Drew Walker & Assil Halimi, Apollo Atomics | theCUBE + NYSE Wired: Powering Tomorrow

0:00 Palo Alto Studio Connection Silicon Valley and Wall Street. I'm John B here with Dave Volante, my co-host. Welcome back to the Cube Studio here at the New York Stock Exchange. I'm J Allen, co host of NYC Wire's powering tomorrow. And we have a huge power problem coming at us. AI and data centers are driving electricity demand higher than ever before. But building traditional nuclear power has meant enormous projects, enormous capital, and very long timelines. Apollo Atomics is betting that nuclear doesn't have to be built that way anymore.

0:37 You're taking proven nuclear technology, dramatically shrinking the system, and trying to turn the reactor from a giant construction product into something that can actually be manufactured and deployed at scale. Joining me now to explain exactly how they're going to make that happen are Apollo Atomics co-founders Drew Walker, CEO, and Ail Haleimi, CEO. Welcome folks. >> Thank you for having us. >> I mean, I'm fascinated by folks like you that, you know, as young men, we're like, I want to be a nuclear physicist, right? And I'm going to change the world and I'm going to monetize this and commercialize it because it seems like it's one of the world's largest and most challenging problems to solve and reimagine.

1:14 >> You know it. So maybe I'll start with you Isil. You begun this journey at MIT if I'm correct have proven a technical thesis that you feel has commercial scalability. Help me unpack that a little bit more. >> Yeah, I mean first the interest in nuclear. Why nuclear? as the first question one nuclear is the cleanest one of the cleanest sources of making electricity that's actually firm and reliable that we have already in the grid. The question is how can we scale it faster. and then we have many many concepts within the academic world within you know what we have tried in history in terms of building new reactors. and so my whole PhD thesis was about how can we deploy much faster using what we have right now and really the speed is the name of the game here. So what we've done and what I specifically studied is essentially all the existing concepts and can we use what we have right now the what we call the water cooled reactors that operate in 70 to 80% of all words nuclear power plant and actually many of them are in construction right now mainly outside of the United States and solve the construction problem which is can we reduce the construction time for you know building these reactors from 10 years to less than two we do that with natural gas plants We build them in about two years or even less than two years. So how can we learn from other industries and actually deploy that within nuclear and the key to that is one single design change not reinventing nuclear not changing everything in the design but really picking one design change that can actually enable that to happen. And so that was a conclusion that and and the component is a steam generator that we have developed at MIT for over 15 years that actually enable us to reduce the footprint of what we call the nfair island or the reactor by an order of magnitude. Now what becomes possible is making the full reactor off the site assembling it testing it and delivering it to the site for a fixed price. And that's really important because right now if you want a nuclear reactor and ask for the cost the answer is typically a question mark. so removing that risk from the construction project and saying hey we will deliver the reactor for a fixed price is really important to create confidence in any developer that want actually to use the clear for their energy generation. So that's that's the thesis how we got into you know using MIT developed technology into solving a very specific narrow problem while leveraging an existing infrastructure and existing supply chain that we have in what we call lightwater reactors.

3:49 >> So your end product it's not necessarily a massive construction investment in a nuclear power plant. What you're saying is that you can create a plugandplay element of the nuclear supply chain and roll that out in so many use cases across the USA. Correct. >> Yeah, that's very correct. >> So, fast forward to 2025, it's the summer. You defend this C thesis and you're thinking, okay, now I'm going to scale this and I'm going to try and understand how I can commercialize this and then enter Drew.

4:18 >> Correct. >> Yes. Correct. >> Okay. So Drew, let's talk about that because when we think about nuclear, we know that there is a huge energy crisis. I mean, it's happening. It's not even on the horizon. It's here, right? >> Yeah. Exactly. >> But nuclear has always had like, you know, if we're being honest, very confusing and mixed resonance, right? There's a lot of like bad aftertaste when you hear the word nuclear power plant. I'm not just talking about the world of Homer Simpson, right?

4:44 >> Totally. >> Even though I mean he didn't help the marketing message for sure. >> Yeah. So talk to me about the decision then to bring this to market to say okay we want to commercialize this. >> So I think where we can maybe start is our we decided to go down our product chain or our product road map of we have a 10 megawatt a 50 megawatt and a 300 megawatt. So we're attach attacking a few different markets based on those size and those scale. So a 10 and 50 we're starting out with which are more of those behind the meter type solutions for industrial power off takers smaller data centers that we can actually put next to each other and actually scale from there. So that's t tackling a different problem because we can actually make that reactor in the same factory like was saying we make it in our factory we put it on a truck and ship it to the site.

