All right. Well, Harry, it's great to see you again. Second year in a row. How have you been over this past year?
It's been great, Ernie. I appreciate you having us back and coming out to Vegas for the conference. It feels a little bit like Bitcoin Data Center Super Bowl, which is fun. I think that the last year has been one of a lot of transition but also really positive. Our portfolio's grown, the management team's aligned in a new way, and both of those have been huge positives for everything that we're kind of getting after over at CleanSpark.
That's awesome. Well, I want to start out with a little bit of your background in case there's some investors maybe new to CleanSpark or just new to Harry and may not necessarily know. So, we're going to go back quite a bit and then we'll kind of step into it in your current role. But first question I wanted to talk to you about is you transitioned from building fintech optimization tools, analytics for hedge funds originally through with Enso Financial Analytics and then also helping out with some of the Eastwood analytics. After that, you joined Grid. And you were there for quite some time. I believe according to some of the research I had, you were one of their very first employees back in 2018.
The first. It was our founder and me, two guys in a WeWork. And like any good startup, it starts with two guys meeting in a chat room on the internet. So we have sort of our version of the startup story that was really fun and really gratifying to go from zero to public company through acquisition and kind of be able to bring a lot of the people that helped grow that business along the way and integrate them into CleanSpark. You know, if you look around our team and our booth today, a lot of those folks were people we hired many years ago on the Grid side. So Rory, who you know, came from Grid. Chad, our Tennessee regional operations manager, he covers all of our Tennessee assets. He was a Grid hire. A number of other folks on the team have taken on real leadership roles within CleanSpark, which makes me really proud because we were able to find and grow them in our previous business and then open up a much bigger opportunity after the merger.
Absolutely. If you can maybe another aspect of some of your earlier career is how you were able to kind of take some of the specifics and lessons learned from your traditional finance side of things. Similarly how Rory kind of started out on that side of things, he's doing the majority of the same now with some of the derivative tradings, but you are now in a different type of role and we'll get into that a little bit later on. But how would you say that initial exposure to finance, everything like that, what was it from that that kind of then drew you into Bitcoin and how do you think you're able to kind of leverage that or kind of see that through a finance type lens?
Yeah, so I think what was really important about that experience and really that kind of education is it taught me how financial markets function. And so that brought me to Bitcoin on the one hand, but it also made me feel really well informed to take on a leadership role in a public company because at the end of the day, you have to understand how your key investors and counterparts interact with the financial markets broadly and then how do you as a public company issuer slot into their set of priorities, needs, incentives, etc. And so what I loved about the time at Enso is that it really taught me from an operational perspective how hedge funds work. And so what did we build in that seat was all of the back and middle office optimization, all of the stock loan. So you published great research about the short interest, right?
We built out entire products around optimizing how do you go short companies for these hedge funds. And so while we're not a fan of getting shorted, having that deep fundamental understanding of how market structure functions around some of these trading vehicles, that's really valuable to us as we look to build an operating business that then lives within the public markets.
That's awesome. That's interesting for sure. So stepping now into a little bit more on the Grid side of things. One of your executive roles there would be your Chief Strategy Officer. So now we're kind of getting a little bit closer to what you may be doing now for CleanSpark. But through Grid's initial growth and then through the eventual acquisition with CleanSpark, what were some of the most memorable changes or transformations that you saw maybe going from employee at Grid and then kind of transitioning into an employee at CleanSpark. What are some of the lessons learned that you took from the Tennessee sites, understanding how to operate multi-site Bitcoin mining sites through Grid and then obviously now take it into CleanSpark and now we're multi-state. Just if you can maybe just give a little bit of detail on that.
Yeah. I mean, I think the first is that I never read a PPA agreement or an electric service agreement. Like, I'd never been exposed to energy markets prior to Grid. And so, one of the things that I was the hungriest to learn was similarly to the way that I got educated on the way financial markets work, true market structure. I wanted to learn how energy markets work. What are the components of all these things? And then how do you build an operating business that marries into the pockets of opportunity that live within these broader market forces. So a lot of the optimization that we did in Tennessee at Grid we were able to bring to CleanSpark and then supercharge the growth around it. So at Grid we had mid to high double digit megawatts of built-out capacity that turned into many hundreds of megawatts just in Tennessee on the CleanSpark side. And so the real lessons learned were around getting granular and specific in how do you purchase energy to create Bitcoin. Then it was okay how do you go through a scaling exercise because running one ASIC is totally different than 10 is totally different than 100 is totally different than a thousand and now we're operating hundreds of thousands of them. And so having the creative and mental and technical flexibility to jump those orders of magnitude, we were able to get to the tens of thousands in Grid but then getting plugged into an organization that was operating under Taylor at the hundreds of thousands, that was really exciting for us because it brought a deeper level of rigor, a greater level of oversight, a greater technical prowess to the optimizations that were available. So, I don't know if you've seen our command center here in Vegas.
