Awesome.
Good morning, everybody, and thank you for joining us here today for a discussion on crypto mining. We have three of the top crypto mining companies in the world here with us today. Good to hear Cipher and Queens Park. Thank you all for joining us. Let's start, and why don't we take a moment for each of you to just introduce yourselves and tell us a little bit about your respective companies.
Sure.
Eric Farrell, Chief Financial Officer for Queens Park. We currently have over 42 exahash of processing power, producing about 22.5 Bitcoin a day. We're solely focused on Bitcoin mining, primarily in the U.S., and we have over 12,600 exahash currently.
Hi, I'm Tyler Page. I'm the CEO and founder of Cipher Mining. We, too, are a Bitcoin miner. We also are in the burgeoning HPC space for data centers. Across both of those businesses, we're probably best known in the industry for our expertise at originating greenfield sites and favorable power contracts. We're known as having the lowest cost of power in the mining industry at $0.027 per kilowatt-hour. We do that basically through active management of containment and then trading and hedging in the Texas market.
I'm Dan LaBerge, Head of Capital Markets and Strategy here at Pixie Gear. Pixie Gear is a Bitcoin miner, but we're also a technology company. We have about 2.7 gigawatts of power in our portfolio. Just over 1 gigawatt of that is currently used for mining Bitcoin, but we are also entering into the Bitcoin mining manufacturing space. We manufacture ASICs just for our supplying use as well as sales to third parties. We are also exploring the HPC and AI opportunity, and given our size of our power portfolio, we believe we've got a number of sites that are suitable for that. We are very interested.
Awesome. Thank you. Let's start off with kind of your thoughts on Bitcoin and what you're seeing in the market. From the crypto market backdrop, what headwinds or tailwinds should we be watching for? Gary, let's start with you.
We're very focused on Bitcoin, obviously, and I think that really the biggest tailwind is the change in administration. Having a Trump administration is obviously very crypto and Bitcoin-friendly. It's been very helpful. SEC is also a lot more friendly these days. I think that on a regulatory front, that's going to be helpful. What we're really seeing is some real tailwinds with demand that's starting to come into the marketplace. There's a lot of talk about the strategic reserves. There are nation-states that are getting in the game to buy Bitcoin, and now coin treasury companies are becoming the new fab. What we're seeing is this build-up of demand for Bitcoin. We saw a little bit of a spike in Bitcoin price this weekend, almost into an all-time high.
Ultimately, a combination of regulatory, political, tailwinds, and some general demand for Bitcoin, I think, is something that we're looking forward to some appreciation in Bitcoin over the next 12- 24 months.
Tyler, do you have anything to add?
Yeah, I would just add that I think we are in the transitional period probably now for Bitcoin. I've now been working in this space for eight years. When I speak to the question around Bitcoin valuation, it has always been one, to me, that is about network adoption of a new kind of network. The fundamental question investors need to ask themselves, and the reason I moved my career into the space, as I thought the answer was obvious, was, does this network with now maybe 200 million users grow like the internet or like cellular networks and have billions of users? I think it's very logical that that will happen. Folks, the market is a very volatile market. It trades 24/7, 365. It gets a lot of eyeballs day to day. It's exciting.
I'm sure my panelists were like me, up at 3:00 A.M. seeing what was going on in Asia after I went to bed and it was $107,000 last night. I'm a bit disappointed, so I'm going to go the other direction. On a longer time frame, all these things Gary mentioned to me are just signposts on the way to broader adoption of the network. Over a longer term time frame, we're just extraordinarily bullish. I think it's inevitable. Day to day, these things sort of swing up and down. Long term, it's really only moved in one direction with some brief setbacks here and there over the last eight years. I think it's continuing.
Great. I have to ask a question. What are your enterprise expectations? I was with Michael Saylor a couple of weeks ago, and he said with the banking adoption in the sector, Bitcoin is going to $5 million. Curious what you guys think over the next year, three years, five years. Jeff, let's start with you.
