Ep4: Data Center Design Evolution

This is uptime now, a Northern Rose Fulbright podcast dedicated to conversations about data centers from development and construction to cutting-edge technology and operational breakthroughs and the people driving it all. Joining us today is Bill Winsoninsky, veteran and the CEO of CWC Consultants, a premier data center consultant firm with 30 years of expertise in site assessment, development and economics and business operations for hyperscale, multi-tenant, built-to-suit and edge data centers. Bill, welcome to the show. Oh, thank you. Thanks for having me. Great to see you again. Likewise. I'd love to start by just talking a little bit about how this all began. You've been in the data center space for, I think as long as anybody, curious, curious how it all started and how you ended up as the CEO of CWC. Yeah, I think about that all the time really. And lately, I've been touring a lot of old data centers that I was a part of and it's really kind of like this is your life type of thing. I was in the Navy and then after the Navy, I went to Berkeley and then after Berkeley, I worked in advertising and advertising as a media planner in San Francisco pulling down 19,000 a year, really killing it in the Bay Area. All the magazines used to come call on media planners, right? So this is why our magazine is better because internet advertising wasn't a thing. It was like, you know, television, radio, print ads, those type of things. And magazines were like information week magazine, network computing magazine, tele.com magazine. And this is the early, not recent. So they were saying, hey, you come to this side of the, you come to this side of the, the, the, the desk. This is where, this is where all the money happens. And I was like, okay, well, it's kind of interesting. I like the magazine. It's kind of cool. So I interviewed. I got the job. And at the time, I was working for information week magazine and network computing magazine in the Bay Area during the.com Booms. And so it was a great education because we got to run with all the publishers of these publications that would teach us and tell us about all the stories they worked on and gave us inside scoop as to what was really going on and what was driving distributed computing. So we were going from mainframe compute to distributed compute and all around.com area. And, you know, back then it wasn't co-location. There was application service providers and hosting companies and those things. So I was fascinated by it. And so we did that for a while. And then started working for a distributor sold data center products. And then I moved to Chicago in 2001. And I started working with my father-in-law who had a company doing barcoding labeling, industrial marking, but they sold a lot of the physical infrastructure components to support barcoding and labeling. And we developed a business plan to sell data center, distribute data center products in the early days as IT and, you know, enterprise companies started to build out their data centers. And so as technology became more pervasive in the enterprise and they started building out these data centers for the enterprise, we sold all the major data centers in Chicago and racks and servers and PDU strips and all those things. And we became one of APC's largest distributor, which later became Schneider Electric. You know, we would go in there and they would say, hey, do you do fire suppression? Do you do structure cabling? Do you do flooring? Do you do this? And we're like, yes, we do. And we figured out how to do it and built the teams around it. Very cool. And follow-up is hired. And so I started working for the Colo because at that time, we would retrofit current data centers, but we would also help people understand their options around Colocation. So third party data centers were starting to pick up in this market, like Equinix and digital and lattices and those type of places. And so we would actually broker those deals and just drop ship gear to that location. And then I started to meet all of them and I started working for companies for the ColoCation companies. And that culminated my last company I worked at before I started CWC was QTS data centers. And we built, you know, we built the Chicago Land Data Center at 2,800 South Ashland. And I ran the team there and it was fantastic. And then right around when they were doing the shift into with BlackRock, I felt it was a good time to start my own company. And I haven't looked back. So it's been about six, seven years since I started CWC. And yeah, we're doing site selection helping customers understand what they have, help people get into the data center industry, help them understand their options and how to commercialize what they have. And then also help landowners and developers with their due diligence and help market it to the asset class based on the designs and based on the land and the power that they can get at. So that's kind of where I spend my time. Very cool. Yeah. Talk a little bit about how, because I mean, you can do this for a while now and data centers have changed and their size and scope. And I'm sure because of that site selection has changed. Talk a little bit about how that's changed over the years. It used to be fiber. It used to be the long pull in the tent, fiber optic cable. And the power that the data center power is just gone through the roof. You know, the heavy power that that sites are looking for. Hyper scale and Neo clouds are looking for pretty large campuses, you know, and that's starting to change a little bit as they're as they're as we shifted more into inference. But you know, the site selection was back in the day was existing sites that you can take over on a sale lease back or it was if it was a greenfield build, you know, the standard 10 megawatt or 10 megawatts was a lot of power at that time, you know, and or five megawatts even in that that wasn't too difficult to get at. But the difficulty was finding diverse routes of connectivity to connect to build a multi tenant data center and