AI's Hidden Power Problem: Why Data Centers Need Different Batteries | Electrovaya CEO Raj DasGupta

Appreciate it. Thanks, Wes. Great to be here today.

Well, great question, Wes. So, yeah, batteries are in my DNA, both from a genetic perspective. My father started Electrovaya, but also in terms of how I've spent my life. So I completed a PhD in material science, which was battery-related, joining Electrovaya afterwards, and really have kept, you know, I've been with the company a long time, over nearly two decades. And R&D has been a key focus of mine personally, and reflected in the company, and that's how we operate, because this is a technology space which continues to evolve, continues to improve, and you have to stay with it if you're going to remain relevant.

Well, the big piece of it is if you're just looking at it in a material science lab, you don't see the big picture. You miss out on how batteries are used. That's a big part of it. Sometimes you're focused on just certain metrics like energy density, certain things that the battery researchers focus on, which is what you see in some of the industry where they're just focusing on maybe a solid-state battery, but they miss out. Does this battery need to last a long time? Is it got to be safe? How is it going to be cost competitive? So when you combine the science, the engineering, and most importantly the customer, the market-driven aspect, that's where everything comes together and where you really have to have your product developed for. And I think that's where it's taken us some time, but Electrovaya has hit its stride in that combination of factors.

Yeah, that's a great point there you had on the large markets not necessarily being the right market, and that sort of reflected in our strategy as well. So if you just look back, you know, 15 years ago, Electrovaya was focused on electric vehicles foremost, and we were going the right direction there. We had a partnership with Chrysler. A little while later, Chrysler partnership with Mercedes. That said, on the energy storage as well, we were one of the early players for lithium-ion battery-powered energy storage systems, which of course today are ubiquitous in a massive industry. What ended up happening in the space, though, was scale came behind certain players, especially large Asian, Chinese, and Korean battery manufacturers. And that scale, that almost government-level push that they had behind them, was a factor which really no North American battery companies could compete with. There have, of course, been a lot of high-profile battery companies in the US that have gone under, especially during that period from 2012 to let's say 2020. And what we did is we pivoted to stay relevant and stay in the space. And that was by focusing on technology and technology that served specific niche markets that we identified, rather than just going wholeheartedly against the bulk. And the bulk, you know, to be fair, these large Asian players have enabled great things to happen, right? The electric vehicles cost way less than they would have if they hadn't brought that level of scale. But from our perspective, I think we've done a terrific job of coming out with new battery technologies that serve cutting-edge markets. And these cutting-edge markets are now at the cusp of scaling. And so we are able now at this point in 2026, 2027 to ride that wave.

Yeah, that's what's happened in certain segments, and batteries are lithium-ion batteries themselves. I often refer to it as like Baskin-Robbins, right? There's so many different flavors, and of course there's so many different applications that require batteries. So the lithium-ion battery that goes in your phone has a very different requirement list than the one that goes into your low-cost electric vehicle or your high-cost electric vehicle or your robot, forklift, or energy storage device. And so what we did is we focused on the applications where it was not driven by costs, and the cost is often a cost per kilowatt-hour. End of the day, everything is cost. All decisions get based on some level of cost, but cost of life cycle, right? So a battery that goes into a robot or forklift that has to be charged three times a day, operates 24/7, operates inside buildings, and is very sensitive to safety and needs to be the same all the time. If you put a commodity-driven battery, which maybe conks out after two years, maybe that's a higher-priced option for that type of application. So end of the day, I do think all decisions really do come down to return on investment, but the ones we focused on are return on investment over time as opposed to a capital cost, initial capital cost, which is what I'd say consumers are more focused on as opposed to industrial applications.

