Good afternoon, everyone, and thank you for joining us for the 26th Annual H.C. Wainwright Global Investment Conference. My name is Dylan Scales. I'm an associate here on the Crypto and Blockchain research team. I'm pleased to introduce Chief Executive Officer of Bitfarms, Ben Gagnon. They are a leading US-based Bitcoin miner, currently operating 11.3 exahash as of the end of August. Yeah, bring it over to you.
All right, thank you very much. Thank you, H.C. Wainwright, for having us. Also like to thank Andrew, CFO of MicroStrategy, for opening up for me. That was very nice. So Bitfarms is one of the largest and oldest publicly traded Bitcoin mining companies on the planet. We've been operating since 2017. We currently have just shy of 2% market share spread across 12 sites in four countries, and we've got some pretty exciting growth plans, where we're going to be expanding to about 3% market share, we anticipate, in Q1 of next year, so we are a company that really focuses on operational exceptional performance. We're focused on growing the company in strategic ways in order to maximize our return on invested capital, and we're not growing for growth's sake.
Although right now, we do think it is a very strategic time to be growing, so we do have some quite exciting growth plans that I will be talking about with you today. First and foremost, we have a plan here to diversify the company beyond Bitcoin mining itself. When you look at the strategic advantages that companies have when it comes to mining and they operate in a highly commoditized market, you need to operate on the tail ends. You cannot operate in the middle, doing exactly what everyone else is doing, paying exactly the same price that everybody else is paying in order to get a competitive advantage, so what we have outlined is a strategic opportunity here with the company Stronghold, which we acquired, or we made the announcement that we're acquiring them on the 21st of August.
We expect this transaction to close in Q1 of next year, and that enables a lot of strategic diversification with the company. First, we're able to vertically integrate by acquiring two strategically owned power plants in Pennsylvania. That enables us to take greater control over our number one cost element in the company, which is our cost of power. Number two, we're expanding and rebalancing our energy portfolio with over 300 megawatts of near-term capacity in the United States, in our favorite energy market in the United States, which is called the PJM market, which is the largest wholesale electricity market in the US. Three, PJM enables tremendous wealth of energy trading and demand response opportunities that help us to minimize and balance those energy costs as well, providing us even greater control over that number one cost variable for us.
Four, it enables a very unique Bitcoin scaling opportunity with HPC and AI potential, which we'll be going into, which is very unique to having the generating assets, as well as the ability to trade power on the grid through the local substations. Five, we are getting access to specific technologies, which help us go beyond just minimizing our impact on the environment, but actually providing a positive impact on the environment and doing so profitably for shareholders, and also helping us to reduce, again, that very important element, our cost of power. Six, we're accomplishing all of this in a highly accretive transaction that is purely done on a stock basis, which enables us to preserve our balance sheet for future growth potential, and this is really going to help us continue on the momentum that we've had for this year.
We've outlined a tremendous plan starting in November of 2023 , that we've been executing on, which effectively we are tripling our hash rate this year and improving our energy efficiency by about 40%. These relative growth metrics are almost unparalleled in the industry in terms of both the increases in our hash rate and the increase in our energy efficiency, which is what's driving that increase in market share, and is also helping us compete across the entire space and regardless of the macro conditions. With the acquisition of Stronghold and the acquisition of the Sharon, Pennsylvania, site, which was announced a few weeks prior, we have a very robust pipeline to 35 exahash and beyond in 2025 , which will help us carry this momentum that we've been building throughout 2024 over the next year.
We anticipate that the threefold growth can continue to grow about 70% next year, and we have a tremendous pathway here to continue to provide investors with high-quality upside exposure to rising Bitcoin prices, with a low-cost basis and a quickly rising market share. In addition to growing our hash rate, we are also growing our energy portfolio significantly. We currently have a little over 300 megawatts online, and we have a pathway with the Stronghold acquisition and our natural organic growth opportunities here to grow to 955 megawatts at the end of next year. This is a significant increase and almost threefold increase in our power capacity in the next about 15 months.
Beyond that, Stronghold also provides us with the opportunity for a multi-year organic growth plan beyond the nine hundred and fifty-five megawatts to over one and a half gigawatts. This has a lot of potential beyond just Bitcoin mining itself for many other use cases, including HPC and AI. The expansion of our portfolio is also a rebalancing of our portfolio. We have a very international portfolio of international energy assets. Right now, we've got a huge amount of growth that's currently taking place in Paraguay, although we do have growth taking place in Canada, Argentina, and in smaller extent, in Washington state as well. With the acquisition of these three sites, we're able to rebalance our portfolio towards the US and North America.
Next year, with 955 megawatts, we should have almost 50% in the US and two-thirds in North America, reducing that LatAm exposure to about one-third of our entire portfolio, and also enabling significantly greater control over our price of energy. The opportunities around the Stronghold sites, in particular, are largely focused around two particular strategies. The first one is the more immediate opportunity around the upgrading of the Bitcoin mining itself. The site has the potential to effectively quintuple the Bitcoin mining operations in 2025 through two different means. One, I'm not sure if this is laser. Oh, wait, look at that! Currently, there's about an 80-megawatt power plant on each site, and there's an 80-megawatt substation, which historically, the power plant has sold 80 megawatts to the grid.
With the development of the Bitcoin mining data center, they now have the ability to either draw the power from the grid or from the power plant. What we are doing is we are going to be islanding the power plant and drawing the power from the grid, effectively doubling the capacity of those sites with almost no capital expenditure. The substation is already in place, the power plant's already in place. The permission to do this is effectively a done deal. What we need to focus on now is building out the second Bitcoin mining data center so that we can effectively double the power capacity at these sites. The second part of the growth plan here around Bitcoin mining operations is their fleet is rather outdated and is due for an upgrade.
