnings we can see from the manufacturing side and bring it back to the design side of things? And so every single decision was geared towards how do we design a reactor that is optimal for manufacturability, optimal for assembly, optimal for transportability. So we don't transport bits and pieces, we transport everything in terms of modules So what are modules? These are basically predefined shipping container-sized boxes that are basically packed with equipment that builds up our power plants. So when you bring them to site, it's basically a bunch of these modules coming together like LEGO blocks, and the only thing you're doing is connecting the interfaces, the structural interface, the piping, the electrical, the power. You snap them together and you get a power plant stood up in a matter of a few months, not a few years or decades.

Gemma Allen

>> So from an order to

Yasir Arafat

>> deployment—that's our fundamental model.

Gemma Allen

>> From an order to deployment perspective, you are saying that this could happen, this kind of in-a-box, out-of-a-box deployment solution in less than a year. Do you have an estimate, a number of months that you're targeting for this? Because that's truly fascinating.

Yasir Arafat

>> Yeah, speed, speed is the name of the game here. And in fact, we are targeting something that most people will say is impossible today, which is ultimately I want to build Aalo's capabilities such that when a customer orders and says, hey Aalo, we need a 50 megawatt or 500 megawatt or a gigawatt or two, from 50 megawatt to multi-gigawatts, we can provide any power level. But by the time they give us an order, from order to providing electrons, can we do it in less than 12 months?

Gemma Allen

>> Wow.

Yasir Arafat

>> Right. So what does that mean? words are cheap, right? Can we put that to practice? Have we built the muscle? Can we demonstrate all of that? In fact, we have. So the first reactor that we have built was in Idaho National Laboratory. That was the reactor that we went critical with. And we didn't just want to go critical because President Trump has an executive order saying, I want at least 3 new nuclear reactors to go critical within 1 year timeframe in order for the US to win this power race against other nations like China or Russia. That was, yes, that was a central goal, but we also wanted to prove some of our fundamental thesis about Aalo. Can we do things fast? Can we construct our building rapidly? Can we build our reactor in a factory setting? Can we build our own fuel bundles? Can we operate them for a completely new technology? We proved all of those questions. First of all, from— let's talk about construction. When we first wanted to build out our reactor building, which normally takes many, many years in nuclear, folks at the Department of Energy and Idaho National Laboratory told us like, hey, it's going to take forever to build out here in the desert. Even if you want to build a tool shed here, it takes less— no, more than, you can't do it under 2 years. So we said, okay, let's go give it a shot. And we built our entire nuclear building in 36 days.

Gemma Allen

>> Wow.

Yasir Arafat

>> Right? And we said, okay, challenge number 2. Everybody talks about, you know, mass-producing nuclear in factories. Can we indeed demonstrate that? So we built our entire reactor here in our pilot factory in Austin in 4 weeks. And put it on the back of a truck and transported all the way to— from Texas to Idaho and installed 80% of that hardware in about a week or so. And we built 540 fuel elements into bundles in 2.5 days. So when you stitch all of these together, from construction to criticality for our first reactor, it took us less than 8 months. That was the fastest nuclear project in the last 80 years in the United States.

Gemma Allen

>> That's incredible. Talk about your TAM. Who are your, do you have proof of concepts of particular customers? What sort of verticals are you seeing the appetite for here? We spoke at the beginning. I know so much is really about, you know, the PR issue around nuclear, but I'm sure it's still somewhat of a taboo topic for many, right? Talk about what sorts of conversations you're having. Like, what does the TAM look like for you right now?

Yasir Arafat

>> Yeah, I mean, I'll tell you this, you know, there has never been in the last 50 years more support for nuclear from a public perception perspective than where we are today. And the reason is because there's not a whole lot of misinformation. People can study about nuclear, they can check the facts out, they can actually see how reliable and safe and amazing this technology is. And that we don't make enough of it today, right? But if you think about, you know, what the market looks like, when you create an energy source that can be deployed fast, that has no emissions, that is clean, that does not require water for ultimate cooling, it becomes an attractive product for many markets. But if we're creating a product for everybody, we're not really designing an ideal product for anyone. Right. So our initial market that we want to aim particularly is to provide on-site power for AI data centers. That is the first market we want to go and tackle because the demand is so high. There's a need for power and, you know, there isn't a similar need that is as large as AI data centers today. And that's our main target. And in fact, We're the only nuclear company that is hyperfocused on creating a purpose-built product for AI data center. What do I mean by that? We don't have a reactor, we're making a power plant just like traditional nuclear. Our power plant solution is actually a 50-megawatt power generating power plant that sits in a 2.5-acre site However, it has 5 reactors and 3 turbines. Why did we do that? So that if one reactor is down for refueling or maintenance, or if the turbine is down for maintenance, we can always maintain 99.9% reliability from the overall power plant.