Yeah, I mean, look, I think I kind of said the direction has been kind of up and to the right long term. It's been obviously very volatile in recent times, but I think what you just made about adoption, I think that is really what is going to take us to the next level. I mean, people talk about Bitcoin as digital gold, but digital gold or gold has been acted more like a safe haven asset than whereas Bitcoin is really seen as much more highly correlated with tech stocks and other risk-on assets. Historically, we've seen when you have broad markets or global uncertainty, sell-offs in the market, in the broader markets, Bitcoin sold off alongside of that. Now, April was, I think, the first kind of divergence from that where we actually saw a lot of global uncertainty, a lot of macro factors coming in.
Key markets were down. I mean, gold was up 5%, but Bitcoin was up 15%, 14%. Small sample size, but I think that is what Bitcoin needed to sort of take it to those $150,000-$200,000 level is to kind of have that broader use case, broader adoption, and kind of a change of view on the asset as a whole.
Tyler, Gary, big adventure of you.
Come on, guys.
I love my strategically. I do not know if I would be sitting here if I went in front of my board of directors and said, "Hey, Bitcoin's going to be at $5 million. So let's plan a business on that." It is really hard to plan a business around that type of optimism. Personally, I believe that very little might see that in our lifetime. Look, this is a hard business to forecast over the next 12 months, let alone the next 36 or 60 months. We operate our business from a very conservative view. Our internal models have it going slightly higher where it is at now. Everything from there is just upside. I think that is the proper way to run a business, just some cautious optimism. If it gets to $200,000-$250,000 in a short time frame, that is great. We are well-positioned to manage that.
Great.
The only thing I'd add is sort of pulling on that same thread I had before. To me, trying to figure out the price is always this question of when do we flip from this network adoption story so it looks more like a tech stock because it is like in an early stage. I would still call it early stage given just the magnitude of the total addressable market of every human company and government on earth, theoretically. Still kind of early, but at some point, that network adoption story will flip, and it will be more like digital gold and be more of sort of an inflationary story. I think that's what's so hard to put any number out there. I think Saylor is a little too impetuous for my taste.
I think if you put a long enough time frame on it, you could have any number. It's just that we will probably grow very fast in spurts at times as long as the network adoption story overwhelms the kind of store of value inflation story. Probably when I'm on my deathbed, it will be hopefully, which is a long way from the future, assuming it is. It'll be trading more like a haven asset, gold, yen, Swiss franc, something like that. When that happens, it's hard to say. It probably raises. I think it'll make all-time highs really soon. Then it might zap up to like a $200,000 number because it's sort of flattened out and trade more like a gold bear or does it go to a million before it does? That's really, really hard to say.
Yep. Makes sense. I wanted to ask about the strategy of coddling versus selling mine to coin fund operations. Gary, on your side, I know you've had different policies over the course of the company and recently kind of shifted to selling a portion of coins mined. Can you talk about the drivers of that decision and how you think about that?
Yeah. At the end of the day, we are not ideological about Bitcoin as an asset. We see it as a capital asset on the blockchain. Just a little context, we were really holding 100% of our monthly production starting in November 2023. Here in April, we started really selling a percentage of our production to cover opt-outs. We went on the record a couple of years ago saying that the 100% hoddle strategy is not attainable. We still believe that. We really hoddled during a point in time when we felt that there was going to be major appreciation in Bitcoin itself. We just see this as prudent capital management, right? Bitcoin is our functional currency. We manage to profitability.
To us, if we could pay the power bills and operating expenses with Bitcoin, service the debt that we might have, and increase our hoddle balance, we just see that really as the right combination for our business.
Great. Separately, I know you talked about it a little bit, Tyler and Jeff at the beginning, kind of your AI data center component. How do you view this market opportunity in relation to Bitcoin mining? What is your strategy around it? Tyler, let's start with you.
