Not Your Father’s Data Center - Data Centers and Energy: Addressing Growing Demand and Power Constraints

Episode Date: July 16, 2024

In this episode, Raymond Hawkins sits down with Ed Socia, North American Insight Director for datacenterHawk. With a background in community development and sustainability from the University of Vermont, Ed transitioned from managing renewable research platforms to working with top data center teams, including a pivotal role at CBRE.The conversation explores the transformative impact of the pandemic on the data industry, driving remote work and digital transformation. Ed discusses the increased demand for data centers, fueled by technologies like AI, and the resulting power constraints in emerging markets. The dialogue transitions into the complexities of power location planning, the potential of nuclear energy for data centers, and the hurdles posed by regulatory and community acceptance.Ed and Raymond highlight the roles of utilities and tech companies in addressing power demands, emphasizing transparent development and long lead times. They also touch on tracking competitor strategies and the importance of market intelligence.

It's helpful to know what people, what your competitors are doing today, but it's, I think it's even more important to know what they're doing next. We've been saying for a while, you know, hey, Dominion, Dominion's telling the story that we're running out of power. We've got some transmission lines to build, but we've also got to figure out the generation side. All right. Welcome to another edition of Not Your Father's Data Center. I am Raymond Hawkins, Chief Revenue Officer at Compass. And today, our friends at Data Center Hawk are joining us. And today, in the person of Ed, you're going to tell me how to pronounce it because I don't want to butcher it. Tell me how to say your last name properly, Ed. Yeah, it's Socia. Socia. All right.

That's right. Just like the catcher for the Mets. I hate it, but that's probably the best reference I can give. Or Mets and Dodgers. Didn't you play in the Dodgers too? Yeah. I think so. I know Mets for sure.

I grew up in a Yankees household. So yeah, you're definitely not. That's not ringing a bell for you. That was the comparison I always got. So you're right on the money. Insight director in North America for Data Center Hawks. You got to tell us what that means first, and then we're going to dig into who you are. Yeah, Raymond.

You know, so, I mean, I'm a storyteller, and a lot of my friends will hear that and laugh, and they'd say, you sure are. But, you know, it's basically my role is to- Especially about your fishing? Yeah, exactly. Yeah, yeah. I got it. But you know, it's kind of my role is to have these conversations with folks in the market and figure out what's going on and, you know,

take all these little sound bites and put them into this story that we can tell, you know, aside from the metrics and the charts and the graphs and the statistics and all that good stuff. But more important than the what is the why. Why are people going to certain markets? Power challenges are tough, but what does that mean? And why is this what everybody's speaking about now?

Yeah. Well, before we get into the data center business, let's learn a little bit about you. You do not sound like you're from South Texas. So get back up to where are you from and kind of weave a path that leads us to Rhett and David's door. How did you get to data center home? Yeah. So I grew up in Syracuse, New York. Go orange. Yeah. Big orange. Yeah. Bayheim and Earl the Pearl. Yep, that's right. See if I can make any other, yeah.

We had McDab back in the day. Lake Effect Snow. Lake Effect Snow. You guys are kind of known for that. Yep. So I grew up in Syracuse. Went to parochial school there with Catholic school. And after graduation, I went up to Burlington, Vermont.

Attended the University of Vermont. And I have a degree in community development with a focus on sustainability. You can't let a Catamount shout out go by. Yeah, yeah. Go Catamounts. Catamounts, that's right. I'm proud of my school. I'm athletically, I identify as an orangeman.

That's good. That's good. You're allowed to do that these days. But yeah, so I went to UVM with this idea in mind, just like sustainability and- And that's before orange man became a political term. Yeah. Right? Sure.

That's right. Some people don't know that's the Syracuse mascot, not a presidential candidate. Now they're just the orange because they had the lady orange when it sounded kind of funny, which my mascot in high school were the brothers because it was a Lasallian school. So the lady brothers sounded even funnier. Even a little clunkier. A little clunkier. Yeah, so I went up there and studied sustainability, which in the mid-2000s, you know, sustainability was like a hippie term.