Yeah. If you just read the press that covers batteries, you know, they'll focus on things like the cost curve, you know, batteries dollar per kilowatt is coming down and where will it be next year, where will it be the year after. And that's sort of the ride to the bottom and very low-margin business, not a technology-driven space at all. And so what we've just focused on the value add where we can add value through technology, and it all comes down to, you know, there's a point in science and technology, of course, that development, but more importantly putting the right pieces together to come out with a holistic solution to solve a problem, right? And I'd say that's common in any business, one where we've been fairly successful at, right? So we had a bunch of different battery technologies that we had, and maybe on their own weren't anything magic, but when we put them together, that really came out. And of course, what we're known for today, a lot of it when people hear Electrovaya, they think of ceramic separator technologies. And of course, that's a critical ingredient, but it's really the combination of a bunch of technologies which enables our cells to be really quite different and quite a bit higher performance than what you see typical lithium-ion battery from typical battery companies.

Yeah, I think our take, it's the slow and steady, and it's not slow by any means. We're growing reasonably quickly, but we've taken a more conservative tack to how we grow the business as opposed to losing a ton of money and building assets and hoping customers come or selling products at an introductory price. We do not do that. We have, I would say, a premium battery technology, premium product, and it's sold at a relatively high price versus typical lithium-ion batteries that has enabled us to get good margins, of course, and make and break even, which is atypical in this industry. The typical players, especially early-stage or small-scale battery companies, lose tons of money, and that's not something we want to do, nor were we in a position to do that. So I think this conservative focus in our approach has benefited us up until now and really puts us in a position to sustain faster levels of growth in a better way going forward.

Yeah, that's a great question, Wes, and something I've heard from other people asking this type of question, right? So, just a few years ago, investors would ask us, you have this great battery technology. It's very safe. It's obviously getting used in all these world-class applications. Why aren't you looking at energy storage? And energy storage over the last several years has become a huge space, huge industry, which has enabled great things to happen, like renewables are becoming more useful. You've seen the penetration of solar especially gain substantial traction and become a really significant force for climate change mitigation. But the batteries that go to support these types of activities, their number one job has been to store energy and store it cheaply. Right? So dollars per kilowatt-hour has been the most important metric, and battery companies, especially Asian ones, have done a terrific job of bringing that down. And the way they've done that is they focused on lithium iron phosphate as a chemistry, which is huge in China. They focused on making the cells that these storage solutions are made from really, really large. So your typical cell phone battery might be the size of your iPhone, and the cells that go into these energy storage applications might be the size of a large laptop, right? And they're huge. They're 500 amp-hours, and they're getting bigger. They're very good at storing energy at a low cost. What they're not good at is a number of things. One, they're not as safe necessarily as the technology we provide. They don't charge and discharge rapidly like ours does. And then the cycle life isn't anywhere close either. So the way we looked at it is, hey, we don't want to compete on dollars per kilowatt-hour. We'll never beat the Chinese at that. But if you look outside of renewables, if you look at data centers in particular, and that's what is topical today, a data center is sort of like a large vehicle. It uses a lot of... think of it like a car, right? Think of the data center as a car. And it's going up and down a hilly road. And so its power demand goes up as it's climbing the hills. It comes down sharply when it comes down, and it's accelerating, decelerating all day long, and it's running all day long. It's like doing a 24-hour Le Mans race 365 days a year. So that's your data center. How is that powered? It's powered through grid connections. It's powered by on-site generators and turbines. Everyone's throwing everything at these things, right? And the batteries are being used to stabilize the power, right? So because the generators can't ramp up and ramp down quickly, they need backup. And so they're putting these energy storage devices at these sites which have been designed for, you know, like I said earlier, using these cheap cells which are designed to be low cost per kilowatt-hour. And I looked at that and I said, "Hey, there is an opportunity here, right? This space is like a hybrid vehicle. It needs a battery which can charge rapidly, discharge rapidly, do it 24/7, many cycles a day. That battery can't be parked at the parking lot away from the car. It's got to be in the car as close to the engine as possible. So you need a safe battery." And so it ticked all the boxes: safety, cycle life, power density. And we said, "Hey, we've got all these things in spades. Let's go after it." And so this has been our biggest development investment push to date for a product in our company history. We're very excited. I think what's come out is this EolaPulse 1500, which is what we called it, and it's an extremely power-dense energy storage system. So effectively, it's like your hybrid car battery for a data center.