have a carrier neutral facility that can facilitate multiple carriers. And so that's, you know, the 350 sur max of the world, the 427 south of south, those type of buildings that became a hub. And then we would all try to find ways to get back to those places, get back to the backbone of the internet. But now it's like, you know, do you have you have grid power? Oh, you don't have grid power? How are we going to get power in the interim? So we're talking natural gas, which is a very new entrance into our world, generally speaking, two or three years ago, if you told someone we were going to do a behind the meter natural gas power generation on your campus, they'd look at you kind of sideways. But those two worlds have really come together in a very rapid pace. And, you know, there's a lot of education around that too around firm lengths and PPAs and who's going to build the power plan and how do you interact with the utility and what of those will serve letters look like and how that how you how you actually get at a ramp schedule or a, you know, confirmation that you can actually build something on that campus. And this is when the power is going to go live. And here's the ramp to scale. So very complicated, a lot of chicken in the egg scenarios where, you know, the utility needs a design. You can't have a design unless you have a tenant. The tenant wants to be able to see some pathway to power or will serve letters. So there's that, that angle that's causing a lot of frustration and fatigue in the market. And there's a lot of speculators in the market, you know, folks that were in commercial real estate or speculative or in light and industrial or industrial. And that that market is kind of softened over the years. So you have really, really good real estate professionals who have land that read in the paper or through their their industry say, oh, data centers is the place to be. And so we get calls a lot saying, I've got this land would it be good for a data center and, you know, maybe it would, maybe it wouldn't. But a lot of times there's there's a education process that needs to go into it, you know, around that site selection process. Yeah. So talk a little bit more about that. Like how, how does, how do people come and work with you? How do you guys come in and help solve some of these issues? Yeah. Well, since I've been in the data center for 200 years, people start to know you and get to know your name. And so I get a lot of it say, like, oh, data centers call bill. So I get a lot of calls that way. But we help folks understand what they have and help them with a quick know a lot of times, you know, just saying, look, it's better suited for something else. This is very capital intensive. And there's a lot of capital you need to put up front and now with with the ability to get load letters or to ability to get studies done, those studies cost a pretty penny. And that kind of weeds out a lot of the speculation, you know, you help them understand the big picture, understand what people are looking to build. And the industry is changing on a regular basis, you know, so every week it seems like there's a new thing or a new design or something. So we try to keep up through AFCOM through a seven by 24, through all the, you know, data center knowledge and just really connecting with a lot of our peers to understand what these neoclaus and what the hyperscalers are designing to and solving for. Yeah, it doesn't require a lot of integration. And you you seem to know like everybody in the industry, but talk about AFCOM, what they do, how that's had an impact. Yeah. So I started with the trade organizations early days. And I just wanted to get involved. I wanted to meet people and I wanted to meet people in my area. And so I started going to these meetings 25 years ago. And when I first started going, I had my list of people I wanted to talk to and I rushed up to them and gave them my elevator pitch. And I didn't know really what to do with my hands at these places, they know I was really kind of an awkward guy in the room and I just kept going and then you started meeting people and I quickly realized that everyone there is really kind, thoughtful, willing to help everybody in the industry. That's where I really fell in love with the industry, actually, and that's why I still stay in it on AFCOM side. I'm president of AFCOM in Chicago now. The awkward guy at the buffet line who was, you know, oh, you're Michael Bordenair, you're, you know, name the OG in the data center that I had on my list. And I would, you know, now I'm trying to help out the guy at the buffet table that doesn't talk to anybody. (laughing) - It looks nervous. - Yeah, it looks nervous, you know? And we have a huge, huge push in AFCOM Chicago to develop the younger professionals. And so we're starting a young professionals group when we've got a really dynamic team that's starting that and these guys and gals are like, business plan and ready to go and there's a lot of energy with the youth and then I'm stoked to see it. - Yeah, so talk about that bit because we found this quote that you mentioned one of your former mentors said, it says, "It never burned bridges. This is a very small industry. Treat people kindly in with respect. You keep the person on your left's cup full. Yours will always be full." So that's all very true, I think, but also it's a very competitive industry. Can you talk about how you kind of balance both things? - Yeah, no, that's, that philosophy becomes real when you separate relationships from transactions. You know, I mean, it's like in negotiations, I try to be at least transparent about any sort of constraints or expectations, even when the answer is not now or no. With vendors, that means really fair scopes of work, realistic timelines, and at the end of the day, honoring commitments, even when projects pause or shift. So during the delays or conflicts or issues, I try to focus on our shared outcome and not really focus on any sort of blame and try to create an atmosphere where we're rolling up our sleeves. So we can