The Future of AI in Data Centers: Navigating the Next Frontier in Computing with Bill Kleyman

This is straight straight out of Crumpton with your host, Greg Crumpton. Hello, everyone, and welcome back to another episode of straight out of Crumpton. I'm Gabby Barr, and I am joined by Mr. Greg Crumpton. Greg, how are you doing today? I am suddenly highly caffeinated due to our pre recording conversation with our guests. So if excited today. This guy, I just found out from him, and I are we're coming up on ten years of friendship and electronic geeking together. And he's the real deal. I you know, I'm a wannabe. He he knows so much more about it than I do. I'm I'm fascinated with everything he does. He's got a lot of stuff going on, which we'll talk about some of it. We don't we don't have all day, but Lot of cool stuff. So why don't you kick us off? Tell us who he is, and then we're gonna hear from the man himself. Absolutely. Well, I am excited to introduce Bill Clement. He is the CEO of Neuro, and he also works with Afcomm data center world He is a program chair contributing editor, data center knowledge, and data center frontier. Bill, how are you doing I am I am doing so well, Gabrielle, and and Greg, it's great to great to be on here. For all for all the audience to know about this, this podcast was in the planning for what, like, I think six months if I'm not mistaken. I think you and I met Greg at a conference recently. And you're like, hey, Bill. Should get you on the podcast? I'm like, yeah. Sure. Thinking it's gonna be, like, in a week or two. No. No. This guy's schedule is so booked up in podcast that If any industry celebrity wants to get on this on this little platform, oh, you're you're gonna have to wait. Oh, there's a there's a line. So I I'm doing well. It's an honor. It's an honor to be here. I should just say. Well, thank you for that. We're glad. I think that really shows how cheap I am that I can't afford more producer help more than once a week. So we are, definitely glad you're here, Bill. It and where were I forgot where we were when we saw shut. Oh, in, DC, I think it was. I don't remember. Somewhere. It was a big conference hall. We had just finished listening to an event. I remember our conversation very well. It was serious, like, five or six months ago at at the very least, at the one of the very many In in my door, I know you all can't see. This is this is one of those podcasts where I wish I could grab a camera and go totally, like, you know, like, youtuber mode. And in my closet right over there, I have a Christmas tree of badges, like my honor badges, like medals of all the conferences that I've been to since day one. And I'm pretty sure that hook is starting to get a little strained from all the way to those bad See, I wish I would have done that. People like you are so much smarter than I am, because I'm such a non clutter guy. Like, I'll throw them away at the airport on the way out of town because I'm that stupid Now if I had any sense, I would have done it from the get go. I'd have forty one years worth. Do you believe that? This is my forty first year of doing this stuff. Crazy. I know. Thirty year old guys hard to pull that off. You know? I I mean, the conference worker keeps young, Greg, clearly. Gabby gave you such an introduction that I'm agreeing with envy just because of how many words are in your introduction. But what do you do, man? What do you I mean, you have a lot of stop. What is the core of your existence other than your wife with and you around the house? My my wife and my five year old. It's a pleasure to be on this hot cast. And what's really great is, you know, behind the scenes, everybody listening, I know when we're recording this, and I know when this is gonna be released. And what's fascinating about that is that You're gonna be hearing some really fun announcements as we're gonna be releasing this podcast in this conversation. I know Gabby, Gabrielle, you're kind enough to call me the the CEO of Neuro. Now hang on a second. You might be like, wait, what is Neuro? That company name may have changed. And it might have changed at this point, but I'm still gonna be CO of it. I wanna make sure that you and I have time for a conversation. I think, like, even a biography on me at this point is gonna take up, like, half this chat. I'm really fortunate. What what is the what is driving factor in the Cora Bill claim. And if you ever take a look at the term extrovert in a dictionary, you'll see this face And, you know, you all give me energy, right? The industry gives me energy. I'm very lucky, Greg, and everybody listening. I'm a native to this the data center space, the IT and technology space, where I've asked this question many, many times, like, how many people meant to be in the data center industry, graduated with a degree or you weren't adopted. It's it's usually like one or two people me who raises their hand very sheepishly and is like, I'm not a former or pediatrician or a marketing professional. Like, I graduate with a network engineering degree, and then afterwards got a master's in business and another master's in information security, but right off the bat, when I got my my network engineering degree, I went to a trade school. There's there's no sugar coating. It went to a trade school, went to Dubai. I was twenty one years old with a network engineering degree, which in parallel, I got with my Cisco certified network admin degree, certification, the CCNA, And I I was literally racking and stacking and working in the data center space before I even realized they were called data centers, you