And everybody's saying, you know, you just want to go around and be Johnny Appleseed and plant trees. And I said, no, no, this is really the future. We're moving in this direction. And cut to, whatever, 15 years later, sustainability. I'm seeing on LinkedIn all the time, people, sustainability officers, and the ESG buzz came around. So yeah, so lived in Burlington for a long time. And I graduated.

And I came out here where I am now to Denver, Colorado. All right. So one set of mountains to another. Yeah. Yeah. These mountains are a little bigger out here. Yeah. Bigger out there in Denver. Yeah. But you know, I basically came out, I just love Colorado and Syracuse has about 65 days of sunny air and 300 days of overcast and Colorado is the exact opposite. So. Yeah. So that was, that was, you know, that was a no brainer. And then I'm a big music fan. My girlfriend and I see a ton of live music. We're about 15 minutes from Red Rocks. It's a great place to be. acquired at the time, it was called Energy Acuity. And it was basically what we do at Data Center Hawk, the market intelligence side of the business, but it wasn't Data Center specific, it was power generation and delivery specific. Is that what took you to Denver,

going to work for Energy Acuity? No, I came to Denver with, I sold my Jeep, got about 2000 bucks in cash in my pocket and packed a backpack and headed out, left on some couches. This is, I think I was like 21 at the time. Yeah, yeah, yeah. It went back when all that was fun. Back when all that was fun. And, and I'm sure some of my friends would say I slept on their couches a little longer than they wanted me to, but, uh, I got an injured ship at that

company and kind of climbed my way up. And by the end of my, I was there for about eight or nine years. And at the end of that, I was managing the entire research platform for the renewables piece. So, you know, that was really fun. And I got my feet underneath me with, you know, basically just the American electric grid and different regulatory bodies like FERC versus, you know, different independent system operators and how it's a really complex system and the generation side and the delivery side, things like that, the policy side.

So then, you know, when I decided to kind of move, and it was a small company, but they were later acquired by a massive company called Inveris. It's a great, great company. You know, I kind of said, hey, what's the next step? And I had some close friends and some family members that had been in telco and data center space. And they were like, hey, this is a really interesting, really interesting time. So cut to, I apply, you know, VRE to be their director of research for North America.

And when I'm talking to my former fearless leader, Mr. Pat Lynch, you know, we're talking about, you know, one of your strengths. And I said, I have a proficiency in research and all this stuff. And I have this really, you know really robust understanding of the power markets and power policy and things like that. Candidly, I have some stuff to learn on data centers. Pat said, well, that's great, Ed. He said, you're surrounded by people who know data centers. What we don't have is people who know power. Ed, help me out.

What year is this? You're in Denver, which is where Pat is. You get connected to CBRE. What year are you and he first talking? So I'll tell you the exact date. And it was March 13, 2020, which was the first day. Oh, wow. It was the first day of the pandemic.

So I got an email after a critical job that I'd been at for the better part of a decade. Yeah. Hey, by the world, the world's going to shut down today. Great job. Good timing. Well, so I got a call or I got an email first. I was like, hey, we're going to the office.

I was supposed to start on Monday, which is the 16th of March. And I get an email like, hey, we're going to close the office on Monday. Just hold tight. We need to figure out this coronavirus thing. So I'm like, okay. Boy, there's the understatement of all time.

That day turned into, you know, and I'm working out of my Gmail. I'm working from like a little, I was in Capitol Hill. So there's riots going on outside of my window. It's like, you're a fast. It was a really, really, it's kind of a, it was a wild time. It was a crazy time. But interesting times. It was very interesting times.