protect our organization's interest while treating people with respect, by documenting things and communicating early and keeping dialogue open. In this industry, tough negotiation partner is tomorrow's critical ally, right? So like, you know what I'm saying? Like when you're talking to somebody and they're rough on you or you kind of look sideways or you think they'll have that person's not or the pretty hardcore next project, you can be, you know, right next to them, trade and paint with them. So long-term trust consistently outforms short-term wins in my opinion, so. - Yeah, I couldn't argue more. I think we see it, I guess, in my world and the legal world, sometimes with either counterparties or counsel on the other side that are just very, and not difficult in a way that's like, you know, I'm protecting my client's interest, but difficult in a way that's, you know, crossing the line a little bit. And then you always wonder, it's like, you know, this is a very small world. We're very likely to meet each other again. And like, wouldn't you rather do that and wouldn't you rather know that when my client asks me, hey, should we go with this counterparty, or that counterparty at this bid process? I go, oh yeah, they're good, they're, you know, they're tough but fair. And you guys should transact with them versus like, oh no, they're gonna be incredibly difficult because their counsel's very difficult. And you don't want to transact with them by another counterparty. I always think, yeah, I mean, we have a mutual friend, Bill Kleinman, you know? - Sure. - That guy's the most optimistic dude on the planet. And I've literally tried to get him to talk, smack about things, you know? I tried to, like, is there any sort of jaded? - Is there a topic or anything? - Is there anything you're jaded about? But it's like, he always finds something to like about everything. - Yeah. - You know, people want to work with people like that. And you know, it's okay to be tough, it's okay to be, you know, have a dissenting opinion, but at the end of the day, for me, the reason why I do this is because of the people and because of the respect that I've been given. And, you know, if I just try to work with people that have that mentality, you know? That makes it makes for a better work day. - Same, same. - Yeah. - I want to talk about something that very much interacts with people, and it's having an impact, and there are different viewpoints on it, and it's AI. Obviously, one of the big drivers for this data center growth, but also being used every day in our workplaces. - Mm-hmm. - How do you view it? Do you view it as a positive, as a negative, something in between? - Yeah, I view it as incredibly positive. It's fundamentally changed the center of gravity. Power densities, thermal management, time to power, that dominates all of our design decisions, all of it. So, I think we're seeing higher rack densities, liquid cool, hybrid cooling, becoming mainstream. There's a huge, really tight integration between IT and facility teams. And then, operationally, our resiliency is being redefined. It's not just uptime, but it's grid interaction, it's power flexibility, it's efficiencies at scale. You know, AI workloads don't tolerate inefficiencies or uncertainties, so data centers are evolving into these highly engineered energy systems. And we've really exponentially increased our design capabilities and just our design posture these days. And I think that has to do with all of the collaboration that we are seeing and just the demand. It's, we have to collaborate and we have to design differently, you know, gone are the days of the old data centers. And that, you know, brings it back to you here, a lot of negative things about data centers. But those are, that's way in our rear-view mirror as far as a design capabilities. These new data centers, I've been through a bunch of them this week and I was completely blown away just how efficient and how they're all incredibly good stewards of the environment and they, it's top of mind and everything they do. So that word needs to be, you know, brought to the forefront in a lot of ways and we're working on that at AFCOM2 to develop almost like a de facto lobby to listen to concerns but also address those concerns with kind of un, like as much as we can with AFCOM, but be kind of like a third party look at, you know, the pro side of design and build in data centers with a, yeah, I mean, one of the, one of the big topics on the sustainability piece of it is the power and where is the power coming from and what is the power source and we had Liz Muller who is the CEO of deep vision, a small modular nuclear reactor company on the show. And it seems like there are just great strides being taken in that space and the time to build, especially with the way that they're doing it has been reduced from like years to like months. What are your thoughts on SMR and how it's going to impact the future of data centers? - Yeah, well, I was in the Navy. So, you know, you got a Norfolk Virginia. You see a whole bunch of SMRs lined up along the dock there. My brother was on a submarine. Some of my, most of my colleagues have been on submarines, you know, nuclear submarines. So, I grew up with that, you know. Look, I don't want power lines over my house. I get it. I don't necessarily want the old school feeling of a nuclear reactor in my backyard either. It's just the narrative around nuclear has had a negative connotation in a very big way. But when you look at it, you know, they offer, you know, long-term carbon-free base load power that aligns well really well with our industry. And always on AI compute needs, you know, especially where grid and expansion lags and demand, you know, lags demand. SMRs are a real thing and they're going to be here faster than we know it. I remember back we had that event a couple of years ago, maybe a year and a half ago. I'm talking about nuclear power. And it just seemed like such a far-off thing, you know, like this was, you know, going to the moon, you know, and the ease or