know, like, I mean, full racks, you know, can came and putting Cisco stuff in there and like, oh, I guess this is what data centers are all about. So I've been doing that for for nineteen years. And yes, despite me being a millennial, I'm so fortunate to be in this industry for a long time. And over the course of those years, I've had a chance to work at the nation's largest partner. It was the largest platinum partner at the time. I worked with end users. I've worked with one of the world's largest DevOps development application, migration, cloud organization called EPM Systems. I had an amazing opportunity to work at switch data centers where I started off with the company they were four and a half billion publicly traded. You know, I was a part of the team that helped them grow. And ultimately, we helped them sell for eleven billion in December. Remarkably, January of twenty three is when I said, you guys keep doing amazing things. I'm gonna do a couple of things on my own. And that's when I joined an amazing organization called Neuro and at a very, very high level. Again, you could probably still go to that website, n e u dot r o. It might point you to a different name a different company, which is really weird because we're talking about this now, but we're gonna hear about this in the future as well. We entered this market. In in back in twenty nineteen, there's an extraordinary story behind it, but right now, we are in the United States, and we are the only platform in this entire industry, the interoperability platform where we are basically the the garage that helps you build any kind of AI infrastructure you can possibly imagine. So We go into data center facilities. If you have any excess capacity, we put ourselves right on top and say within five days You can compete with Amazon, Microsoft, Google, and literally become an AI company, but here's the cool part. You put your name on there, you put your logo on there, and it becomes a part of your organization. Now There's no other platform in the industry right now that can enable both multi tenancy and self-service. So literally, you could put a couple of GPUs in there. And sell it to the market with your own platform, your own portal payment system, and everything. We're definitely gonna use this market by storm because it's a it's a hungry space. And I'm missing a whole bunch of really fun pieces on them. I'm originally from Kiev, Ukraine, came to the United States in the early nineties and continue to support, you know, my my friends and family out there, Just get a chance to work with this industry, which is a lot of fun and never stopping. I've told you he was wide open and he is. What I really admire about Bill is his ability to and this is a gift because a lot of people that are really skilled in a technical fashion are necessarily as skilled in the human side, the EQ portion. And I think Bill brings a super duper balance to that because He knows all about that as we just heard. And even some before we started recording of how into the space he is technology wise, but as you see, it trans it transfers over to people. You know, he connects with people just as well. So it's just a a cool mix of a human. So thank you for that, sir. So this is kinda fun being able to talk into the future a little bit because we know when this was gonna be released and things may may change. But what had you done recently in a technology way or had vision to or got to play with that was really cool. What do you what have you been into? We've been heavily into the AI space. And What's fascinating about this, you know, the the straight out of crumpton conversation in this podcast is that I I am gonna talk to the future here for a second. There's, one of my mentors, his name is Peter Gross. You may have heard me say this a million billion times, and I'm gonna get to a point where I'm gonna challenge this statement because a very dangerous statement in a way, but the data center industry loves innovation as long as it's ten years old. Kind of a dangerous statement to make. Right? And so when we start to take a look at this idea and concept of of the data center industry loves innovation as long as it's ten years old. We realize that we are in this position where we are right now, where the hyperscalers, these large cloud organizations are winning this AI race. Right? Let me let me answer your question some of the stuff that I've been getting a chance to work on. Over the past year, I've had this extraordinary opportunity to work with data centers and to modernize them, right, we're literally talking about, okay, we as an industry have been operating at eight, nine, ten kilowatts per rack for a long time, and it's been fine and solid Now, all of a sudden, you've got these AI companies and workloads. Forget companies per second. The customers that are saying, cool, cool, cool. Can you manage this level of density, which now is like thirty, forty, fifty, sixty, seventy kilowatts per rack. And it's like, whoa. These are massive increments. So one of the most fascinating things that I think we've been working on is the evolution of things like liquid cooling. We've worked now with several data center partners where we've help them evolve from traditional rack and stack style of compute to working with rear door heat exchangers directed cooling where they've gone from ten, fifteen, twenty k w to sixty, seventy kilowatts per rack, now capable of handling some of the most high density NVIDIA GPU intense kind of workload and architectures out there, and what's been most fascinating are the conversations that we've been having