And, you know, that week turned into a month. That month turned into like six months, that six months turned into over a year. And the whole time, basically, I'm just thinking like, what the hell is this? Is this the new normal? I guess a benefit of all of that is when everything else in our economy and all of the, you talk about REIT and you talk about, especially with commercial real estate, you talk about different verticals or different asset classes, data centers, people were like, man, you landed in it right at the right time because data centers were, you know, responding. I mean, everybody felt the squeeze, but data centers were really with the dispersed workforce, with, you know, people going to this sort of hybrid remote setup. The statistic I always used to throw out was

Zoom in beginning of March of 2020 went from like 30, something like 30 million calls a day. By the end of March, it was up to like 300 million. And then by the end of April, it was up to like over 400 million Zoom calls a day. And, you know, I'm learning about this, this digital infrastructure at the time and that everything has to traverse this physical fiber, you physical fiber to get from point A to point B. So while all this is going on, at least everyone's

patting me on the back going, man, you've really accepted it in a good way. So we're taking the sustainability researcher and we're teaching you about IP packets. Exactly. And we're saying this stuff bounces around and amazingly video packets are really, really big and lots and lots of work. And yeah, all right, boy, as if our industry, I mean, our industry had been growing for almost two decades at that point and it didn't need gasoline thrown on the fire, but boy, and that's one of the things I know the pandemic was a horrible experience for mankind and a horrible experience for our planet.

But man, it did have an unbelievable, I heard somebody fairly early, it was in that first summer. So June or July of 20, say that we've had three years of digital transformation in the last three months. And I think it forced organizations to go, hey, how do I do this differently? And I know you use the metric for Zoom, certainly changed the way, you know, everybody communicated and had meetings and where they sat. But that forced a lot of change on the back end, right? You alluded to networks, and of course, then there's got to be data centers and servers and all of that. So our industry got a huge boost from the need to be able to do things remotely. And then I've heard something lately about, I'm not sure I'm going to spell it right.

I think it's AI, something like that. I'm not 100% sure, but I hear it's doing something too to our industry. And GPUs, I think, I'm not sure, but seems to also be providing another accelerant to what was already a pretty well-performing market. So, all right. So you go work with Pat Lynch, who we think the world of at CBRE in early 2000. You're basically a researcher with a sustainability environmental background and you get dropped in the middle of the best data center team in the brokerage world, right? I mean, I think that's fair, right?

I would say it's fair. I think they're the biggest. Yeah. And you go, okay, on top of that, now everything's going to go gangbusters because we just lit a powder keg because everybody's got to do this thing from home. All right. We're caught up. Keep going. What goes on from 2000? Yeah. So- And 20, excuse me.

Yeah. So yeah, I kept going on CBRE and I still to this day, I love CBRE. I love the green machine. I grew a lot there and I have nothing but respect for that company. I dipped my toes in the world of brokerage for a little while and I realized that I really do like the research piece versus the sales piece. So I kind of said, hey, how do I downsize a little bit in terms of company? Because I'm a very social person. I like to know everybody that I work with and there's like 140,000, 150,000 people in CDRE.

I know a lot of people. I don't know that many people. It's a tall ass. It's a tall ass. I think about everybody could do it and me. But anyway, so, you know, then basically took a sabbatical for a couple months and got outside and was hiking the beautiful Front Range of Colorado. And basically, you know, reached out to Mike Netzer, who you've spoken with before at Data Center Hawk. Yeah, yeah.

And Mike kind of laughed, and he was like, Ed, I'm going to send you an email, or I'm going to have David send you an email. And it turned out, I was basically saying, hey, I took some downtime. I've grown my hat back in the ring. Just let me know if anything comes up. And unbeknownst to me, they had been looking for a director for North America. So I talked to them about it. We chatted.

It seemed like a great fit. That was back in, geez, August of last know, I talked to them about it. We chatted. It seemed like a great fit. That was back in, geez, August of last year, I think. So I'm still a little under a year at Data Center Hawk, but it's been a great, you know, fight around. And David and I really hit it off. He's just like Pat Renschen, and he's second to none when it comes to being a leader. Good human being. These guys in the leadership positions, they're lead by example, and both Pat and Dave did that. You know, there's no call they wouldn't get on. There's no, you know, they wouldn't send you into battle if they themselves weren't going in with you. And I don't know.

