Exploring 2025 U.S. Data Center Trends

[00:00:00] Mike Netzer: Mike Netzer: Welcome back. I'm going to say that to the Data Center Hawk podcast because it's been a hot minute since we've been in the chair. You have quite literally traveled around the world since the last time we talked. We're not going to talk about that today. We're going to talk about domestic issues. U.S. North America data center market today. Welcome. I'm Mike Netzer, joined as always by my good friend. I think it's worth mentioning that we are actually friends. We actually enjoy one another's company. It's not just a facade for the podcast. Data Center Hawk founder, David Liggett, what is up? Thanks for being in the chair. [00:00:36] David Liggett: David Liggett: Nothing. In fact, one time we were at a place outside of work and someone came up and said, "Y'all really hang out outside of work, too?" We were like, "Yeah." [00:00:47] Mike Netzer: Mike Netzer: Why wouldn't we? [00:00:48] David Liggett: David Liggett: Oh, great. Yes, I'm doing great, Mike. Thanks. [00:00:51] Mike Netzer: Mike Netzer: That's what friends do. They hang out. David Liggett: That's right. Glad to be back. Mike Netzer: Glad to have you back. Mike Netzer: Okay, so there's been a lot going on in the space, which we want to talk about, specifically around North America, U.S. That is the focus of today's podcast. But before we jump in, I just thought it's worth mentioning that if you are in your organization responsible for data center real estate strategy, that is exactly who we are built to serve around competitive analysis, supply and demand, big end user analysis, power infrastructure, fiber infrastructure. That is what we exist to do is help you both inform and accelerate your data center real estate strategy. So if you are not familiar with data center, check out datacentero.com for lots more info on all that is what we do. But today we're going to talk about some of the big things that have happened in the first half of 2025 we are in, specifically in the U.S. So there's a handful of, I think, big trends that you're seeing. I think we'd love to get kind of a download on each of those. So let's start with kind of the big one is, and it's something we've really probably talked about on this podcast for three or four years now is, you know, the shifting in the geographical shifts in demand and where it's going. So kind of the headline stat is, you know, we track core markets, secondary markets, even as far as far as to say, maybe tertiary markets. And then there's another group of markets beyond that, that have had little to no data center activity demand for, you know, historically. And so what we've seen is a growing amount of demand land kind of outside of those, those top three tiers of markets. So talk about where you're seeing that. What do you think is [00:02:23] David Liggett: driving that kind of the ins and outs of that? Yeah. The, the first thing I would say is, you know, just historically, our, the data center market in the U.S., you know, really centered around bigger cities in, you know, 10, 15 years ago. And so those cities typically were, you know, Dallas, Chicago, Phoenix, Northern Virginia, Northern California, like areas that had other businesses around them. I swear, every time we do a podcast, there's like a guy on a motorcycle and he goes back and forth on a podcast. It's, I mean, it's like our arch nemesis out there somewhere. Anyway, it's a price you pay for this beautiful view of 635. If you're listening, I hope you enjoyed that. So, but you know, basically that's where the, the space kind of started and you know, the shift this year, I think if you're like in the, you know, kind of this midpoint of the year, I think we have seen these larger projects go to more rural markets. And those have been nailed it. Yes. Thank you. And those have, and that's been a big shift in the space because we, we had never seen that before at the size that we're seeing today. I think the one takeaway here, you know, if you're, you're new to our spaces in the, in the second quarter, one trend that we took away was the, the growth in those markets actually outpaced the growth in the major data center markets. So I think Atlanta, Chicago, Phoenix, Dallas, those, those markets that I mentioned now that that's not a trend. We, I don't, I don't believe we'll see that moving forward. I mean, I really think, and, and I think you're seeing this in the way that people are, are buying land that most groups believe in the, the thesis that having sites closer in closer proximity to these larger markets in the long run is really the, the thing to do, but we're seeing major investment in areas that have had no data center growth in the past, probably we'll do one or two larger deals. And then that'll probably be it in those specific areas. Yeah. I think that, that speaks to two [00:04:28] Mike Netzer: things. One, we've seen large end users buy land developed for themselves in everywhere. That's right. Non-markets middle of nowhere. Yep. That's been happening for several years. Not so much on the leasing side. Yeah. And now we're starting to see them leasing in those far-flung markets, which again, speaks to one, probably some additional demand that can't be met by self-building to an [00:04:50] David Liggett: expansion of that buyer pool. Yeah. That, that I would say that's, that's certainly a big part of it. And, and two is like how much development you really need to do before, you know, you're speculatively building anything or, you know, so it's, it's a much different view. You know, if you have a site in Mississippi or, um, you know, North Dakota, how you would look at like spec