Sure. I would say the space has evolved over time. I think starting really with the ascendancy of the large language models now roughly two years ago, there has just been an explosion in desire for these large interconnects available sooner. Historically, on the HPC side, a really big data center was 30 megawatts in the Dulles Corridor, something like that. Suddenly, the need for training to have hundreds of megawatts has forced that industry to look to non-traditional places like West Texas, where we happen to have a lot of large interconnects. Our company has a total operating and pipeline of 2.8 GW of power. What has happened is there's been a huge push and interest in these large interconnects available sooner because everyone is racing to have the winning model.
We happen to have angled the company really starting a year ago, leveraging our expertise at developing greenfield sites to aggressively position ourselves for these large interconnects available in the next three years. As a result, a lot of that interest in that market has come to us. We have been very busy speaking to intersealers, the type of tenants that can take down 300 MW at a single site. On the back of this, we have had some announcements. We took a $50 million investment from SoftBank right around the time of Stargate. They had an exclusivity on one of our sites, Bruce Pratt.
We recently announced a financial partnership framework with Investment Group, where they are our financial partner at our Barber Lake site and are backstopping the entire cost of the build of a data center so that when we are speaking to hyperscaler tenants, they can feel comfortable that the big pockets can stand that up. I think one of the things that's difficult for a company that has been traditionally in Bitcoin mining is that's a little bit exotic if you're going to be the partner of a trillion-dollar market cap company that is building a $3.5 trillion data center. I think we're working through the necessary sort of partnerships that we're sitting on what we think are the really rare valuable pieces there. Everyone in that space, the bottleneck is power. That's how we're thinking about it.
Look, Bitcoin mining is a great business as well. They have different characteristics. Over time, I see us having a balance of two businesses.
Jeff, do you have any different?
Yeah. I mean, look, I agree with everything Tyler just said. I mean, we're all kind of looking at this opportunity differently. Some not at all. Some are looking to make a full transition to it, and some are kind of right in the middle. The way we are looking at it is that, look, we have a significant power portfolio right now, 600 GW. Demonetize it all for Bitcoin mining doesn't necessarily sound like the most prudent thing to do. Looking to diversify some of that revenue into other areas, which will ultimately attract multiple, given the stability of the revenue, diversify the base. To us, what we're really looking at is we have a way to ultimately reduce the cost of capital for what we consider our core business, which is going to be the ASIC manufacturing.
We are, like you said, having the right expertise on your team to be able to execute on a data center like this, much different from Bitcoin mining. I think a lot of people are finding that out now. We made the decision to work with a strategic partner, both a development and a financial partner, because hyperscalers are looking for people they've worked with before, people they know that can execute on a project of this scale. It's not just about having the megawatts. It's about surrounding that with the right team and the right capital providers.
Interesting. You mentioned your own Bitcoin mining machine. Can you talk to us about the Seal Miner and your rationale for the project and what advantages you think it brings?
Yeah. I mean, the rationale is this is technology development is really in our DNA. I mean, if you look, most people know our founders, Qihan Wu, who is also the founder of the team. Our Chief Strategy Officer is a guy by the name of Harris Astrid. He also has a pretty deep semiconductor background. He actually developed the current IC design that's basically used in every Bitcoin mining rig since 2014. It is kind of derivative of that that his team put together. It is really part of our team's background. I think it gives us a different moat around our business than just the pure play Bitcoin mining. It also gives us a lot of strategic advantages. Just on the mining side, we can produce ASICs for our own self-mining fleet at a meaningful discount to the rest of the market.
We can also sell them to third parties as well. It is a lot of flexibility, a lot of optionality for us. As far as what is different than others, right now, you could argue maybe not that much because everything that has been done over the last 10 years has really been built on that 2014 architecture. You can certainly operate through that. Most of those advances have come really at the foundry or the fab level. TSMC and others have gotten better at their processes. Every time they come out with a new wafer or thinner wafer, you see a big jump in efficiency, pretty juicy little incremental jumps from there. We are working on a new IC design that we think will sort of revolutionize that. We are hoping to take that out later this year.