I find that to be a very admirable quality from a leadership standpoint. Yeah. Hear, hear. Well, I will just say at Compass, we are friends of the program. We're big Dave Liggett fans. We love him and have known him a long, long time. And the Liggett family, which we'll talk about another time. And thank the world of him as well. So, okay. So you meet Netser and he says, we're looking for somebody in North America. And then you meet Liggett and it sounds like it's a thing and you've been there a year. And give us, I know you gave us a little bit at the top, but insight director, what's the real mission? The real mission is just telling the story of what's going on and where the ball is moving. And it's helpful to know what people, what your competitors are doing today, but it's, I think it's even more important to know what they're doing next.

So when you can find these trends and you can kind of identify, hey, why, you know, why is everybody moving south down the I-95 corridor when Loudoun County's, you know, a massively, a core ash where it's massively developed. Again, flicking them on in terms of fiber connectivity

and then everything kind of sprawled over to Prince William County and Manassas. And, you know, you can look at land pricing out there. And we've been saying for a while, you know, hey, Dominion, Dominion's telling the story that we're running out of power.

We've got some transmission lines to build, but we've also got to figure out the generation side. So we're thinking, okay, well then there's NOVAC territory, there's all this other region. So when you start to look at this stuff and then you start to track the providers that are buying the parcels or they're doing their site selection and you start to say, okay, Bothellvania, Lockyer, Culpeper, and you see this trend that we had kind of been talking about for a while. You guys have that incredible data center in Red Oak.

We said the same thing in Dallas. Stuff's moving south in Dallas on the I-35 corridor, going towards Austin and alternatively coming up from Austin, Lancaster and Red Oak. And what Compass is doing is a great example of that. So yeah, just taking those bits, taking those bits and pieces and figuring out, hey, what does this mean? What's the story? And then I get on camera with gentlemen such as yourself and talk about it. And if you to put your chips on Lancaster, Texas in 2020 as being a data center market, that would have been some real insight. Hey, David might have done it.

He might have called out Lancaster. In fairness, yeah. South of town in Dallas certainly became a thing. If you go back five years, no one was talking about it. But three years ago, it really became a thing. And boy, now it's just gangbusters. So I know we can talk about South Dallas because it's headquarters for both of us. Let's talk a little bit about what you see in those markets that are going gangbusters. You mentioned going down Prince William County, down I-95.

With your energy background, what do you think is leading us to where the industry is going? You know, I know I joked about AI, but certainly it's changed the size of what we talk about with people. It's changed the pace of what we talk about. And I also think it's changed the location. So take any one of those three you want to take. I think the important thing to note is when we used to, and it was almost before I even got into the space, that people would talk about markets as tiers, a tier one market, a tier two market. I was sort of introduced to them in, through the lens of primaries, secondaries, tertiaries. What we've seen recently is sort of a,

I don't know if I'd call it a flight, but a migration from these, from these more established markets to essentially non-markets. think rural Ohio, you know, Southwestern Dusty, things like that. And it's all driven by that availability to acquire power sooner than later and then scale that power, you know, with your growing requirement. You know, a lot of these providers are sort of kind of hedging their bets and they're deploying multiple, you know, multiple markets at the same time. But, you know, if you need to be in a market, if you have a latency-sensitive requirement or your end user, I should say, has a latency-sensitive requirement, you need to be in that core market. But if you have, and if you want to go back to our alphabet,

our alphabet soup, if you want to go from the AI and even the pre-AI stuff, the machine learning stuff, it's not latency agnostic. Some of it is, and then some of it is latency sensitive, I guess I would say. So I think we're starting to see folks say, hey, while we're waiting for these will serve letters, while we're waiting for these load letters from utilities who, you know, are giving us a two, three, four year out, you know, estimated delivery date. Where can we go now?