Gemma Allen

>> Am I correct that you guys have just entered a partnership with Crusoe?

Yasir Arafat

>> That's correct. Yes.

Gemma Allen

>> Talk about where that is headed. You mentioned data centers. They're very large, NeoCloud, very successful. So perfect fit for you, I guess, based on what you've just told us. What— where do you see this going? What percentage of Crusoe's footprint do you envision as one such example could be powered by nuclear 10 years from now?

Yasir Arafat

>> No, absolutely. And I think, to answer your first question, Where is that partnership headed to? Recently we signed an agreement, a partnership agreement with Crusoe, which we're very excited about. It's taking that first step, how do you have a co-located and co-built AI data center and nuclear on the same site, and how does this get built together and connect and empower each other without necessarily putting that strain on the grid, right? That demonstration is key to unlocking essentially the whole market. Is it going to work? Absolutely, we know it will, but it's important to show and demonstrate and show that it can actually be done. So that's our first step, and from there it's not the screw-up, but then there's— we're talking to multiple array of customers where, some of— we're always in the near term going to be supply limited. There's way more demand than what we can actually supply, even from our big factory. Right now we have 40,000 square foot facility. We're going to be expanding to 1 million square feet over the next 3 years so we can get to a 100 reactors a year cadence in the production level. Even then, we're going to be supply limited. That's how big the TAM really is. So how are we tackling that? What are some of these partners really looking at? What we are seeing, interestingly enough, is that companies like Crusoe and others are not saying Hey, I want a reactor or two for this site and that power plant or that project. They're saying, I want 20% of your production capacity from your factory at any given time. That's a very interesting model, right? So if you think about if we create multiple customer tracks, each for a 20% slot, by the time we hit 80% capacity reservation for the factory, I can then go build my factory number 2 and create that balance between supply and demand, right? So ultimately, in the near term, from each of our factory, we want to get at least a gigawatt or more minimum from our single factory, maybe 1 to 10 gigawatts from each of our factories. And we want to be able to build multiple factories as you go along, as there's more demand coming in. And that model is not impossible. We've seen that happening in the chip industry. We've seen that happening in automotive. We've seen that happening in jet engines or cars. There's so many other deep tech industries that have pulled this off. So we're essentially creating that SpaceX type of a model in the nuclear industry, essentially.

Gemma Allen

>> So, Yasir, last question to you. Your energy is palpable. This is a very interesting concept. You certainly talk a great game. We wish you guys all the best. Close us out. Tell us what's ahead for you and the team at Aalo. Are you raising money? Talk to our audience for a second. Give us the pitch for the last few seconds.

Yasir Arafat

>> Yeah. At Aalo, we're mass manufacturing nuclear reactors and we're going to be the first ones to make it a reality. And we are scaling and executing very rapidly. We're only a 2.5-year company from a 2-person team. We are now a 200-strong team, and we're one of the first companies that have built a reactor pretty rapidly and went critical. So we have proven that first stage of execution. Our goal is to really scale this up to a point where we can make a positive dent into the whole AI data center space and we can bring down the cost curve we can then tackle other industrial sectors that need power. When we can bring the cost down even further, then all of a sudden we can become a universal energy source for anybody in the world and make it ubiquitous. So, in the future, what we want to see is not a question of, oh, when is nuclear going to come? I think we want to create a future where we're no longer talking about, hey, is there a reliable energy source that is carbon-free? And scalable and can be deployed anywhere in the world. Me and my co-founder, we have a vision that we say, hey, if you think about the population of this Earth, 50% of the world's population live on 3% of the Earth's surface. There are many areas on Earth that is uninhabitable. That's because we don't have access to clean water or energy. And so one day we would like to see our technology unlocking all those areas for humans to expand on the surface of this Earth and also go beyond to the Moon and Mars and so on and so forth. So we want to make unlivable places livable with our product. Ultimately, that's the ultimate goal.