Great. For Gary and Tyler, how do you see the market for ASIC mining hardware evolving? It seems you have a pretty dominant market share that's sustained itself over time. Does this bring concentration risk in terms of suppliers? How do you think about tariff risk? What does this market look like going forward to you, Gary?
Yeah. So there's a number of players now that are coming to market. I'm looking forward to seeing the bigger machine come out here shortly. But we're primarily Bitmain right now. Bitmain, it's kind of like Tom Cottage, right? Like no one ever got fired for buying IBM. Bitmain's just been very solid machines, and they've been very successful in having the majority of market share. But we obviously welcome competition in the marketplace. There's one miner that's apparently going to be manufactured onshore. We'll evaluate all of the miners. We think that the days of $90 a terahash aren't at least I don't foresee that happening. Sorry, Jeff.
No, no.
Anytime soon. That is just about competition, right? We think that it is going to keep the cost of CapEx a little lower. In terms of tariff risk, I mean, it is a whipsaw, right? You got to be able to plan for it. We comply and implement miners from China contractually. You need to make sure that they are manufactured in Malaysia. It is appealing to have miners manufactured in other areas, but you never know what is going to go on with this administration and where the tariffs are going to land. Ultimately, you got to have access to capital. A 10% tariff, if you are really worried about that, you got a problem with the business model and capitalization of the company. I think that is one of the things that I feel really comfortable and I hope the business people are right. It is like we will capitalize.
We have access to a decent cost of capital. If there is that tariff risk, we can address it head-on or we can delay purchases, which ultimately, it is a level playing field, right? Because it is going to apply to everyone. If the tariffs are so significant that it reduces the purchases, that just is going to keep a lid on global hash rate. Absolutely. I think we have always taken the approach that certainly Bitmain has the most market share, and you have to have a relationship with them. They make good machines. There is a reason why they have that market share, and we are a key client of theirs. We have always tried to diversify and look forward to also sustain and perhaps finance and Seal Miner in the future. Our fleet, traditionally, we have worked with MicroBT and Canaan as well.
I think we've always viewed there being a lot of risk with those companies because they're privately held. They were Chinese or are Chinese in origin, so you couldn't always get the best picture of their financials. Earlier in the business several years ago, one of the risks was sort of like you send this giant bag of money across the world and hope these machines show up in six months. It definitely caused me and our CFO quite a few sleepless nights early in those expensive dollar per terahash days. I think we've always tried to stay sort of diversified among the manufacturers just to sort of protect against any particular company risk. Those have all done pretty well. We always want to find other manufacturers. I think what's also unique for us, our sites are in Texas. It's a hot place.
It's a harsh environment sometimes. Some machines that often have the best specs for perfect environments are also the most finicky in harder environments. It's not as easy as just kind of like, "Give me the one with the best horsepower because maybe it's only the best horsepower in perfect conditions." You can tweak the conditions, but that also costs more CapEx. It's sort of a constant ROI evaluation battle to figure out where you want to strike the balance on those factors. That will always be thrown into our analysis. I think to that end, tariffs are challenging in the sense that, and it's not unique to our industry, it's the uncertainty that they cause. It's not so much the numbers or the fact that you can't plan for a 10% tariff. There is a tweak and you have a 34% tariff.
I think that's the challenge because we have always been really focused on being pretty strict in planning on what's the payback look like, what's our ROI on this investment when we make a choice of machine and timing and so forth. Certainly when it's sort of you wake up the next day and again, the cost spikes on your largest cost CapEx of 34%, it's not going to cause financial distress so much as, "Wow, that really changes how I feel about the ROI, and maybe we would have thought timing differently or whatever." I'd say right now, the landscape with 10% across the board tariffs outside of China is fine, very easy to plan around.
I would say, and it seems like the administration's on this path, it seems like they have gotten kind of broader message on industry that clearer sort of projections about where things are going to go and maybe more measured steps make it easier when you plan projects. We're an increasing crucible for industry in general. We make these decisions and we have a data center coming online in a month, Black Pearl, this month. That's been an 18-month, 16-month dream project. Part of that is planning on when you make these big payments that you've across the world to Asia. In the interim, when you get uncertainty on what the input costs are, that's a challenging environment to operate. Hopefully we continue towards that kind of like, "Here's the level. Here's where it's going to operate." It seems like it's going to be set up.