something, you know. And now with those companies you're working with and some of the companies that we've been talking to as well, yeah, they're moving fast, really fast. Yeah, I think they may have a project that's going to come online. I mean, in the next several months here. This is really cool. And they only started working on it earlier this year. So I think to give you-- Wow, and for how quick that timing is moving in part of it is the result of a program that the DOE has in place to help kind of fast track these projects. On that, what do you think is the role of the government in helping to make things easier for these data centers? That's a good question. I think that regulation-- it depends on the administrations. You know, I think, you know, when we were under our previous administration, there was more regulation, more of regulation. And it seemed to slow things down. But I think that we kind of rolled up our sleeves. And we're talking about it more because there was regulation. Whereas when it's less, I don't know. I just think that when the government gets involved, my fear is it causes a lot of bureaucracy and a lot of mess. You know, it's no secret that data centers cluster in places where it's easy to do business and, you know, places like Indiana, they're, they're taxed. incentives through the legislation is like a pamphlet, you know, I think Illinois is, you know, ours is like war and peace. It's like 900 pages or something like that. So I think they play, I think they play a role in the national lab type of work that's going on. You know, you just saw the argon and Nvidia and Oracle partnerships around that. I think those type of things make, make the news in a big way and I think we'll lead to expediting that process and really showing that, you know, it's safe and we can do it. You know, I would love, I would love for there to be a formal sort of like, you know, marketing or almost like a campaign saying, dispelling some of the the negative attributes that are out there around the nuclear or around, you know, data centers in general coming from coming from local governments and the government and I think that's starting to happen certain certain states. Yeah, I think it is too. I think I had a friend who actually reached out to me because he knows I do work in the data center space and he was talking about how he was asked to come in and talk about his opinions on data centers. He was just selected randomly out of, you know, the public and they were trying to figure out just generally how the public feels about data centers and energy sources for data centers and this is being conducted by the federal government. So I think they're definitely thinking about it. If you could advise new data center startups, what would be your top piece of advice based on, you know, your time in this industry and based on your experience starting a company from scratch? So a new data centers start up, are they a co-location? Are they enterprise data center? Are they a multi-tenant data center? Let's say let's the enterprise enterprise data center. So they built their own data center. And are they leasing out of Kolo? Let's say that. Yeah, because the enterprise, the center Kolo. So for co-locations for new startups and co-locations, I think it's very important to have a purpose and have a process and understand and set the culture from the top down because we're in the process, we're in the business of mitigating risk. And so that needs to be pervasive throughout the enterprise. And starts and stops kill you. You know, when you change and you start and you stop, I think, you know, attention to detail early on, specifically around your process on the mobs and the shops and those are the things that organizational process that needs to happen operator, having a good team and following that process and really kind of staying true to your mission. And your core, I think that really, really helps. You know, I had a really great tour yesterday with Edge Data Center. And I was very impressed by their executive team and then also impressed by their design. And so they have a very clear design and they know exactly what type of customer they're going after. And they know it inside and out from from the top all the way to the bottom. Everyone knows the story. Everyone knows what they're designed to. They're designed. They have a hybrid cooling methodology where it's 70 kilowatts per rack on the air side and then the liquid cool side could go well north of that. And they have delineated floors and they have it set up. And that that model, I think, is going to be they're going to do the leader in that space. And they've partnered with sister companies that that make the air units of the CDUs. And so there's there's integrated vertically integration. So as much as you can do that to cut down on these times for scale and being able to go to market and tell your end users or your customers that you can grow with us. And we stand behind it. And here's how that's that's I would recommend. We're we're do you think it's difficult to find to build a team to find a team and build the team in this center. Yeah. Yeah. There was never there's never been a template for it. I'm on the leasing side and sales side and sales engineering side. And so we we thought we went round and round and I can tell you that I give people advice around I say it's really hard to quantify the value you get at a company like Equinix or a company like digital reality or a company like QTS. You know, the young kids that worked for me at QTS, I think that they learned through osmosis just by being around other really high caliber folks that really cared about what they did. And we all shared that common goal and that that process through the leadership from the top. And that you don't find that everywhere. And so it's hard to quantify the value you get at those places. You know, we tried to create that at QTS, you know, and I hired folks from all different backgrounds, you know, so from from telecommunications from other industries that were, you know, top