All right, it's time now for our daily Wall Street Week conversation, and the rise of AI is posing challenges to data centers. Applied Digital, which designs, develops, and operates data centers, is looking at how to adapt its infrastructure. Wes coom and See is the company's co founder, chairman, and CEO. He joins us now alongside Wall Street Week's David Weston and data centers. I mean, it's one of the hottest toppings out there, it really is.
Every day we're talking about data centers somehow, some way. But we thanks so much for joining us because normally when Katie and I talk with people who at data centers, it tends to be the commercial real estate side of it and investig in it. But you actually developed these things and actually install and operate them. So give us a sense of what's the big deal with data centers now that didn't exist before the rush to AI. We've had them around for some time, right.
Yeah, David, I think that the key to understand, the primary thing to understand is the difference in the compute workload that's going into data.
Centers now versus what it was in the past.
So in the past, you know, we were building data centers that were in population centers that we're providing ultralo latency communications so that.
We could get YouTube, so that we could get Netflix.
So we could get TikTok, right, so we can get all of our things that have been driven mostly by video applications.
Now, as we're moving.
From applications that need ultralo latency communications outside the data center to applications that need massive amounts of compute. So think of, you know, doing large math problems algorithms. This is what AI is doing inside the data center. Compute equals power, and then inside the data centers you need a lot of compute close together, so you need really
high power density. And so these are kind of some weird numbers, but you know, historically data centers were let's call it seven kilowatts for an entire rack of computing, and Nvidia a one hundred server one by itself takes ten ten kilo wants. So and then you need to stack all of these together, so you have a massive amount of power consumption that needs to be close together.
The latency inside the data center now versus outside, needs to be ultralove, and so you have to build a completely different style of data center, you need large amounts of power in the same location, and so that's what we're doing. We find large amounts of power, typically different locations than you would do city center.
So it's no.
Longer New York or Virginia or Dallas or Los Angeles. We're building in different locations where we use a lot of renewable energy that have these hundreds of megawatts, if not gigawat of power so that you can run these workloads because they're just very different style workloads versus what we're used to.
Right and when it comes to that adaption process really sort of converting those existing data centers to be able to handle the AI workload that people are trying to do. I mean, how long the process is that? How long is that timeline?
So I'm a big believer that you're better doing green Field in that you just build an entirely new style of data center versus trying to retrofit.
Cooling is different.
Again, the power density is significantly different, so there's there's a big issue.
So the number one issue right now is finding power. Where's near term power? We have we have a very.
Unique way of finding power, which is we find things that are called stranded power, and we typically use a lot of renewable energy that's more in the middle of the country.
So number one is power, and then number two is supply chain.
So this is electrical components, it's just you know, it's transformers, it's chillers, it's high voltage switch here. So the supply chain because of the demand for this has become really stretched as well.
So number one is power. Number two supply chain.
So wes are you building your own power? I mean you're building those wind farms or solar event or even nuclear for that matter, or are you tapping into some other part of the grid. And similarly on the supply chains, how long is the lag time on all this? It sounds like it's going to take an awful lot of time to build that power and also come up with all that supply chain.
So right now we're not building the power that's something for the future for us. So there's there's a unique setup in North America with that we take advantage of what I call it is stranded power.
So there is power that is generated and not used.
And I know a lot of people would think that if you can generate an electron anywhere in the US, it can be used anywhere else.
In the US, and that's just not the case.
You need the transmission infrastructure to actually be able to do that. So we find places where, you know, for Ellendale, North Dakota, for example, there's a massive amount of wind power that feeds into a substation, and oftentimes they have to curtail the wind farms because there's not enough demand, there's not enough load to use the power that's generated.
So we go to that location.
And we have several of these locations where we use power that's currently not being used now to go to the other part of your question is for new power, so you can build power gen on renewables over you know, let's call it a twenty four maybe thirty six month time frame.
The bigger issue is transmission.
So transmission to transmit that power that typically is built somewhere in the middle of the country to end users. You know, that's more of a tend thirteen year process, and so we're kind of shortcutting that by taking these workloads directly to the point of generation.
I want to talk about the competitive landscape here because you think about AI and it's just really made its way into every part of the market. It seems like every investor in every asset class is trying to figure out what's the AI narrative in their industry. When you think about your own industry and data centers right now, how fierce is the competition.
So it's an interesting time for sure.
So if you go back a year ago, everyone is trying to figure out on the data center side, what's happening in this market right Why is all of this demand?
Why for the high power density?
Now in this industry figured it out, and now it's trying to make this solution because again, getting the amount of power that's needed and operating the style of data centers is extremely difficult. And so think of on the data center side, we're the picks and shovels. This is digital infrastructure. This is the digital infrastructure was built out to run everything that we're you know, everyone's used to running on their phone or on their TV at home,
and now it has to be rebuilt. It's not a replacement, by the way, this is an add on, but rebuilt for compute and AI workloads. So we're just we're at
the building blocks. We're the infrastructure that runs all the AI that you know, everyone wants to do and we talked about you know, everyone knows chat GBT, but if you're doing anything in in uh, you know, in video and visual anything generative AI runs on this infrastructure and a massive amount of this infrastructure used to be built, you know, and when Jensen and in Video talks about
AI factories. So so what we've done we actually designed our data center and you know, in partnership with in Video, we work with their team to build the style of data center that we're doing because it's this is the issue in the data center industry is it's just a very different style from data centers that were done in the past. But we're running all AI workloads and you
know these are really split between training and inference. So the terms that are used in AI, but the entire industry is i would call it scrambling to build the amount of infrastructure that's needed to house the compute to run these applications.
Well, it wasn't that long that we hadn't even heard about all these data centers. What's going to prevent absolutcence?
So that's a great question.
So when we build our facilities, we think about what we call future proofing. Right, so we think the power density goes higher cooling, but I think you know, the last twenty thirty years were again really driven by applications that required call comms, high speed communications, and as we go for the next twenty or thirty years, applications are really going to be driven by compute and that's what AI is, and so we're working to future proof that.
And by the way, the older style data centers, those are not obsolete. We still need all of that capacity for the applications that we continue to use. This is absolutely a complete add on to what's been done in the past and so but trying to not be obsolete is making sure that your future proofing your facilities as much as possible. With power density, with cooling, we're trying to run an extremely high efficiency so that we lower our carbon footprint. You know, we build where there are
renewable we mostly use renewable resources. But those are the things you try to do obviously, you know, not being obsolete in the world is one of the keys here.
All right, future proofing, that's a good place to leave it. Really enjoyed this conversation our thanks to wes Coomins. He is Applied Digital co founder, chairman and CEO