with the end users, the customers, What we've noticed that we haven't been driven by the data centers necessarily, we've driven by their by their customer base. They don't wanna be ultimately in Amazon. They'd want data locality control. They don't want their information to be used to train some big foundational model, and they want cost control as well. So they've been going to data center partners and saying, Do you have anything for us? And the typical response from our community has been, this doesn't look like traditional compute to me. I'd I don't I don't have this. So we have a market that's driven to move, in the industry that's not necessarily ready to pivot to that. Now we don't have ten years. We we don't have ten months. When the foundation of the internet began, it was the hyperscalers that helped mold and craft that. Whereas we, in the data center space, we're relegated with, you know, Exchange servers, SQL databases, VMware, and Citrix workloads. And the same thing started happening with artificial intelligence beyond anything else, generative AI. Right? And so all of a sudden, these big hyperscalers are molding the way once again, Amazon, Azure, GCP are the ones that you gotta come here and do all your work with us. Whereas again, the data center was left a little bit behind. Well, we're not doing that anymore. My past year, and even into the release of this podcast, has been working with data centers both very, very large, as well as very small to help them democratize and capture this market. That's been a tremendous amount of fun educating this industry on the types of software tools that you need, how you revolutionize your infrastructure to support higher density workloads, what that even looks like, how to ascertain this kind of architecture, and most of all, fact that this is gonna be released sometime, in the near near future, Nvidia as of this recording has about seventy percent of the market. In terms of GPUs. You better believe they're not the only horse in this race, and we've been having really fun conversations with AMD and Intel as well, they're gonna be impacting it. So very long with an answer here. My past year has been an extraordinary exploration into the evolution and innovation of the data center industry to help them increase their capacity and ultimately handle and support some of the most advanced artificial intelligence, generative AI, large language model workloads that are in this industry, and and none of it slowing down. He brought up several points of of which I have really in particular interest. I think about existing space and as many millions of square feet that we have that are operating at, you know, eight, ten, twelve k w. We have these high scalars that are they have the resources to go and and totally, you know, retool at a whim. My thought and and question is is kinda twofold. What happens to, you know, a legacy twenty thousand square foot colo facility and how do we adapt? Because of density, because the spaces weren't built either electrically or mechanically to support that So I think about how did we retool, you know, and I I've got, I don't know, a hundred in my head that I can name off of physical spaces that I know the infrastructure well, and they they simply can't do it. So that piece, and then when you're out and about and people are so people centric. They wanna work with people. They don't wanna work with an acronym. They don't wanna work with AWS as an acronym. They wanna work with Bill. How do you and I know that this is kinda jumping all over the place, which I normally do, but how does the people side interact blend with explaining the new data center coupled with what do we do with the old data center? If you can make heads or tails of that, good luck. But I try to take down your your multiple threads of thought here, and I didn't we'll we'll see what I did, as far note taking here from Greg Crumpton's streamer thought. So it it's important to note, and this is something that we're seeing. Every data center's gonna become an AI center. Ultimately, it all depends on how quickly you can get there. Over the course of the next five years, large language model based applications are gonna be as dominant as your very own traditional web applications that you go to right now. Right? They're gonna have some kind of a large language model backing it. What I want everyone on this phone call to understand And, yeah, we're throwing out some really fun acronyms. Yeah. We're having these really crazy, cool discussions around the future, and yeah, it's all cool, cool. Cool. What we're talking about isn't in innovation and technology. We're not talking about, you know, obviously, we're gonna mention GPUs and, you know, data center architecture. I promise I'll answer your question with legacy to modern. But what I want everyone on this phone call to pause really quick right now, get off your Amazon shopping list or whatever you're doing, and understand that we're experiencing right now is not a technological shift. It's a shift in humanity. Right? How how you and I and everybody listening to this extraordinary podcast Krompton works with data. For the past twenty to twenty five years, all of you, every single one of you has been conditioned to search for data in a certain type of matter. Let me give you an example. So Greg Crumpton, right, is gonna go to his favorite search engine, ask ajeeves, maybe Yahoo would if he's feeling feisty on a Monday morning, maybe even Google. Who knows? But he's gonna type in his favorite question. Right? He's gonna type in his question, and he's gonna get a blue link. That