development or even like the work, pre-work you're going to do versus a site in a Dallas or Chicago or Phoenix. So again, on the geography, the shift is continuing to happen in the U S and in North America. I think it will, um, you know, probably continue this year and into the next, but, but I, but I would just say that the watch that demand, um, in those markets, because it's really fascinating to see, you know, if you go buy land in a rural market and are planning on building something, you know, while the, while the group of people that can lease, it has grown is still a small number. And that's an important thing from an investment perspective to recognize as you're doing that. [00:05:52] Mike Netzer: Yeah. And if you forecast out to say like 2050, the entire country will be just one big data center, one big facility, it will be one market, one monolithic and one gigantic facility. I can't wait. Uh, okay. Second piece is like, you know, kind of the way demand has started and stopped. Oh yeah. And there's been a lot of announcements and maybe if you're kind of dialed into the space, you've heard rumors that different companies are ramping up their leasing or they're ramping it way down. And so talk about that. And what do you think is driving that? And what's the impact on the [00:06:25] David Liggett: space from that regard? Yeah. Um, I mean, demand is something we're all paying attention to and how these, uh, companies come in and out of the leasing space is really interesting to watch. You know, one thing I would say it's, it's really important to note is that there's a lot of new people in the data center space, new companies that have come in, in the last, you know, 12 to 18 months. It's a hard space to be in for a couple of years and, and, you know, get the return that you want and they get out of the market. I mean, those have been in it, that have been in it for a while that are going to be in it for a long, the long term, you know, don't get, um, don't maybe lose their minds when they hear a group has all of a sudden stopped or turned off or backed out of a deal, like an LOI or something like that. So I think there's a couple of things I would say here to this at kind of the midpoint of, of the year. One is the nomenclature around the starts and stops is really important to understand there's groups that are looking at 20 different sites. It doesn't mean they're leasing 20 different sites. It doesn't mean there's facilities that are going to be built on all of those sites. So I think understanding when somebody says, Hey, they, they backed out of a deal, you know, what really does that mean? Like the industry needs to be thoughtful in how we communicate that out. That's first thing. Um, you know, second is if 20, 2018 to 2020, we had a similar period like this in the space at a much smaller scale, but, but there were some similar themes. One was that we had a big ramp up in growth and then, um, and then it kind of cooled for a bit. And, and those that remember back then we, we called it kind of the digestion period of the cloud growth, and then it ramped up again. And so I think that, you know, there's some of this that I just go, Hey, there's groups that like take a lot of capacity and they need time to maneuver through some of the shifts that are happening and take taking of the space. So that's the second thing that I would say, but the third, during that period, there were about three or four users that were doing that work. Now that the user list is much larger, we talked about it. So there's a different, so when there starts and stops, the challenge the market faces now, and actually those users face is that there's other groups coming in and you can tell by the capital spend that the, the, uh, hyperscale companies are talking about publicly, how much they're spending on the space. Some of these, uh, AI groups that have come in, the GPU as a service groups that have come in and are doing certain things. Um, you can pull back from space, but different today than back then is those spaces potentially will not be there when you come back to the market. And back in 2018, 2020, those companies, I would believe always thought that like, Hey, I can pause this thing for a year or two and come back and it'll be there. That's not the case in today's data center market. So I think with the starts and stops, um, the minute we hear one group might be stopping, we hear two others are back in the market. And one anecdote I'll share is, you know, there's a group that paused in, I don't know, let's say March or April, you know, for a month or two, literally a month or two. And they come back in June and July and they're, they're already behind again. So that's the pace that these companies are working. And that's not a shot at anybody. It's like they are dealing with, uh, trying to scale a business and scale growth, uh, at complexities we've never seen in the space before. So it's a really challenging thing to do. Yeah. [00:09:35] Mike Netzer: I mean, again, you can do the math on to your point, the publicly announced spend and just rough translate that to what that demand, what that equates to in demand. It's massive. [00:09:44] David Liggett: Yeah. I don't, I mean, I don't know how it can't happen fast because it's the, the infrastructure it takes to, to get that money going or get those sites going at things like that. It's just, it's, it's too much. So it's, we're in a very unique time, very unique time. We've been saying that for [00:10:01] Mike Netzer: a few years too. And it really, it persists, uh, the dude abides. All right. Uh, related to that on the scale. And so I think for the number of years, it's been