From a technology perspective, you mentioned the conditions. How important are some of the technologies we talk about in this industry, like liquid cooling versus air cooling?
From our perspective, again, we tend to go back to the numbers and think of it as a series of levers. If you pull this one down, the other three move in different levels. There is not kind of one correct answer across the board, which is why I think every company has a bit of a different approach. We have tended to focus on air cooling historically. What that has meant is that we, in general, are trying to plan for a lower CapEx there. There may be implications on what we pay for OpEx because it is a harder operating environment for the machine, and maybe the specs on the performance of the machine are a little bit worse. When you think about Cipher's positioning, again, we have the lowest cost of power, which drives like 90% of our OpEx.
The way we've thought about moving the levers is we can deal with a little bit higher cost of ongoing OpEx in terms of how the machine performs or repairs that need to be done because it matches up with our low power cost. Let's save the money on CapEx on the front end. We think about ROI on the total project. When we think about it and we've looked at it, and we probably will be doing projects in liquid cooling. Again, this is a broad generalization. The way you think about the sort of return on investment is if I move more of this CapEx piece up front to fund it, do I kind of back in the performance over time with a better performance and operating? I do not know that there's one right answer. For us, it's really about sort of maximizing that ROI.
Jeff and Gary, do you have a similar difference here?
Yeah. I mean, obviously, we have a view from both angles, obviously, as a seller and a user. But liquid cooling solutions are definitely going to play a big role. Hydro cooling, immersion cooling. Like Tyler said, I mean, it's a quick turnaround conversion cycle as well too because the refreshing your fleet, you don't know where technology is going. Two years from now, it could be another jump from there. Making a big CapEx investment on the front end for immersion or hydro cooling and knowing that you're going to have to refresh that fleet maybe in two years, maybe three years, maybe even get four years out of it. Where does that efficiency kind of curve land during that time? You need to be competitive. I think we are kind of entering a much more power-constrained market than we had in previous years.
Where if you want to grow, sites were available. I'd argue they're still available, but they'd be more difficult to find and scale and likely more expensive. There is a balancing act. Do more from what you have versus growing and kind of balancing the CapEx and the OpEx and looking at a total ROI.
Great.
I'll just add that really 30% of our new rollouts are going to be emergent. We've had a 20-megawatt site operating in the Greater Atlanta area now for probably ongoing almost three years. We've become very in tune with running an emergent facility. The cost of the CapEx has also come down to almost the same amount per megawatt building an air cooling facility. It is really going to come down to, as Tyler said, the ROI, right? Can we completely scrap a site and put emergent? ROI is not going to look as good as if we have just raw dirt there and build from the ground up. Particularly with some of the innovations that are going on, emergent now where they actually have mobile units. We can have those manufactured off-site. We can see you run copper wire. Maybe throw down a concrete slab.
The timing and the expectation and construction time is going to get compressed, which we really like. Also, the great thing about immersion is that the immersion project we have, like we have some running in the state of Wyoming right now, pretty much put on a pad, seal it with immersion liquid, mount the machines, lock it up, put a fence around it, maybe have a security guard, and then you can control it from another data center or our NOC headquarters. It is very labor-light. It is such a stable environment that you really do not need to take machines out of the liquid.
Again, we really see that this is really the evolution of cooling because the chips can only get so thin and so much space in between the chips, as you know, which means that the fan air cooling can only be so much that can be spent that heat. So immersion is where we think is the industry's going, and we're leaning heavily into that.
Great. Tyler, you mentioned Texas. In terms of geographic footprint, it does seem like many mining roads lead to Texas. What makes the state so well-suited to Bitcoin mining?