Gemma Allen

>> Well, Yasir, that is certainly a fascinating statistic, and I also did not know that to end on. And certainly the type of conversations we love to have here on Powering Tomorrow on NYSE Wired. So, Yasir Arafat, thank you so much for joining us. Hope we can have you here in our studio in the NYSE in the not so distant future.

Yasir Arafat

>> Awesome. Thanks for having me.

Gemma Allen

>> I'm Gemma Allen here at theCUBE Studio at the New York Stock Exchange. This is Powering Tomorrow, one of our segments at NYSE Wired. Thanks for watching.

Yasir Arafat, Aalo Atomics

search

Gemma Allen

>> Palo Alto Studio Connection, Silicon Valley and Wall Street.

Yasir Arafat

>> I'm John Furrier, co-host here with David Vellante, my co-host.

Gemma Allen

>> Welcome back to theCUBE Studio here at the New York Stock Exchange. This is Gemma Allen with NYSE Wired, powering tomorrow. And one thing we know for sure is that AI has a power problem and nuclear may be the answer. Aalo Atomics is building a new generation of small, factory-made nuclear reactors designed for reliable power right next to data centers consuming it. And this summer, they hit a huge milestone. Its reactor reached criticality, meaning it achieved a sustained nuclear chain reaction. Joining me now to understand exactly what that means is Aalo co-founder, CTO, and president Yasir Arafat. Welcome, Yasir.

Yasir Arafat

>> Thanks, Gemma. Thanks for having me here.

Gemma Allen

>> So Nuclear. what a fascinating world. It's a world that we all know of for many reasons. But you guys have really done something and achieved something very interesting this summer. As I said in my opening, for those of us not in the nuclear space, maybe unpack that for us. What does criticality mean?

Yasir Arafat

>> Yeah. So just literally 8 weeks ago, we built our first nuclear reactor of our design. And took the reactor to criticality, which basically means, for your reactor, you were able to sustain a fission chain reaction inside the machine. So in a nuclear reactor, you have always a bunch of neutrons flying around. They go and hit a uranium atom and then make 3 more neutrons, and they can kind of propagate There's a lot of neutrons being born and a lot of neutrons being lost. When you can balance those two things right in a controlled manner, that's what's called criticality. Essentially, it's very similar to hopping into a car and turning the ignition on, but you haven't yet pressed the gas to go from point A to point B. Everything works. The engine is running, all the lights are on, your AC is working. So that's essentially very equivalent to what a criticality event looks like. Wow.

Gemma Allen

>> And let's talk for a second about nuclear broadly. I know we spoke when we were chatting off screen there about The Simpsons, right? And the whole world of the nuclear power plant and the PR issue it's had for a long time. But it's certainly on somewhat of a PR reimagination. We think about the world of AI and power and how much power will need to be consumed to make things work in the way this industry is headed. Talk to me a little bit about, the decision to really— and what you're seeing too, to really commercialize this, bring that to life for us. And also tell me about the decision to start in this space with data centers as opposed to working with maybe some of the traditional telecom or, you know, power suppliers. Like, give me the backstory here.

Yasir Arafat

>> Yeah, absolutely. So a lot of folks don't know this, but 20% of the United States electricity comes from nuclear power plants. And these power plants were built from '50s to mid-'70s, '80s, and then we stopped. For nearly 50 years, we haven't built any new nuclear power plants. Those old ones have been continuously running, and they work plenty reliable, plenty safe, unlike the public perception and whatnot. But because we stopped for a few decades, the country essentially has lost its muscle to build new nuclear power plants. Can we do it? Yes, but it takes a very long time. I think after many years, the country tried to build several gigawatt-scale nuclear power plants, and instead of taking 9 years, it took us about 18 to get those built. And that gives you a scale of how long each of these projects really takes. So when we started Aalo, I've been in the industry for many years and trying to kind of solve this challenge on how do we really fundamentally come up with a framework of a nuclear technology where we can deploy faster? It doesn't take a decade or two. Can we deploy 10x faster potentially? So that whole idea about factory manufacturing, mass manufacturing is at the center of all of this. Now, factory mass production of nuclear power plants is not a new idea. It's been around since the '50s. But back then, there was a missing link. nobody could really pull it off because we didn't have a big demand the way we are seeing today. And that big demand is coming from AI data centers. Just in the next 5 years, the country wants 80 to 100 gigawatts worth of power that we currently don't have. And the grid is too old to pull all that power so rapidly.