5:30 So it's a much much smaller much much smaller project alto together. So as far as like the the messaging that we're doing is for the whole nuclear industry, what we want to do is we want to make this approachable to the common person that understand that what this new nuclear renaissance is bringing to us as a nation and how we can actually solve the power problem and beat China along the way. So >> I mean when we hear nuclear I mean we think great, no carbon, no CO2 emissions, fantastic, right? We also think but hey are we going to get like a whole ton of other problems on the other end from the perspective of chain reaction radiation, right? And I think that we've never and I don't think any country really has done a very good job of really messaging that you know of really helping us understand well what the future of this industry could look like.

6:15 >> Yeah. >> So I want to understand a couple of things from the perspective of the supply chain opportunity for you guys. Who are you predominantly targeting? Like who is the if you had you know if you could make this product at scale tomorrow who is the immediate use case like what sorts of conversations are you having right now what commits are you entering into with companies? >> Yeah at the moment we have over 20 gawatt of lois from people going from the industry data centers that want to power AI models to university campuses to foreign countries that are looking for you know clean firm electricity for their grids. and within those we have several of them we're at the contracting stage. We want to first start in the United States making sure that we can develop this and actually deploy the first model correctly on time and on budget and then use that model as as a template to deploy beyond the United States and also beyond that first project specifically.

7:17 >> So where are you at with the P for this? Have you guys actually developed this? this I mean you've obviously created the technology within your PhD but have you actually created like a tangible product >> and and have you worked with anyone like on any particular manufacturing sites across the US that have basically taken this PC and said yeah this really works and it's the backbone >> of you know a particular element of a supply chain right it might not not again like I said be the depiction we have of a nuclear power plant instead it's part of an entire holistic supply chain that's more efficient >> definitely I think just quickly I would just say that we have built a a 40 kW demonstrator with MIT showcasing our technology and showcasing our compact stream generator that is our energy innovation to our reactor. and I would say from there we've been able to showcase and bring some really great partners along the way with us and and really securing that supply chain. So we have I would say one of the largest fuel manufacturers in the world that we have a commercial agreement with already pumps u everything else that is a normal pressurized water reactor supply chain that's already there. We're we're aligning them up and getting them to hold to our timelines which are very aggressive. I don't think we've talked about that yet, but want to deploy our first plant, commercial plant in 2028.

8:28 So, so really bringing so what keeps us up at night is going to be aligning that supply chain to hit those quick deadlines. >> Yeah. And maybe one important addition is that the reactor demo that we're running out of MIT is at commercial conditions. this is not a simple R&D test here and there. It's running at high reliability factors at the same temperature, chemistry, and pressure that you find in a commercial plant. That's really important because the data are important in terms of getting the right financing for the first project that we have. so the data related to this new component. All the rest of the components are, you know, traditional. So we have already a knowhow. So that's really important to get into a point where you can actually contract. And this is what we're discussing at the moment. make sure that we have enough data to contract for 10 years, 15 years. you know, we call power purchasing agreements. and so that's what we're producing. That's for 2026 our reactor demo that we're running out of MIT. And we're building another demo that is 50 times more powerful by mid 2027.

9:31 >> Wow. >> And then in 2028, we'll have our first commercial reactor scaled up from the 1 megawatt test that we have in 2027. So this has very much left the lab and it's about to enter the factory >> flooring. Yeah. >> In terms of your own supply chain for building these reactors. Where is that activity that footprint happening? Like if you were to create, you know, again something that's highly repeatable and you can build these at scale >> where where will that happen? Where is this currently happening? And what is the kind of you know design to implementation time frame for something like this? like how quickly can a hyperscaler for example I don't know if you guys are pedalling in the tech space just yet but I have no doubt that you will yeah >> how quickly can they deploy this >> yeah so the first we call the beach head market this these are applications are we call behind the meter people who cannot wait to get connected to the grid they're actually looking for more redundancy at less of a scale so looking for smaller reactors that they can have multiple of them in one installation if they lose one asset they still have honors. so this is kind of our beach head market where we have these small reactors, our first 10 megawatt reactor which by the way we're assemble it in the Bay Area and then >> expensive choice.

10:47 >> Yes. Yes. This is so our installation in the Bay Area is mainly for engineering, licensing and manufacturing of our key components and the first assembly. And then when we want to scale up the manufacturing, we're looking at two states mainly Louisiana and Texas. Okay. in in the US. >> Okay. So, I was going to go to cost, but first I'm going to go to regulation cuz I want to talk about Louisiana and Texas. We had an interesting conversation yesterday with a fellow MIT graduate and brains of tomorrow, Bobard, who is building he's he's not talking about fision, he's talking about fusion, right? You guys are probably familiar.