blue link is how we've been conditioned to look at data for the past twenty to twenty five years. Right? And god forbid Greg Crumpton ends up on page two of Google, Nobody belongs there. That's like mortar. You should no one belongs on page two of Google. Right? If you need you need to find your stuff on page one, but I'm being facetious That's how we've been doing data for a very, very long time. Ask a question. Here's a blue link. You do the rest. Well, today, that's fundamentally changed. Right? If all of you, in fact, I'm gonna challenge you. Every single person listening to this has most likely used Google or Bing in the past, my twenty four hours. That means that every single one of you has become a user of generative AI. Because when you use those search engines right now, the first response is no longer Blue Link. It's a generative AI response. Okay? And that is something that we've never ever ever seen happened to us. So it took five days for chat JPT to hit one million users. It's about a month for them to hit a hundred million. And after about nine months, nine to ten months, they were about a billion unique users. There's only one application, one application that was capable of hitting one million users after one hour. And I'll tell you what that application is. And some of you might be like, what? It was threads. If you remember Instagram's attempt at the Twitter takeover, threats came out there. People like, this is so cool for, like, an hour. And then afterwards, you're like, I'm good. I'm good. Alright. Back to chat, GPT. Right? Whereas, if I ask that same question of ChatT, almost everyone will raise their hands. So that was a very long introduction to to your question because what I want everyone to understand is that we're experiencing a shift in how we interact with data. For the first time in humanity, for the first time ever, we're asking data a question and getting a conscious answer back. I'm using the that word very carefully. Right? I'm not talking about this sentient being, but we're getting a conscious answer back. That's huge, everybody. Please please understand. This isn't just some guy who's caffeinated. Right? Actually, I've switched to tea now. We've changed the way we ask in interact with data, and this is the foundational time where all of this is happening. Now the fact that this is happening in the near future might mean that we're new working with a new iteration of GPT. Some new architectural models are on AI, but the fact of the matter is we are a year into this, and none of this is slowing down. So to answer your question initially, what happens to legacy data centers? Look, there's going to be an evolution in this space. Right? We're gonna see more solutions that are gonna be around build around sustainability, working with less water, closed loop systems that have very little impact on local community watersheds, you know, obviously, you know, ways where we can we can sustainably create these architectures The the one of the big challenge that we need to work around is something that we were founded on. AI ethics, AI transparency, and AI sustainability. Those are core foundational elements of my organization. And a lot of it's being transferred to to other organizations as well and and and other ideas and concepts of how to adopt these technologies. So from a physical perspective, again, the challenge and this is from the latest Afcom state of the data center report, our industry operates about eight to ten KW rack. You can't work with artificial intelligence if that's your density just to give everybody perspective. Right? Because we're talking about AI. We don't have a way to conceptualize it. A a DGX node from Nvidia, right, the green mean machine, of h one hundreds. These are the ferraris, right, of of the GPU space. One of those units is six u, right, six u, and it slides into a forty two unit rack. Right? A six u unit of just one of those d g x h one hundred nodes, and there's eight cards in there will consume anywhere between ten and a half to twelve KW rack. Just one. Just one. So what are you supposed to do? Put one in there to whole bunch of blinking panels? That's not gonna look very good, and it's a horrible use of space. So you need to advance architecture to support some of these most modern workloads. So that being said, we've been heavily involved in working with data center organizations that are hundreds of thousands square feet in space, including smaller ones as well, ten, twenty thousand square feet of space. And as long as you have a good foundation, you know, good control, good containment, you have an extraordinary capability of moving into that more modern type of architecture. Now we work with countless partners right now, physical data center spaces that are stuck in that ten and a half eleven twelve k w range, and really all they need is a few sensors to understand whether they have hot cold spots, a little bit of improved containment, a rearchitecture, a hybrid hybrid environment of a good air cooled system, a good rear door heat exchanger architecture, and if you wanna double dip. I use that, you know, there there's my pun. You could do direct to chip liquid cooling alongside a rear door heat exchanger, and you can really create some advanced. Our detector with your very traditional standard racks that can now range from forty two units all the way to sixty you so you can really use that vertical space very, very intelligently. The point is you can get creative. You can get very creative in how you do not need to go out there and put in a giant immersion pod, which I think are very