Gemma Allen

>> Okay. So on the topic of building power plants, a couple of years ago, we had a lot of press, a lot of headlines around the Microsoft investment in Three Mile Island, right? There was a lot of industry speculation on what this could mean. Are we going to see more of this, etc.? I think it's due to go live in 2027, if I'm not mistaken. But what interests me when I read about this was actually the economics of it, because at no point are they really saying that immediately this is a cheaper option, right? Like, I think they're saying it's about $115 per megawatt hour for nuclear, which is, what, maybe 2x on what we expect. Talk a little bit about, how you think about the economics of this market. Are you planning to create an option here that could sell into hyperscalers. Where do you see the numbers adding up kind of longer term and what is the overall economic viability? Yes, Yasir.

Yasir Arafat

>> Yeah, totally. And this is what I was talking about earlier is the model that we are taking is instead of building a gigawatt-scale plant that takes 10 years and not knowing how much it will actually take, we're shrinking that down to less than a year. So every year we want to generate a gigawatt-scale power plant. And the way we do that is instead of building a single 1,000-megawatt plant, we build 100 of these 10-megawatt nuclear reactors that we can mass manufacture the way we build cars or jet engines or smaller machines in factories. And this ties in essentially to our ultimate thesis in Aalo is why Aalo really even exists. It's not that nuclear is not safe. It's plenty safe if you look at the data right now. It's not that it's not reliable. It runs compared to any other power source, at very high availability. But it's unpredictable in how long it takes and how much it costs. To your question, though, the advantage of building, design once, build many, is that you can actually come down the cost curve pretty rapidly compared to just building one giant thing. And so our— to answer your question, instead of giving specific numbers, economics and economic viability have been at the very forefront of every major tech decision we have made in Aalo since the beginning. Right. It all boils down to what kind of supply chain is available, what kind of fuel are we using? Do we get the right performance? How long does construction take? What are the costs of materials that make up the overall power plant from a first principles perspective? So all that being said, now we have a product that we're very confident is going to have a very strong economic posture. Now, a lot of people say, hey, nuclear first of a kind will be expensive, and then over time it's going to get to a point where it's attractive enough. We did not wait for that to happen. We're saying What if we can make the first of a kind attractive for today, for AI data center hyperscalers? What if we can make it low cost enough that it's attractive from a first of a kind perspective? And over time, as we build more, then it becomes even more cost effective over time, right? So that's kind of like the goal that we are marching towards. And if you think about economics, there's two fundamental drivers. One is the fundamental cost of materials that build up the hardware that makes the energy, but also the speed of execution. If you can handle— if you have a handle over those two things, then certainly you have full, full grip over your overall cost structure.

Gemma Allen

>> I guess if I'm correct, what you're kind of saying, if I'm summarizing correctly, is that you're thinking about nuclear more like an airplane factory than like a massive construction bet, right? That there's opportunity to scale the economics of this in the traditional way we've scaled other parts of our infrastructure supply chain. Talk to me though about that supply chain, maybe just for Aalo alone. Talk to me about the footprint, about the CapEx. this must be a huge undertaking to create a factory that can create in-a-box nuclear reactors, right? Bring it to life for us. help us visualize this and what your life looks like day to day.

Yasir Arafat

>> Yeah, no, totally. And there's multiple segments of that question that I want to slowly unfold here. So if you think about what our day to day looks like, what our mission looks like, how much capital is needed to pull all of this together. Unlike software, deep tech is very capital intensive. And you have to figure out how to not only just build the right product that has a great product-market fit, but also create the capability, the infrastructure, the manufacturing muscle, to create the right supply chain. There's a whole host of things, licensing, there's like 9 major areas that you have to really excel in to bring a successful nuclear product to life. So a key central area in all of this is our factory mass production. And why do we really overemphasize this? Because if you think about any product that you build, the cost of a product is entirely dependent on the process that builds it. So we did not wait to build a factory until our product kind of nuclear design matures. When we started Aalo, within 6 months, we built a 40,000-square-foot pilot factory here in Austin, Texas to figure out how do we learn and how can we build such hardware as we're maturing the design are there learnings we can see from the manufacturing side and bring it back to the design side of things? And so every single decision was geared towards how do we design a reactor that is optimal for manufacturability, optimal for assembly, optimal for transportability. So we don't transport bits and pieces, we transport everything in terms of modules So what are modules? These are basically predefined shipping container-sized boxes that are basically packed with equipment that builds up our power plants. So when you bring them to site, it's basically a bunch of these modules coming together like LEGO blocks, and the only thing you're doing is connecting the interfaces, the structural interface, the piping, the electrical, the power. You snap them together and you get a power plant stood up in a matter of a few months, not a few years or decades.