11:22 He talked about the regulation space in nuclear and this idea of like agreement states where states essentially regulate themselves, right? And I think part of the challenge that we have with the resonance and the image of nuclear is that we haven't really solved for that. Right? I think if you were to ask the average person on the street, how is nuclear energy regulated today? Most folks would struggle to tell you. So when you guys are thinking about this and thinking about, you know, deploying this and building these manufacturing footprints in different states. Talk about the policy and regulatory element to that. Like it must be a huge part of it. And how willing are these states too to say, you know what, we need this.

12:01 This is a clean energy solution. >> Yeah, maybe I can touch on it real quick. Is that I think from a regulatory perspective, we're trying to build a coalition both at the state and federal level from the governor governor's office on down. So because from the local and state buyin, we have to have it. And same thing with the federal government like we're going through the tens of our part 50 pathway for our like commercialization of this reactor. But to get those to site these reactors locally, we have to have that buy in from the beginning. And we're having those conversations even before we have a a specific site that we're citing these at their location because we have to get the community buy in. We got to get the state and and find make sure the incentives behind it are do align with the project. So takes a lot of coordination and effort to make sure those >> and from the regulatory point of view I think that the current administration has been doing great work in terms of bringing more predictability. U so we have a reactor that the US nuclear electric commission had approved dozens of times in the past. This is not an exotic technology that they don't know.

12:59 They already know about it a lot and we're derisking specifically this new component that makes us the best for the rest of it. They are very familiar and actually two executive orders from last year reduce the cost but also reduce the uncertainty on the timelines. It used to be you submit an application and then hope for the best. Now you submit an application and there's a clear timeline of when you're getting are you getting a response for a construction permit and this is why we can assert that you know for our first commercial plant it's going to happen this commercial this construction permit is going to happen in 2028. Let's talk about costs for a second because what I found interesting from the interview you had yesterday and other interviews you've had in this show is folks are willing to like upfront a lot of money on these commitments these 10 15 year 20ear commitments in the nuclear space on plants that haven't even broken ground yet right like Google the Italian Microsoft it means you might island with their you know consella partnership again a lot of money is being spent on a promise right yes >> and I mean it's a promise that I think people now realize probably has to come true because of the demand in the energy space, but the capex up front is huge to and and from what I understand and correct me on this, if you actually price out nuclear versus, you know, the less clean tech energy sources we have today, it's actually right now more expensive. Correct.

14:23 >> Yeah. >> So, it's not necessarily cheaper, but it has huge opportunity like long-term or midtail opportunity. talk about the conversations with the investors in this space. Like they obviously have to be super familiar, super comfortable already. I'm interested to understand when you're in those rooms how those economics are playing out. >> Yeah. And we we want to be clear, we're not developing nuclear just because it's clean. We're actually developing nuclear that is going to compete with natural gas. That's our clear goal unlike some of the other players.

14:53 when we look at you know comparing the cost specifically the capex between nuclear and let's say natural gas because that's the best in you know outside of nuclear category natural gas is typically about 70 to 80% cost of the fuel and then the capex is a smaller portion and the know operating the plant and maintaining the plant. So 70 to 80% cost of the fuel means that you're highly exposed to the volatility of the >> you know natural gas and we're seeing it every day. the good thing with nuclear is that the capex is the largest portion. The fuel is a smaller portion.

15:26 We're talking about you know 10% mostly depending on the size of the reactor but roughly about 10%. Which means that even if although the uranium is well spread across the planet there is an issue on the fuel supply if you do 2x on the the cost of the fuel you're still not affecting the price of electricity down the road. So that's one good thing about about nuclear that we have in terms of resiliency to supply chains. and then as a company we see ourselves not only as a hard tech company, we're also a software company. We have detailed cost models call them bottomup cost models where we look at every component and every system that goes into the power plant and actually bring that as a proof based on feedback that we get from the suppliers.

16:12 Remember, we're starting from a technology that we already know how to build and how to operate and so we have the cost figures and so we aggregate all that and then build it in a model where we have our new steam generator that enable us to have much higher compactness and that's how we build our proof and I think that's important you know the technoeconomics beyond the promises is really important because when you get to the you know the stage where you're actually contracting we get to the stage where you're actually you know asking for financing the project That's what people will ask for. They will ask for reliability. They don't want a reactor that breaks after 2 weeks. It needs to be reliable at least for 101 15 years. So they can sign a contract for 105 years. And they look at the costs and where are the sources of your costs. And so having that ability to develop a very detailed model to actually aggregate all the costs from the suppliers is very very important.