cool as well because that's not the only way you can enter this game of artificial intelligence. So we've been working with a lot of the legacy organizations, data centers, but there's a specific reason. It's because of their customers have gone to them and said we are motivated to move from the cloud. We wanna have a part of our infrastructure with you because we already have our exchange servers with you our storage environment, our core critical pieces of infrastructure. So all you have to do is put a little bit of extra, right, GPU capacity. And again, this is where Neuro comes in as that platform to enable these organizations, how you become a part of this game and impact a customer base that's really, really willing to move. So there's good news here. Right? Now that being said, if you have a raised floor environment that's got, like, you know, a rack spacing between each of your of your, you know, racks. And you have no idea what's underneath the floor of your raised floor, and you're like, we haven't looked under there in fifteen years just to ignore it. That's gonna be a problem. Right? We're gonna have to talk about so we're architecting some data center ninjaing to bring you up to par to make it, you know, work a little bit more successfully, but I'll give you a great example. We've got a partner out in in Lafayette, Indiana WinTech. And I'm hoping that by this time our conversation has evolved even more what an extraordinary use case, great local market, secondary, a lot of capacity, a lot of power, really good green resources, and a local community that's force them. They don't have to be deployed in Northern Virginia. And what we found at those secondary tertiary markets are prime. Oh my gosh. They're their prime for, you know, creating GPU farms and accessing this really hungry market. Heck, it's part of the reason why I went to why I went to New Zealand as an emerging market, and I'm gonna be out in Australia as well in Sydney in September of twenty twenty four talking about these very same concepts of where you can build these powerful AI architectures sometimes in very remote spots. So legacy to modern is happening right now. And here's who's benefiting the most. It's the retail facilities, retail co location facilities because the larger hyperscale colocations are kind of stuck here. Let me explain why. Their primary customers are Amazon and Microsoft and NGCP. So while they can support the market, they can't necessarily participate in it, which means they can't deploy their own services. They can't deploy their own GPUs. They always have to do it for somebody else, whereas, like, a retail data center, like, our partners That's where they put a little bit of GPU space. They put Neuro on top. They put their logo on there, and they're like, hey, guess what we are. We're an AI company now. And that's how that level of modernization happens very quickly. I love what you said about the hybrid model. I'm a I'm a huge fan of a little bit of this and a little bit of that in your existing space. I think that totally makes sense. You touched on one thing, Bill, that I I find really interesting thing. I've been following a company, DC blocks, and where they're building. And, you know, they're definitely looking at tier two cities. They're they are laying a a big old fat pipe from from Myrtle Beach, South Carolina across the southeast pretty much. How visionary do you have to be to start thinking about tier three markets, or are we so early in tier two that that that's premature or should we start thinking? And I was reading a piece this morning about the Quantum loophole deal up in Northern Virginia with the generator issue with the county. They proposed two hundred and six some of my generator aligned did, and they put them on hold because they start to become a power generator at that point instead of looking at each generator individually. I'm sitting in a in a position of having to think through. Alright. We're brand new at the AI game, as far as local usage or common usage, maybe is a better word. We're starting to think about what's next tier two cities and and different toucher of how to get there, is it too early to think about what's next? And do we have dark fiber going to places we think we might want it? You know, what what are your on that, kinda looking ahead. I I know we're looking ahead already, but now now we're, like, if you've gotta make a ten year investment and land, what do you start thinking about? It's a great question. It's a great question. I think a lot of the focus right now is obviously on artificial intelligence and how do I support that market ultimately? Funny you mentioned DC blocks. Heck who knows, by the time, you know, we we release this podcast, I'll be able to, you know, alongside of it. Have a really much more fun announcement that I can talk to right now. But that's one of my targets, for example, for neuro. Right? And and I'm just gonna give you an example of DC blocks. This is just me looking at my market. No. No. No. It's it's them. It's both like element critical, for example. You know, my partners at Scott Data Center is another one there in Omaha. I'm in Omaha, Nebraska, WinTech is in Lafayette, Indiana. Talk about tier two tier three markets. Let's let's focus on that for a second. But for someone like DC blocks, whose customers are largely, retail. Right? They have some other kinds of customers as well. They have this extraordinary opportunity. This is again where we come in They put in GPU space, they put Neuro on top, they put their logo. It becomes DC blocks AI, and all of a sudden literally They