Gemma Allen

>> So from an order to

Yasir Arafat

>> deployment—that's our fundamental model.

Gemma Allen

>> From an order to deployment perspective, you are saying that this could happen, this kind of in-a-box, out-of-a-box deployment solution in less than a year. Do you have an estimate, a number of months that you're targeting for this? Because that's truly fascinating.

Yasir Arafat

>> Yeah, speed, speed is the name of the game here. And in fact, we are targeting something that most people will say is impossible today, which is ultimately I want to build Aalo's capabilities such that when a customer orders and says, hey Aalo, we need a 50 megawatt or 500 megawatt or a gigawatt or two, from 50 megawatt to multi-gigawatts, we can provide any power level. But by the time they give us an order, from order to providing electrons, can we do it in less than 12 months?

Gemma Allen

>> Wow.

Yasir Arafat

>> Right. So what does that mean? words are cheap, right? Can we put that to practice? Have we built the muscle? Can we demonstrate all of that? In fact, we have. So the first reactor that we have built was in Idaho National Laboratory. That was the reactor that we went critical with. And we didn't just want to go critical because President Trump has an executive order saying, I want at least 3 new nuclear reactors to go critical within 1 year timeframe in order for the US to win this power race against other nations like China or Russia. That was, yes, that was a central goal, but we also wanted to prove some of our fundamental thesis about Aalo. Can we do things fast? Can we construct our building rapidly? Can we build our reactor in a factory setting? Can we build our own fuel bundles? Can we operate them for a completely new technology? We proved all of those questions. First of all, from— let's talk about construction. When we first wanted to build out our reactor building, which normally takes many, many years in nuclear, folks at the Department of Energy and Idaho National Laboratory told us like, hey, it's going to take forever to build out here in the desert. Even if you want to build a tool shed here, it takes less— no, more than, you can't do it under 2 years. So we said, okay, let's go give it a shot. And we built our entire nuclear building in 36 days.

Gemma Allen

>> Wow.

Yasir Arafat

>> Right? And we said, okay, challenge number 2. Everybody talks about, you know, mass-producing nuclear in factories. Can we indeed demonstrate that? So we built our entire reactor here in our pilot factory in Austin in 4 weeks. And put it on the back of a truck and transported all the way to— from Texas to Idaho and installed 80% of that hardware in about a week or so. And we built 540 fuel elements into bundles in 2.5 days. So when you stitch all of these together, from construction to criticality for our first reactor, it took us less than 8 months. That was the fastest nuclear project in the last 80 years in the United States.

Gemma Allen

>> That's incredible. Talk about your TAM. Who are your, do you have proof of concepts of particular customers? What sort of verticals are you seeing the appetite for here? We spoke at the beginning. I know so much is really about, you know, the PR issue around nuclear, but I'm sure it's still somewhat of a taboo topic for many, right? Talk about what sorts of conversations you're having. Like, what does the TAM look like for you right now?

Yasir Arafat

>> Yeah, I mean, I'll tell you this, you know, there has never been in the last 50 years more support for nuclear from a public perception perspective than where we are today. And the reason is because there's not a whole lot of misinformation. People can study about nuclear, they can check the facts out, they can actually see how reliable and safe and amazing this technology is. And that we don't make enough of it today, right? But if you think about, you know, what the market looks like, when you create an energy source that can be deployed fast, that has no emissions, that is clean, that does not require water for ultimate cooling, it becomes an attractive product for many markets. But if we're creating a product for everybody, we're not really designing an ideal product for anyone. Right. So our initial market that we want to aim particularly is to provide on-site power for AI data centers. That is the first market we want to go and tackle because the demand is so high. There's a need for power and, you know, there isn't a similar need that is as large as AI data centers today. And that's our main target. And in fact, We're the only nuclear company that is hyperfocused on creating a purpose-built product for AI data center. What do I mean by that? We don't have a reactor, we're making a power plant just like traditional nuclear. Our power plant solution is actually a 50-megawatt power generating power plant that sits in a 2.5-acre site However, it has 5 reactors and 3 turbines. Why did we do that? So that if one reactor is down for refueling or maintenance, or if the turbine is down for maintenance, we can always maintain 99.9% reliability from the overall power plant.