Not yet. But we'd love to get you out there to see it. But just the level of eye in the sky and optimization that that team is able to introduce to our mining deployment really is just fundamental to how we try to run with true operational excellence across all of our deployments. And it marries really nicely into our AI ambition because it's going to take a lot of active power management. These are, you know, 99.999% uptime type of opportunities, but the procurement of that power, the management of that power, that's the SLA for compute availability. It isn't necessarily the way that you can optimize the power service into the data center. And so, we're going to bring a lot of our lineage and heritage in energy markets and in Bitcoin mining into the AI opportunity. And that's something I'm really excited to kind of see us bring to life.
That sounds a little bit like a new coined term, mullet mining. So, that may take a couple minutes on that and then we'll kind of get back to some of these other interesting questions. I got a couple more about your background before we move on.
Yeah. So, I think Matt's been out there really pounding the table for this mullet mining idea and we think it makes all the sense in the world. It's an idea that we've also been working to rapidly stress test in the tenant community as well, which is the ability for the energy to be managed actively and holistically outside of just the critical IT compute environment. There's a big gap there right now. And so why is that? These sites are built to a particular peak. So think of an energy market. There's peak and there's average. And we know that the whole market is engineered around this peak concept because the hospital can't go out of power, the school can't go out of power, the households can't go out of power. So the whole system is peak optimized and then utilization never comes in at that peak rate. And so there's a lot of opportunity in that gap between peak and average in the electricity markets. In the data center environment, that's actually a microcosm of the same idea. All the PUE is designed for peak utilization. Hottest day, coldest day, most humid day, whatever the case may be. But the average utilization of that PUE energy load is actually lower than the peak.
And so that leaves this sleeve of flexibility in the deployed operating model where there is room for load management. There is room for a demand response flexible layer, which is all that we've been working on for the last 11 years. And so having this very flexible sort of buyer of last resort attached to the data center that's the buyer of first, second, and third resort because they need to have that SLA met, but there's still this opportunity for some squishy load in and above that that sits in between that average PUE level and the peak PUE level that the whole site's engineered around. So we think there's a big opportunity there. We think that it's going to come sort of in bits and pieces as people get comfortable with this integrated in. There's got to be some software engineering, mechanical, etc. that's got to get plugged into a deployment of that nature. But ultimately, we believe that monetizing megawatts is our fundamental reason for existence. And this is a way to be able to express that vision across a number of different deployment types but achieve ultimately a higher and better end than if you tried to just do one thing or another.
Yeah, that's really interesting. So you may know this but my background is actually in power generation. So when our customers, they have basically two different types of natural gas power plants that they'll be running. They're either going to run a base load unit, which means they're running 24/7 at a steady-state type load, which is exactly what you want for HPC data centers. And then they have within that same power plant stack, they actually have peaker units as well. Some of them, one out of three or maybe three out of nine of these power plants are going to be peaker units. So, they're going to be turning ramping up and on during those peak times like you were talking about. So a very hot portion of the day, these peaker units turn on if you're down south. If you're up north and it gets really cold, the same thing kind of happens. So it's a great kind of orchestration that they utilize that for. However, when you do turn on and off those large natural gas power plants, you're actually degrading the actual hardware itself, and that's what we're repairing. That hardware has a specific amount of start life as well as overall hour life. So a lot of units that run this peaker type profile, they end up having to get the parts refurbished a lot faster than if you had a steady-state base load. So when you look at the amount of overall megawatts that are available, even gigawatts that are available, it's really these peaker units that could be running steady-state base load that's going to be able to provide that extra buffer for things like what you're talking about, which is the whole Bitcoin mining aspect. So you're going to find those times when you're not peaking out and you're still going to be able to maintain that and that's going to help the utility company because they can still monetize that excess power.
The utility wants to sell more megawatt hours without having to turn on more megawatt generation.
And so what we've found is that over the last 30 years we basically have been running, and not we CleanSpark, we America, have been running a variable gen environment with how we do load management on the grid.
In the last five or six years we've started to be able to run real demand side variability to meet the needs of the power systems. And so if we can continue that trend we can build more steady-state gen. We can build more flexible demand and we can generate a lower electron cost to a much broader universe of people and that's going to help things like community relations, the attractiveness of bringing a data center to your region. All of those types of conversations start to become about how much can we save the average household on their power, not how much is their bill going to go up.
Yeah, absolutely. Well, I was going to talk a little bit about your degree being with economics from Bates College, but I feel like we're kind of already past that a little bit, so I'll kind of put that off to the side and maybe instead we can talk a little bit more about how you guys are now hunting that power, how you're able to get in there and do some of the boots on the ground community relations and be able to secure multi-hundred megawatt types of contracts in these different sites. Recently you guys had the Texas sites both at Brazoria County which is somewhere in the excess of about 600 megawatts and then you also have the Celly Texas site which is around I think 260.
Thank you. No worries. So with that, obviously I remember on one of the earnings calls you guys talked about the area of Texas, specific area being very fertile for this type of power. Without giving up any of your secret sauce, what's just maybe very broad generic ways that you would kind of go about just screening through all of the various pockets of power that's available.