leaders. And then also kids right out of school that were hungry that that impressed me by calling me incessantly asking me questions all the time and being really incredibly inquisitive about the asset class. And I put a lot of our teams through third party training to be able to get certified around understanding 360 degrees about the data center, not just your sales role, but understand what operations does, understand what facilities does, understand the physical security aspect of it, the supply chain aspect of it, the capital side of it, how this all ties together because that comes across in your tour. And then understanding the applications that these people are being able to deploy at your site and how your your data center solves for it and what from the customer standpoint always reverse engineering from the customer standpoint. So when you get down to the facts where you're starting to quote people and you're starting to quote monthly recurring or you're starting to quote non-recurring charges, you're less likely to waive those things because you understand these are very real people that are doing very real work to make this happen for this customer. It's very difficult because in co-location in that world in multi-tenant and also in the hyperscale side, you know, paces everything. And so in getting on the other end of the return of the invested capital is so we don't traditionally we haven't had a lot of time to train people. And that really has you see it a lot in some organizations where the sales people there's a lot of turnover in the sales organization because they get a computer, they get a phone, they get a ramp schedule and they get KPIs and back to that that you know the the adage of transaction, you know, it's like this is not a transactional sale, this is a solution like no other and we need to treat it as such. So do you do you think besides you know taking power constraints off the table, do you think the people constraints is the biggest challenge for data centers right now or do you think what would you say is the biggest challenge other than power? A human capital without a doubt. Without a doubt, yeah. It's just the trades, you know, you're seeing collaboration like we've never seen it, you know, GC's working with other GC's and subs and everyone working together. But yeah, the the need for electricians, the need for mechanical engineers, the need for, you know, labor sales and all of it, you know, we have a huge need for that. And you know, there's I don't think there's the right answer to try to attract. I've been trying to demystify the asset class my whole career and I haven't really devoted a great deal of time to it, but it's always been kind of in my radar and I don't think I've done a very good job that way. You know, I try to personalize it and give people my story, you know, and say that I've created a really nice living for myself, but the real thing that I'm most proud of and that makes me feel really good about where I'm at is the people I work with and the community that is our data center community and that our customers, you know, it's like a different job every day when you're working with multiple different customers, you feel like you have a different job every day. The diversity in this space around what you're doing, you could be, you know, I'll give you for instance, we had a woman that we hired as our organization. So she was the person that came in and when we did tours, she would step up the place, make people, you make people comfortable, you know, you know, show them around, tell them where they're going to go, brought them up to a conference room, got them whatever they need, ordered lunches in all the things and she was amazing. But now that was eight, nine years ago. Now she is an executive and she's done almost every single job in the data center and that education, you know, just, you know, paying your dues and getting that education and sucking up all of that knowledge from all these great people around you and staying in it, I tell you, as this industry grows, it's like, it's like being in the Navy, you know, during war or something, you go from like a ninson to a commander like that. That kind of thing. Yeah, the people is absolutely the reason I stay in it. So looking ahead, both for CWC and for just the space in general, what do you view as success when you're looking forward and you're thinking about, say, 2030? I think, I think 2030, 2030. Wow, I think success will be measured by by resilience, people, and credibility. I definitely think, I think sustainability in the terms of measurable performance, not marketing claims, right? Talent pipelines will be intentional, diverse, and continuous, not reactive. So I think we're a very young industry, so that diversity will come, the pipelines will be intentional, and a continuous process and not a reactive process like it feels right now. And then resilience will include things like grid flexibility, onsite generation, adaptability to new workloads. I think 2030 we're seeing quantum computing be very pervasive, so that's gonna change the landscape like crazy. And I think the winners will be those who plan early and execute consistently. - That's a positive and a great note to our end things. I want to thank you for being on the show. All the work that you do with FCOM, it's always a pleasure talking with you, and I mentioned this to Jim, he was also on the show, and it was a conversation with you two that I was having, and it was such a great conversation that made me think like, oh man, this would be great to do as a podcast, and then here we are, and now you're on the show. - Nice show. - So thanks for the inspiration and everything that you do. - Sumer, it's always a pleasure to see you and talk to you, and I'm glad that your friends. - Thank you for listening to UpTime now. If this episode sparked an idea, pass it along to someone in a new network, and leave a comment. I'd love to hear your thoughts, ideas, and recommendations for guests. Till next time, thank you, and remember to like and subscribe.