Gemma Allen

>> Am I correct that you guys have just entered a partnership with Crusoe?

Yasir Arafat

>> That's correct. Yes.

Gemma Allen

>> Talk about where that is headed. You mentioned data centers. They're very large, NeoCloud, very successful. So perfect fit for you, I guess, based on what you've just told us. What— where do you see this going? What percentage of Crusoe's footprint do you envision as one such example could be powered by nuclear 10 years from now?

Yasir Arafat

>> No, absolutely. And I think, to answer your first question, Where is that partnership headed to? Recently we signed an agreement, a partnership agreement with Crusoe, which we're very excited about. It's taking that first step, how do you have a co-located and co-built AI data center and nuclear on the same site, and how does this get built together and connect and empower each other without necessarily putting that strain on the grid, right? That demonstration is key to unlocking essentially the whole market. Is it going to work? Absolutely, we know it will, but it's important to show and demonstrate and show that it can actually be done. So that's our first step, and from there it's not the screw-up, but then there's— we're talking to multiple array of customers where, some of— we're always in the near term going to be supply limited. There's way more demand than what we can actually supply, even from our big factory. Right now we have 40,000 square foot facility. We're going to be expanding to 1 million square feet over the next 3 years so we can get to a 100 reactors a year cadence in the production level. Even then, we're going to be supply limited. That's how big the TAM really is. So how are we tackling that? What are some of these partners really looking at? What we are seeing, interestingly enough, is that companies like Crusoe and others are not saying Hey, I want a reactor or two for this site and that power plant or that project. They're saying, I want 20% of your production capacity from your factory at any given time. That's a very interesting model, right? So if you think about if we create multiple customer tracks, each for a 20% slot, by the time we hit 80% capacity reservation for the factory, I can then go build my factory number 2 and create that balance between supply and demand, right? So ultimately, in the near term, from each of our factory, we want to get at least a gigawatt or more minimum from our single factory, maybe 1 to 10 gigawatts from each of our factories. And we want to be able to build multiple factories as you go along, as there's more demand coming in. And that model is not impossible. We've seen that happening in the chip industry. We've seen that happening in automotive. We've seen that happening in jet engines or cars. There's so many other deep tech industries that have pulled this off. So we're essentially creating that SpaceX type of a model in the nuclear industry, essentially.

Gemma Allen

>> So, Yasir, last question to you. Your energy is palpable. This is a very interesting concept. You certainly talk a great game. We wish you guys all the best. Close us out. Tell us what's ahead for you and the team at Aalo. Are you raising money? Talk to our audience for a second. Give us the pitch for the last few seconds.

Yasir Arafat

>> Yeah. At Aalo, we're mass manufacturing nuclear reactors and we're going to be the first ones to make it a reality. And we are scaling and executing very rapidly. We're only a 2.5-year company from a 2-person team. We are now a 200-strong team, and we're one of the first companies that have built a reactor pretty rapidly and went critical. So we have proven that first stage of execution. Our goal is to really scale this up to a point where we can make a positive dent into the whole AI data center space and we can bring down the cost curve we can then tackle other industrial sectors that need power. When we can bring the cost down even further, then all of a sudden we can become a universal energy source for anybody in the world and make it ubiquitous. So, in the future, what we want to see is not a question of, oh, when is nuclear going to come? I think we want to create a future where we're no longer talking about, hey, is there a reliable energy source that is carbon-free? And scalable and can be deployed anywhere in the world. Me and my co-founder, we have a vision that we say, hey, if you think about the population of this Earth, 50% of the world's population live on 3% of the Earth's surface. There are many areas on Earth that is uninhabitable. That's because we don't have access to clean water or energy. And so one day we would like to see our technology unlocking all those areas for humans to expand on the surface of this Earth and also go beyond to the Moon and Mars and so on and so forth. So we want to make unlivable places livable with our product. Ultimately, that's the ultimate goal.

Gemma Allen

>> Well, Yasir, that is certainly a fascinating statistic, and I also did not know that to end on. And certainly the type of conversations we love to have here on Powering Tomorrow on NYSE Wired. So, Yasir Arafat, thank you so much for joining us. Hope we can have you here in our studio in the NYSE in the not so distant future.

Yasir Arafat

>> Awesome. Thanks for having me.

Gemma Allen

>> I'm Gemma Allen here at theCUBE Studio at the New York Stock Exchange. This is Powering Tomorrow, one of our segments at NYSE Wired. Thanks for watching.