How can we build for, respectively, or building for a client? How can we do that today in markets that we have power today? And do they have applications or do they have requirements that, you know, would would work in those more, I guess, remote locations? So I think, you know, again, we're seeing, I guess, remote locations. So I think, again, we're still seeing folks building in Loudoun County. We're still seeing folks building in Prince William County and all down that I-95 corridor. We're still seeing folks build in Hillsborough.

But all these markets are power constrained. And so while they're building in these markets, they're trying to say, hey, where's the power today? Yeah, and the where's the power today, to your point, you already talked a little bit about in Loudoun County, the issues there with Dominion. It ain't in those markets anymore, at least not at the scale people need it, right? And because of that, we don't want to call them tier ones and twos anymore because looking in Wyoming or Reno or Iowa or Alabama for hundreds of megawatts, that's no longer size-wise a tier

two thing, but you're going there because you can get power, right? And you see people in our industry doing that. It's surprising the where getting originally driven by network, and now more than anything, it feels like it's being driven by power. Yeah. And I would say, you know, I think people are a bit more or a bit less rather beholden to the energy or the fuel mixes of regions now. We used to see people go to areas because, hey, this place has, you know, a plethora of wind or, you know, maybe it's even large. Yeah. Hydro. Yeah, if you think hydro, Quebec up in Montreal, large hydro. Yeah. Yeah, if they think hydro Quebec up in Montreal, large hydro. But now, you know, if you're talking about a gigawatt, two gigawatt deployment, you're basically saying, hey, who's got two gigawatts to sell us?

And you're kind of saying, we might have to put this ESG thing, we might have to put these renewable portfolio standards. We might have to stretch that target date out a little farther than we had initially intended because of this AI thing, you're not going to find multi-gigawatt PPAs with straight wind energy in middle America unless... Yeah, not a thing. It's not a thing unless you're Bill Euro Windfarms, which I know AWS and companies like that have certainly done. Or they're saying, let's go behind the meter and look at the Foswahana deal

at the nuke facility. But that's obviously kind of a, I guess I'd call it a test case. It's a unique situation, but I do think it speaks to where the future is going, right? To be able to do the amount of energy we're going to need at scale and neutral in a carbon world. I personally think that's the direction we're headed, is that you're going to see more deals get done. We're going to have to fix the regulatory side of it because we can't wait 10 years to get a plant online. Yeah. And I think that's what it is. I think that's what it is about a decade out, I would say, the SMR, SMB, whatever you want to call them, the nuclear story. But I say

this all the time, the United States military and the United States government has hundreds of thousands of these things cruising around the oceans on submarines with no detriment to the service. They've been doing it a long time. So I think it's certainly where the market is going. I do think we're probably eight to 12 years out from that being a commonly adopted sort of power generation source. All right, Ed. So I got to challenge you on that one. Why do you think it's that far? Because we've seen Microsoft announce a project with it.

We've seen the AWS Susquehanna deal. So clearly the guys, the biggest guys who move our market are already making public announcements about it. I'd love to hear your take as an energy guy, why you still think it's that far. Aside from power availability, one of the biggest threats to data center development is the NIMBYs, right? And so a lot of these folks who they hear nuclear and they think Fukushima, they think Chernobyl, they think, you know, they just absolutely assume the worst. So they're going to push back on nuclear as a generating source. If you go to Europe, you know, they're huge on nukes.