bigger, bigger, bigger, right? It was, 32 megawatts is great, huge deal. 72 bigger, a hundred. Wow. Yep. Now 500. We won't hear, we won't talk to anybody unless they've got at least 500 with a path to a gigawatt. Sure. I mean, I think we said that verbatim on this podcast probably in the last year, um, that started to change. Yeah. And there's been a, I mean, I think there is still demand for those large campuses, but there's a new evolving demand for maybe some smaller. So talk about that. Yeah. Um, I think that [00:10:40] David Liggett: the, the ambition to get capacity in, in one place, you know, in a market on one site, et cetera, you know, is, is the ambition to do that is different than reality. And I think we all would, would understand that. Um, but you know, the trends that we're seeing around, you know, growth, like the large growth is really, if you can get to those rural markets, it, it really needs to be, you know, 500 ish megawatts with additional phases beyond that, you know, for someone to like commit to that rural type growth. And there's probably five to 10 companies that will. And right now, maybe, maybe a few more. Um, I think one of the things, maybe one good takeaway for people listening would be on the scale side of things is because of some of the regulatory challenges, uh, states and different municipalities have put out in the market, you know, there's, there's some gaming of the system going on at like, can we stay beneath a certain number and not face the same type of scrutiny? We might relate it to permitting or money we have to put up. Um, and I think that's something to look at on like the lower end. So like in Texas, as an example, I think the numbers I'm saying are right. There's like a 75 megawatt number that if you're below that, there's less scrutiny. If you're above that, there's more scrutiny. So, you know, trust me, there will be people that gain that system. If there's one thing people hate, it's scrutiny. That's right. Or try to gain that system all day long. And so that's, that's one thing we're seeing there. But I think even like finding a portfolio or establishing a portfolio of areas that have 200, 300 megawatt sites, and that's it isn't necessarily a bad strategy today, because if you can reduce the time, I'm a lot of, a lot of the value created there is around the timeline of the delivery of the power, which again, it's not a, it's a pretty obvious point today, but it's worth saying. So that range at 200, 300 megawatt range is still pretty attractive for some of these companies. And I think that as you know, this next round of users matures, you know, they're going to mature from the 36 to 72 megawatt site into the 100, 200, 300 megawatt campus range. And so understanding the demand is not just all up at the one gigawatt range. There are some companies that are trying to understand that. But there's still this middle of the pack that is, you know, those 200 to 300 megawatt sites would be interesting. [00:13:03] Mike Netzer: It would be interesting. And probably worth mentioning that we still believe there is demand for the even smaller, the sites in that, in that 20 to 30 megawatt, that is still a bit of a moving target from, from both sides, from the buyer and the, and the seller in that regard. So that's probably another topic for another day. You mentioned alternative energy or just energy and talking about being a whole nother wrinkle. Once again, probably we do a whole podcast on that, but talk at a high level about what you're seeing from, again, alternative energy being something other than a grid connected site. Yeah. What are the different options being entertained? Do you see anything that's become [00:13:34] David Liggett: more common or less common? Yeah. Again, Mike and I will say, and it's obvious, but we're not energy experts. But a year and a half ago, I think I've shared this probably on the podcast before, but I was with somebody that finances energy projects. And I was just asking him about the data center space and the growth that we are seeing and, and, and the industry's kind of evolution around energy use. And the guy, the person I was speaking to was like, natural gas is the only way that you can accomplish the timeline side of that, you know, in the US with, you know, the growth. And so, and I think that's what we've seen most of the evolution be behind lately. I think there's, there's a number of other energy kind of pathways being, you know, talked about, thought through. I don't, you know, I think we think about like, can it work? Then there's also the financial component of that. Then there's also the, like, like social responsibility and community pushback. And so there's, there's a number of different things to consider. [00:14:43] Mike Netzer: There's another aspect which we'll talk about in a second. I will tease. Oh, I can't wait. That's a good old podcast tool, the tease. [00:14:48] David Liggett: Get it, get it. [00:14:49] Mike Netzer: It sounds like what you're saying. There are some other groups like, like solar or wind or hydrogen, but it sounds like Nat gas is still kind of the first option. First option would obviously be great connected. Yeah. Second option would be natural gas, whether that's a permanent or even bridging. [00:15:04] David Liggett: Yeah, I've seen, it seems, it seems to be that way. And I've even, I've even talked to some data center operators that are like, Hey, we're not, you know, we're going to let people go chase the alternative energy thing. We are going to double down on utility and we're going to find places where utility can be delivered and we are going to focus on that. And so I think, you know, you can see a