Yeah. I think one of the things that investors maybe do not have a full appreciation for when you are evaluating Bitcoin mining is the value of this power usage in that it is really the only large-scale user of power as an industry that is so instantly curtailable. At first, I think most industries do not pay much attention to the value in that. A large part of what Cipher is, and I would argue distinguishes us from our competitors, is realizing the value of that curtailability. Again, I am sure most of this room probably knows this, but power grids need to be stable. You need to manage supply and demand, and that becomes complex. To overgeneralize and oversimplify this, in Texas, it is its own power grid, and it manages supply and demand through price signals.
You can have prices be virtually free or even negative to balance the grid for large parts of the time and then see power prices spike $1,000 in five minutes. If maybe it's a day when there's no wind and the sun's going down and it's summer and everyone's running their air conditioner, there's tons of demand. That price signal is how to get more supply incentivized to come on and demand to come off. For us, it's the distribution of those prices that, if you look at the average price in Texas, it's still competitive, but there are other states that are cheaper if you're just going to put power 100% of the time. The way we operate our entire fleet is we never run our fleet 100% of the time.
We monetize this component so that the distribution of prices in Texas are such that 5% of the time, prices are really high and maybe not economical for Bitcoin mining. The other 95% of the time, they might be near zero. If you just chop off kind of 5% of uptime, you can drastically reduce that price. That enables sort of knock-on things that you can hedge the power. You can sell power back to the grid. You can trade connections to move power around the grid, which all is kind of in our DNA. We have a lot of Wall Street at Cipher, and hedging and managing that power is a huge part of what we do. Texas is a unique market for that.
Gary, I think you have a different view in terms of your geographies.
Yeah. We operate in four states: Georgia, Mississippi, Tennessee, and Wyoming. Everything Tyler said about it being a curtailable load is right, and it's one of the distinct advantages of Bitcoin mining. When the grid needs power, we can shut down in a matter of one or two minutes. Our approach is we really like to focus on rural America and partnering with these rural communities. We bring a lot of capital to these communities through taxes, sales taxes on capital, as well as emergent. We're talking populations of 10,000-15,000 or less, the number of communities we operate in. What's important is that in our demand response programs that we have and our agreements, at least most of the agreements, maybe 100-120 hours a year, we might have to turn down because we need that power for their local citizens.
Those are times that we would not run anyways, right? Because power prices might go up 5-10X. It is $0.50 a kilowatt hour, $1 a kilowatt hour. We are not going to be profitable anyways. We turn down. That allows the grid and the local municipality really to keep the prices as low as possible because we are run through that. We have to now import power, which means that they are probably paying higher prices anyways if it takes them out. It is really citizens that end up getting burned at the end of the day. Plus, if we were allowed the municipalities to keep most of that margin rather than importing that from elsewhere, that could account for 20%-30% of our cash flow or profitability per year. It is significant dollars, and plus we see this as a partnership, right?
We can go in, we can get power at the cheapest rates possible, and then be completely curtailable once possible to allow in their share of profits.
Great. I know we're running out of time, but I definitely want to ask one question on geography for you, which is, what do you think about international geographies? I know you have operations outside of the U.S.
Yeah. Look, we see a lot of value in it. I mean, obviously, we're in Texas. We're in Tennessee. We're in a lot of the U.S. markets right now, but we are probably the most globally diversified miner. We've got 50 megawatts that's out in Norway, and then we have another 600 coming on in Bhutan. We did just transfer to Africa, acquired a 50-MW site in Ethiopia. So we see a lot of advantages in having geographic diversification. I mean, one just kind of played out more recently here is when we had the tariff situation. We were getting ready to ship a lot of miners to Texas, and the tariffs backed up, and we were going to essentially reroute those to Bhutan and Norway and really just kind of reslotting them in there.
I think it gives us a lot of flexibility beyond just the energy price and the energy. Looking at stable political regimes is important to us. I think there's going to be a lot more international mining in the future.
Great. We are out of time. Thank you to all three of you for joining us here. It was a very interesting discussion.
Always.