I just think that, I think it will be adopted. And there could be a South Korean thing. You know, I mean, that facility was there. Microsoft has been talking about this for a while. When it comes to nukes as the preferred power source for, you know, future data centers. I just think the proliferation of that technology, it's going to require regulatory bodies, chime in on that. It's going to require a ton of capital to develop these things. And then you're going to have to, like I said, you're going to have to get the communities on board. So maybe it's going to be shorter than that. I would imagine, and from a

lot of my conversations, a decade is the sweet spot that I've been hearing. That doesn't mean that some will pop up before that, but I think in terms of being pretty readily adopted or widely adopted. Yeah, yeah. And when you qualify it with the go-to standard of power for a data center, yeah, that's probably right. Before it's the main street way we do things, it's probably a decade. I think we will see projects before that. To your point, Susquehanna was already there. I think we're going to see, this is one guy who has nowhere near the energy education you do. I think we're going to see strain on the generation side. I mean, we already see strain on the distribution

side, but I think we're going to see, and transmission side, excuse me, more than anything. I think we're going to see strain on the generation side across the energy portfolio, not just for data centers. We take big bites at one time, so we get noticed. And I think the data center and tech companies are going to go, hey, there's a great way to fix this. If you give us regulatory approval to do this, you know, two gigawatt generation thing with a nuclear reactor, hey, we don't need to take anything off the grid. I don't know how it's going to play out, but I can see that, right? When I think about who has enough money to lobby the regulators and the legislators to say, hey, I know it's 15 years from day one to commission a nuclear plant, but hey, I could take two gigawatts

off the grid if you give me a five-year plan. That's kind of how I think it gets done, is as a way to relieve pressure from the grid. And I totally agree. I think right now, today, it's both a generation and a transmission distribution issue, but the delays are more heavily on the transmission distribution. I mean, look at, again, the Dominican territory and that transmission line they're building out. But when you talk about, when you look at, again, the dominion territory and that transmission line they're building out. Right. But when you talk about, when you look at the future growth of data centers, and we were talking about it, 14 gigawatts in a year or something, or more than that, whatever

it is now, I mean, that's egregious as a majority of the kind of face you do, but that's an egregious amount of power. And I think the something like a 2.4, 2.5 gig offshore wind farm. And I was asked by a data center provider, is this going to solve our problems? And I was like, no, it's not even three gigs. Look at what's in the pipeline just for Northern Virginia. So they're going to have to figure that out. I think that they will. I guess Dominion has been thoughtful and they've done it well with, you know, coupling this up with the Loudoun County Economic Development Board, Buddy Riser in the late 90s,

mid 2000s. They really understood what it takes to- The great Buddy Riser, I have to add that. Because he'll listen and he'll call me on it if I don't say it. But, you know, they basically figured out, they said, hey, we have to be transparent with our development and what we're able to do and lead times and what it takes to get that car to litter. Now that these emerging markets or non-markets are starting to see these massive swaths of data center demand, I don't want to call utilities out, but some of the utilities in middle America are now saying, wait a second, they bit off more than they could chew. And so now they're back trapping a little bit saying, wait a second, we got to figure this out before we just say, instead of seeing just dollar signs, we say, hey, wait a minute. One of the issues we're seeing is if a hyposcaler puts out an RFP and then four third-party data center guys go to a utility with that FAMR-FP for

easy conversation, it's on 100 megawatts. It's not at four loads. That's right. It's shown as four or five hundred megawatts and thin of 100 megawatts. So understanding what that requirement will be in the future has been obfuscated. And I think there's work to do on that side, but I totally agree. I mean, the generation thing is going to be more and more important. All right, Ed.

Energy expertise, insight director. We love David Liggett. Rhett and Mike are good guys too, but we're kind of big David Liggett fans. You made a good call going there. Hard to leave Pat Lynch because he's a great guy too. I love it because it's all true all the time. Yeah, yeah. I love that you're still in the space with Pat.

And when I get to Denver, I'm looking in the space with, with, with Pat. And when I get to Denver, I'm looking you up. We'll go, we'll go hike up something. We'll bring a beer for the top and we'll talk about data centers after we've consumed said beer. Sounds good. Assuming I can make it to the top. Ed, thank you so much. We've enjoyed it and we will look forward to talking to you and other friends at Data Center Hawk again in the future. Thanks, man. It's been great.