17:06 And then there is a third layer which is the financing layer. this is why we're focusing on having shorter construction timelines. When you look at the total cost of nuclear, we talk about capex. Most of that capex is actually cost of depth and equity >> that you have to pay during the construction period because it's massive upfront cost and then you have to wait for the construction time. Now that we have decoupled the reactor from the construction on the site, we can do that in parallel and shorten the construction period. that will enable us actually to reduce the capital cost, reduce the cost of depth, reduce the cost of equity and that's the the other layer that we have in terms of how can we reduce the cost of nuclear in the next 5 to 10 years.

17:49 Although you didn't say the magic word AI. So like we also I think there's a lot of you know the one of the biggest costs of nuclear is the soft cost like the regulatory side and bringing some of these things that AI and these systems that we're actually developing as as said we're that software company not on top of that hardware company and building some of these systems internally so that we can take on and you utilize the agentic workflows of today to help us bring some of these costs down that will help us deliver on these projects that we're agreeing to. And you know it's an interesting conundrum in the technoeconomics of AI and nuclear because in the minds of tech investors in the tech community sure right like it's you know chicken and egg one will solve for the other in the public mind though and the you know public opinion and the hearts and minds of Americans it's it's confusing because these two things at once you know where we you have challenges right now out there in like public perception right so like it is it is kind of an interesting dance >> it is it is >> but okay so last question to you guys. I mean when I think about a company like this, a product like this, you know, and the chops that you guys have in terms of what you've already proven MIT, I mean, people are like, "Yeah, these guys are smart. They know exactly what they're talking about, right?" Does the go to market kind of happen itself?

19:04 Like what is the goto market like for you? And how much priority is spent right now on winning these like you know m min commits from whoever it might be a hyperscaler or you know some large manufacturing >> global fortune50 company like talk about how naturally that happens >> I would say it happens it happens fairly naturally and like it has evolved over the course of that as we've worked together because I you know we were very fortunate to get into white combinator and after that we started white combinator I think with one gigawatt of an LOI And then we ended up, you know, with 20 gaws and and now we're kind of doing this transition of a go to market from like, okay, it's great to give us a non-binding LOI, but we're now looking at slotting actual projects, whether they be PPAs or selling our reactors as we scale up and actually slotting them into our manufacturing plan because we're now at a point where we're no longer, you know, having to show these these great LOIs, but, you know, we're at the point where we want to do some real projects and actually build and bring power to the grid. So, >> and is that the priority like for you guys for the next kind of six to 12 months like you know where is the majority of your time going to be spent?

20:11 >> Yeah. So we're we're we're doing technical work obviously as founders and also you know working with you know the government making sure that we get we're getting the right financing for these first of a kind nuclear projects because you know speaking of go to market no one wants to be the first. so you need to create a structure and a contract where actually the risk is shared actually books of risk is on on our side to deploy that first reactor and then we have an order book of many people actually are willing to sign contracts after the first of a kind is being deployed and we're talking about contracts before 2030. and the only way and authentic way to actually get these offtakers to sign and get investors to put money in these projects is to be credible. And we spent a lot of time you know with utilities. We have a utility advisory board that know how to operate pressurized water reactors that actually help us both on the design aspect but also in operating these plants. They know how to operate them at over 95% capacity factor. That's by the way higher than natural gas. So it's a very reliable asset. We want to keep that as as we're deploying nuclear. And then on top of that, we're partnering with construction companies.

21:21 We're not a construction company. We make their lives easier. we make the construction period much shorter, but we're not a construction company. There are people who do it very well. and so we focus really on our product and we build this thesis and go and tell it to people and say, "Hey, this is what we think is the future of the nuclear industry in this country." Because, you know, in the past six months, China connected 6 gawatt of electric power into their grids. That number is zero in the United States. So that needs to change. We need a road map. We need a credible product. And that's, you know, we're happy to sit and talk to people and tell them we have the right the right thesis.

21:56 >> And a French man might be here to fix it. >> Exactly. >> Well, Drew, thank you so much for joining us in NYC. I think my key takeaway is that even nuclear needs a village. >> Yeah, exactly. Right. >> Thanks for joining us. Thank you for having us. Pleasure. >> I'm Jim Allen here at the Cube Studio at the New York Stock Exchange. This NYC wires powering tomorrow. Thanks for watching.