continue to do what they're amazing at, which is selling space, capacity, and power to emerging markets, but all of a sudden, they become AI company as well through our support where they don't have to do anything other than again, put that platform on there and service their customers. Now they have a very hungry customer base. Obviously, that right now has quite a bit of that workload in Amazon and wants to move it on prem. Let's talk about the market specifically. Right? It's been fascinating to us to see how a lot of our attractions been coming from tier two and tier three markets. When we flew out to New Zealand to talk to them, it was very quickly the datacom and Spark and Zee came to us as data center organizations like Bill, we love this AI chat. How can we become a part of it? How can we be relevant? And their concern was latency. Right? So between Auckland, New Zealand, and mainland Australia, I think the ping time is, like, thirty two milliseconds. That's not a little bit. It's not a little bit, especially for latency sensitive applications. But a lot of AI work that we've been doing is not latency sensitive. Right? Imagine for a second. You go to New Zealand, you build a facility, high density, very clean power energy, low cost per kilowatt, and you create a GPU farm. Right? In those situations, you just need the data to sit there for a long time to train it and then ultimately you can send bits and pieces back. The idea behind large language models and AI training is that you don't wanna speed the process too much. You want the data to be accurate. You want it to be low bias, high accuracy. You wanna make sure that this information is well, well trained to make sure that you're not creating hallucinations when you ask a question that the data output that you're creating is very, very accurate. Now the ramifications of improper data is even more impactful if you're like a health care or financial services or science and resources or pharmaceuticals, right? So you have to make sure that you're trending that data to its fullest capacity. You know, there's one time we were asked to do a proof of concepts for an automotive company. They gave us a week. They gave us a week, Bill. We want you to create this really cool platform and we're gonna give you a week to train little bit of data. And it was garbage. It was horrible. Like, the you asked the same question of the model, and it would give you two different answers every other time. And one of them could be completely not coherent. And that's because we've never had a chance to train that model properly. So it's not about getting the data out there fast. It's about getting the data out there accurately. So these second tier, these third tier markets are very much emerging spaces because you can go out there, create these great data pools, these architectures to farm access, and work with GPU, architecture support artificial intelligence, large, medium, small language models, and everything sort of in between, as well as legacy AI as well. So one of the most interesting things that we've been seeing are some of our biggest partners have come from tier two and even tier three markets as long as there's some decent connectivity there, again, part of the reason why we're going to Sydney, Australia, New Zealand to talk about how those areas may very well be emerging markets as well. So, you know, you mentioned the whole generator's thing in in Prince William County and London and all that. And I've been in a part of a lot of debates a lot of conversations with people holding the picket signs as well, and they've had a lot of really good concerns out there. We certainly need to become a better part of the community, but there's been so much incredible work being done to make sure that these data centers are no longer being built as these nondescript remote facilities, but rather being part of the community directly. This is a very long winded way of me answering this. The secondary and the third tier markets are gonna become much more laser focused for a lot of these organizations to try and enter the data center market, especially in the focus of AI. Let me give you one very specific statistic. According to Jones Langlacile JLL, as well as a DC byte as a really good research company that I've been working with. If I'm not mistaken, when this podcast is gonna be released, the wholesale availability in Northern Virginia of data centers is gonna be zero percent. There's gonna be no availability. So what choice do you have? You have to look at these other markets to deploy into? I think last week, I looked at a piece of property up there that's gonna house four hundred and fifty six megawatt data centers, seven buildings, but you don't build that overnight. It takes time. So we have to be able to have places for people to go in the interim. I don't know. It's just really interesting to me. I love I love the realism, a structure, a facility and all the infrastructure it takes to make it happen, but we have to keep this thing growing while all of this is happening. And it's like, It's just a great time. I I love being in the industry right now. There's so many cool things going on. So many cool people that, you know, we get to interact with and work with. It's a fabulous time. I'm thrilled I get to be around and participate in it. I had a flashback while ago when you said Cisco gear and containment. I remember the side draft thirty six hundred Cisco's blowing hot air into their neighbor in the telecom space. And I remember making hot aisle containment specifically for that thirty six hundred where we could get