bifurcation of some of the groups that are wanting to, to, to do that. But it certainly provides, it's, it's one of the reasons we've seen so many new entrants into the space from the energy market. Um, it's cause they're really coming at it going like, Hey, we know energy really well. And if we can set this up in a way that works for you longterm or in a bridging solution that allows you to get to utility, um, there's opportunities there. So I think it's really, you know, the natural gas, um, source has really been what we've seen really drive a lot of the interesting growth, but there's, there's other options being talked about too. [00:16:00] Mike Netzer: Yeah. Yeah. And one of the things to consider with, uh, with the natural gas specifically, and you brought this up, I don't know, a couple of weeks ago is the, the load profile of, of an AI deployment and the ramping up and down. We've talked about that before. It's different than like a normal cloud or data center load. And that doesn't really play well with the natural gas, um, you know, onsite generation solution. So that's, I mean, we mentioned that it's like, there's a design implication, but you, I think you'd mentioned some other design kind of changes or trends you're seeing kind of emerge. So let's talk about that. [00:16:33] David Liggett: Yeah. The, the, you know, the, the design implications, I think kind of stem towards, um, you know, just overall, like construction of the building and, and how many stories and what the actual, actual, like physical structure of the space is today. Um, you know, there's groups that are trying to scale up quickly with, you know, GPUs and, and that, you know, you've seen, I think Meta has been pretty open about some of the shifts that they've had on the, the, uh, build of their sites. Um, very different than what we've seen, you know, three, five, 10 years ago. Um, you know, I think the co-location provide the, the data center operators are still in, in the challenge of like building something that's going to appeal to, you know, this 10 to 20, um, and 10 to 20 end users that are doing things at large scale. So being able to like be thoughtful in the designs, um, related to how the space is together, how the size of the data holes, you know, obviously the density, how water is being introduced into the space through liquid cooling. Um, you know, we've even heard of some groups that have felt like they over provisioned for liquid cooling and are now thinking about how to, you know, build something that is more, uh, in locations that could be cloud. You could have a cloud function to, um, you know, network year traditionally is like air cooled. And so thinking of how to like consistently, um, cool that, I mean, there's just the design side of things, just really fascinating. And we're not, I feel like everything we're talking about, I'm telling people we're not, we're not, we're not energy experts. We're not architects, you know, but I think that like the, you know, the design piece is really fascinating to watch and in our space, you know, and, and, and AI has changed such, um, like the norm, the way that we built data center facilities for so long has shifted, you know, by a bunch and it's going to be fun to watch over the next, you know, three to five years. But that's that, that design, those design changes can impact things like, you know, how big of this, how big of a site do you need? Uh, and that, that will impact where it's going to be based on power. So there's just, even the design implications have fascinating impacts, you know, beyond just like the size of [00:18:48] Mike Netzer: the building and how much power can go there. The building data centers is hard. Why would anybody? I don't know. It was quite a masochist. It's really, really hard and it's getting harder. Uh, but no, this is great. Uh, anything else? I mean, that was great kind of synopsis kind of at the [00:19:02] David Liggett: half year point of the year. No, so like, so, so again, the, we talked about what geography, we talked about, uh, scale design, kind of demand starts and stops, demand starts and stops and alternative energy. I mean, those are five really good things to kind of build a framework around, you know, and just do check-ins. And so that's why we wanted to provide this. I mean, some of them change dramatically. Some of them don't change that much, [00:19:23] Mike Netzer: but until they do, you know what I mean? It's like, some of these things seem to be a relatively straight line trend and all of a sudden it's like, Oh, left turn. Yeah. So it's a, that that's, [00:19:32] David Liggett: that's part of it. But I mean, I think as we continue to, you know, we, I think people probably know this, we digest a bunch of information, uh, through a number of different platform or tools, and then we analyze it and put it on our platform. And so these are just some of the findings from that. But if you want to understand the space, I mean, in the U S and why we did this in the U S is like, this is where the large AI growth is happening right now. Now it's certainly things are happening in Europe, in APAC, in some ways in Latin America, but it's, it is, it is front and center right now in the U S and so this is where to pay attention if you're wanting to be in the game from an AI [00:20:09] Mike Netzer: perspective. Awesome. Well, great thoughts, David. Thanks again for being on the podcast with me, your host, NMC, uh, rural. Good one. You nailed it three times this podcast. I was very impressed. Not a slur or anything about it. That's right. That's uh, anyways. All right. Hope this was helpful to you, the listener. Anyways, we'll catch you next time.