Yeah, it uses an incredible amount of energy, that's I think well understood. The pattern on how it uses it, though, is less well understood. Right. So if you look at your lights, right, they're on during the day or in office hours and then they're off at night, and your home is predictable as well, right? And the consumption of electricity in typical industrial sites is relatively constant and somewhat predictable. In a data center, the magnitude is higher, right? The amount of energy used in these buildings is for the density, for the size of those facilities, is really, really high. But the consumption varies dramatically. It'll have huge power swings. From a microsecond perspective, it's very noisy. And if you look at it on a higher resolution, larger resolution, it can have 40% power swings, right? So it is, you know, my analogy to a car is not too dissimilar, right? Your cars, when you're driving it on cruise control on the highway, yeah, your energy consumption is pretty flat. But if you're in the city and you're in stop-go traffic, that's more similar to how these buildings are being operated. And that's where the hybridization really can have actually a meaningful benefit, both in reducing electricity or peak electricity consumption. And I don't want to get too into the weeds here, but the way utilities price things is not, you know, they price things on your amount of energy you've used, but a large portion of their charges are demand charges, right? And that's linked to the peaks, right? Because you have to build out infrastructure to the maximum power consumed. And I think our solution really will be able to have a significant effect on reducing that.

Yeah, it's going to be evolving, right? So more and more attention is actually going to be placed on microsecond-level response times. That's something we're working towards on a new battery product, which is for their next-generation data centers, which you might hear this term 800V DC being used, and that will be more DC-connected batteries, right? So it'll be able to respond at really, really, you know, 20 microseconds and less. Now, when I look at the power level, power requirements overall, when you're looking to hybridize turbines or generators and things like that, which is where our product also has a huge impact. Yeah, the time scale there are really seconds to 30, 40 minutes, right? That's where I think our technology will come in and play a huge role. Obviously, there's this is a sector which is going to need a broad range of solutions, and we're certainly not the only one, right? So there are opportunities for sure for longer-duration energy storage, but energy storage, and this is where some folks get it a little bit misconstrued. You know, a battery doesn't generate electrons. It just stores them, right? Ultimately, more energy needs to be generated to support these types of assets.

Yeah, it came down to we looked at the current state-of-the-art being offered in this space and typical storage solutions that they look like 20-foot containerized boxes, and typically these range about 3 megawatt-hours to maybe just over 5 megawatt-hours per unit. And they're typically designed to be charged and discharged relatively slowly, right? So 2 hours is really the highest level that they operate at. So when we looked at that and said, "Okay, that's the state-of-the-art." Now on a cost basis, on a cost per kilowatt-hour, as I said, we won't be able to match these guys, but because we've got these ceramic separators, we're making the cells in lower volumes, we're making them in the US. So on a cost basis, we're not necessarily going to be directly competitive. But when you look at it on a power basis, right? So we said, hey, let's run this system much faster. So 3 megawatt-hours per block, we're the lower end of the average there, but on a power basis, we'll be 7 megawatts where the typical solutions are more like 2 or maybe 2 to 3. And so that huge difference there enables us to go to customers and say, hey, you need energy storage to deal with these peaks and troughs, but you can use half the number of batteries if you use ElectrifiPulse. And that's where the compelling argument comes in.

Yeah, the other solutions are good solutions to tackle the same problem. Though what you'll end up having is you'll have to have a lot more of these conventional batteries to deal with the same, to come up with the same solution. Right? So you'll need a lot more space, and the life cycle of the conventional solutions may not be the same either. Right? So you'll have to replace these units more frequently than an Electrifi product. Ultimately, we would provide a, you know, that's how we provide these end customers a value, right? Well, you need less of them per unit. Yeah, they're more expensive, but you need half of them. So overall cost is lower. And then you need less land, you need less space. And it's a proven technology, right? Electrifi's battery technology, this Infinity technology, is used in lots and lots of warehouses already. Amazon is a key partner of ours, and it's an extremely safe battery technology. We've made, you know, 35,000-plus systems with this technology with a perfect safety record. And of course, people also know data centers are expensive. They're often located in urban areas. Safety of these energy devices is critical, too.