that hot air out of that rack in a sideways configuration before you could buy it. We were making it out of plexiglass. So, I mean, that's how fond it was back then. You know, when people were running, it was a little bit like the Wild West, and and I'm fixin' you with your your black DIC hat on that I saw recently with your cowboy hat. So Just really, really good stuff. So let's transition. We start to head toward the top of the hour away from technology and more toward the people. You you mentioned Peter Gross, who is, in the street legend, as being one of your mentors. Who are some of the other people that you look at that have either helped you along the way or really brought foundational worthwhileness, I guess, for lack of a better word to the industry. What what who are who are some of those folks for you? I'm gonna try and rattle this off from the top of my head. And if there's anybody I forgot, and you're listening to this, it's it's not intentional. It's because there were no rehearsed questions, literally, Gabriela and Greg's, like, here's the time this, you're gonna record this, you're gonna be there. It's a Monday. It's we don't need we don't need a script. So, Peter Gross is definitely definitely one of those people Nancy Novak, found out this miss Young from from switch, Eddie Shooter, formerly of Switch as well. Amy Marks, she taught me a lot about pre fab that that lady's name. The modular queen. Yeah. Yeah. Carrie Gets, who I always enjoy working with. She's she's an extraordinary people person and has opened my my purview of how to work with different types groups of people, which I really appreciate. Carrie's been on with us. She's an old friend. Yeah. Yep. Yep. She's she's she's one of the people I tried look look at Rich Miller, founder of data center knowledge, and then data center frontiers, been a huge, huge, huge mentor and Kevin Armando from the publish side as well. Those are the folks that run like data center knowledge, data center frontier. They got me introduced to Afcom and data center world as well. Trying to see who else. There's one more individual who's been a huge mentor of mine who's not even in the industry. His name is Sean McCarthy. I'm actually having lunch with him tomorrow. He is the CEO of a manufacturing from worldwide fittings, and this dude gave this I think I was, like, twenty five, twenty six years old. The keys to the entire data Kingdom up as an organization. And we're talking, like, you know, several hundred million, five hundred people, six locations in the United States, two in China, one in the UK, and I manage the entire technology infrastructure for all of that. It was an extraordinary lesson, and he taught me so much about learning the business. This guy made me get a forklift driver's license. He made me pick orders in the warehouse so that I could deeply and intimately understand the business before I get with the technology. Amazing, amazing, amazing industry, you know, just knowledge from that. Those, you know, at the highest level possible. I I feel like I'm I'm forgetting a few folks from the top of my head, but those are those are some of the real big individuals that I've I've seen as mentors, so folks that I I sort of just, like, try and follow on the periphery, people that I can I can ping regularly and ask questions about, and there's there's people that are younger than than myself as well, who I consider mentors like Skyler Holloway, former peta? I work with him at Switch as well. Valeria, who I work with at. I'm Aisons and nomad Futrious. I think she's awesome because she's she's just like a go get her and tries to go out there as well. So did you say you broke up a little bit, Bill? I didn't hear the last name. Oh, Valeria. She is the current cochair of the IMA a millennial and gen z resource group when I stepped down from that. And and so I had a chance to work with her. And she's younger than myself, and she's great. I got a chance to learn from her and what the the up and coming know, generation requires to be successful in this industry. My god, that that was just off the top of my head. And I I know there's, like, at least a dozen more names that I wanna talk about. Obviously, Peter Gross, we we mentioned him as well. But these are individuals that I will unabashedly ask questions of, you know, where can I learn? Where can I become better? Where am where am I not good? And I'll tell you what becoming a CEO of the company quickly made me understand all the things that I'm good at all the things that I'm not good at as well, almost simultaneously, which is an extraordinarily humbling experience in having to go to people and and ask for help. So And and Chris Crosby, from Compass, he's another big, big friend and and a mentor of mine, and he's he's a big supporter of fail fast and get up faster? Well, I I ask. I think more so. All those people, you know, from and I only know a handful of those personally, but the ones I do know are all open minded. They love collaboration. They wanna help the industry as well as help people But what's really intriguing is the difference of the people that you talk about, you know, like what they do within the trade or within the industry that's different. And I think that's so helpful. The the fellow who gave you the keys to the castle, I so much applaud him for giving you the instruction to go pick orders with a forklift because if you don't know what everyone's struggle is, you can't solve the problem. And, you know, if if you're picking orders and you've got this stupid app or or a piece of software that you have to use that's crappy and you have to use it. There's no better judged than you of crappy software or application. So now you know the real struggle of what the common folks are out on the floor. And I think that's so important. So important. Here it is. Here's one of the parts. You could see the logo on there. It says worldwide fittings. Right? There's what these are one of the fittings that these guys made, and I have it on my desk always. I've got a couple little things. Like, these are these are things that I help pick, you know, from you're right. You're right walking a mile in the end user's user shoes. You'll understand why the CD round doesn't work. My job, I get to get out and go visit with our companies, and we've got fifty some of my companies around that North America And when I get to go get on a rooftop and hear and watch one of our technicians, piece of software or an application that he's using, or she's using for either diagnosis or for reporting, you get to see what works and what doesn't. And when you have to sign out and sign back in, instead of using, you know, single sign on, what are we doing? We're we're giving people nine times the work when we don't have to, and you don't get that sitting in a damn cubicle. You have to go. You have to get out and go. And I just love that. But we're up on the hour I feel like we could go for three more, but your your tea may run out before we get there. I don't know. It's a big mug. Look forward to hearing us in the future talking about what's happening today that you can talk about in the future. So that's gonna be fun. When this gets released to hear what's new with you. But I value your friendship. You're a good dude, and I love when I get to see and talk a bit. But more so, for what you're doing for the industry, you're helping a lot of people even though they don't know it through what you do every day. So thank you for that. Appreciate it. I I appreciate you, Greg. And for everybody listening out there, one piece of final, something to think about for everybody. Right? Find me online. I'm hoping after this gets released, my my inbox and LinkedIn get gets blown up. You can see the videos that I do for Intel for Oracle for all these really fun places, and you can see me wearing my cowboy hat in weird data center locations as well. Just remember one one little thing, everybody, vision without execution is just hallucination. So when you hear about all these crazy things, whether it's artificial intelligence, data, just all this stuff that might sound scary. Learn about it. Ask questions. You know, there's my Ted Lasso quote of the day. Be curious not judgmental. And I I really challenge everybody here again to to better understand this, not this shift in technology, but the shift in humanity. Because it happens so quickly that you don't even realize it. Right? Before you know it, right, we're running around with these really advanced cell phones in our pocket. We're just a little bit earlier, we we didn't have any of this. I mean, it was it was June of two thousand and seven when the first iPhone commercial came out. It was during the Oscars and the commercial was simple as Hello. That's all they said. So the one thing that I want everyone to just think about is that technology moves at such an extraordinary pace, extraordinary pace, And right now, we've never had this pace to compare anything to. I know this sounds like an over caffeinated excited little chipmunk telling you this, but this is your moment to at least understand this market, understand your influence in this whether you're an operator person putting an HVAC gear, wiring, building MEP, whatever the case is that you're listening to this. I'm talking directly to you right now. Take a second to really understand how everything that we've discussed, generative AI, these solutions are gonna impact your life because they will. They will. Again, these solutions will become a prevalent as your regular applications that you use every single day that are going to be backed by the technologies that we're discussing right now And my my challenge to you is is have a vision. Have a vision against anything that can help you and then execute on it. Well, Bill, that's strong words, great advice for any number of things in life. But in particular, what we what we do with our digital world So well, gabby, another I'm still laughing at chipmunk because I'm thinking chip as in computer chips. So that's a that's a good one, Bill. I'm a I'm a have to remember that one, so really do appreciate it. Gabby, why don't you, kind of wrap us up and we will head down the the trail toward, more chipmunks. Yeah. Absolutely. As Greg said, Bill, thank you so much for joining us today and telling us kind of about what you do and where your interests and passions lie. It's very clear from, you know, the moment you got on the call, you really are super passionate about what you do and what you explore when it comes to the world of data. So it was super interesting to hear about what your world looks like on a regular basis, and I'm really excited for this episode to release for everyone else who's listening to hear about some of the exciting changes that are going to be true by the time this episode releases and be announced and everything like that. Thank you again for joining us and as for everyone else, thank you for listening to another episode of straight out of Crumpton. For more episodes like this, please subscribe to Apple or Spotify Podcasts And don't forget to check out greg crumpton dot com for all of your Greg crumpton content. This this is Freet out of Crumpton with your host, Greg Crumpton.