Just like we've done throughout 2024 with our transformative fleet upgrade, we have the opportunity to dramatically increase from the four exahash that they have now to between around 20 or 22 exahash with the kind of high-end miners that are available on the market right now. This ability to increase from 4 to 20, plus exahash is pretty unique to these sites because of the ability to double down on the energy infrastructure that's already in place. So this gives us a robust pathway to effectively double our potential growth potential from the 21 exahash we'll end this year, with an additional roughly 21 exahash that's possible at these two different sites. Beyond the Bitcoin mining, opportunity here, there's also something very, very unique and strategic with HPC and AI.
The traditional data center build is you draw power from the grid, and you spend a bunch of money to buy diesel generators and turbines in order to provide you a very, very small amount of redundancy and reliability. Bitcoin mines generally can save a lot of money by not having this redundancy and this reliability. This is why we're so much more cost-effective than a Tier 3 data center. But Tier 3 data centers, HPC and AI, they need this level of redundancy. When you're looking at what is the cost to build out a Tier 3 data center, the numbers vary, but I think the good ballpark estimate is about $10 million a megawatt, plus or minus a few million. One of the largest single-line items in that equation is the redundant aspect that is equal to about 10 basis points or 20 basis points of the year.
So these substations can provide power for 99.8%, 99.9% reliability, but that 0.1% and 0.2% reliability is incredibly expensive. What we can do here, and have the ability with these sites, is we can actually build out the HPC data center, provide the primary base load power plant from the power plant itself instead of the grid. Then we can save anywhere from $3 to 5 million a megawatt by not investing in the diesel infrastructure in terms of the gen sets and the turbines, the diesel storage, all of the electrical connections, in order to provide that redundancy that's only used for 10 basis points or 20 basis points of all the time in a year. Massive, massive cost reductions for us to build out HPC and AI infrastructure in our portfolio. In addition, diesel is incredibly expensive to operate.
In order to generate energy with diesel, you're looking at kind of a minimum $0.15 to 0.16 per kilowatt- hour. By doing so with the grid connection here, what we're able to is actually have a more stable data center, where we're able to monetize this redundant connection with the Bitcoin mining piece. So instead of having a 30% to 50% of your infrastructure allocated to something that's only going to serve you 0.1% to 0.2% of the time, we can now save all that cost, and we can actually monetize it through the Bitcoin mining data center, saving us costs on the CapEx, but also improving the operational revenues that we receive out of the site, and also the operational expenses that we have on the site.
The other thing that is a benefit here for having this merge data center is that the HPC data center is not a reliable, consistent load. The HPC data center load consumption is based on when a customer actually wants to use the HPC and AI, so it's dependent on the customer submitting work orders and instructions. The Bitcoin mine, we can operate at 100 megawatts consistently forever, assuming the economics make sense. But with an HPC and AI data center, the load is a lot more variable. So in addition to being able to monetize the redundant connection, we can also overclock or underclock the Bitcoin miners to consume that extra load, provide a more stable base load, and also improve the operational efficiencies and the costs on those sites. This also makes the site actually environmentally cleaner, right?
Because we do not actually have to spend all of the money, all of the carbon that goes into building out the diesel footprint, having the gen sets, having the turbines, all of the infrastructure that goes into that, we're able to actually power the data center with the Tier 2 alternative energy power plant that we acquired, and we're able to provide the redundancy instead of from diesel from the grid, which is increasingly decarbonizing and providing a lot more nuclear, wind, solars, hydros, all the other stuff, which is a lot cleaner than the diesel. This is a very unique opportunity that's specific to these plants. When you look at the universe of Bitcoin miners who are getting into the HPC and AI space, I think you see a variety of different strategies.
One, you have companies who already had GPUs from mining Ethereum, which are old, outdated, and really not what most companies are looking at for HPC and AI. You have, two, the companies who are looking to invest in the GPUs, and then you have the companies who are just specifically looking to invest in the infrastructure. When you're looking at the cost to invest and bring HPC and AI into your portfolio, the costs vary a lot from Bitcoin mining. With our growth plan that we've announced this year, to build out one megawatt of infrastructure and compute, it costs roughly $1 million. That's the substation, that's the data center, the miners, the logistics, the duties, the labor. It's roughly $1 million a megawatt. But to build out that Tier 3 data center infrastructure, it's roughly $10 million a megawatt.
If you want to add in the GPU compute, $30 to 40 million. So it's a significant capital expense that is not really being properly understood by, I think, a lot of the companies who are looking to get into this space. What we are seeking to do is achieve a competitive advantage over the hyperscalers and the traditional data centers through the combination of the power plant and the Bitcoin mining data center combination. Doing so gives us better return on invested capital, lower capital needs, and better operational expense and revenues out of the site. In order to do this, we do have quite robust liquidity that's funding this growth. Our growth plans for the remainder of this year are fully funded. We've got ample cash and Bitcoin, and we also sell the Bitcoin that we mine out of our operations quite systematically.
We are able to fund the business with our business. We do not need to dilute the business in order to fund the operational expense of the business. We're able to do this because we are quite efficient, both from an operational perspective, but also from a corporate overhead perspective. Additionally, we do something that's quite unique in the Bitcoin space. We have something called a synthetic HODL strategy. When you're looking out at the universe of publicly traded Bitcoin mining companies, there is generally two camps: you either sell every Bitcoin that you have, or you HODL every Bitcoin that you have. There's obviously pros and cons to both strategies, but when you are HODLing every Bitcoin that you have, and you're not able to actually fund your business out of your revenues, you can see how unsustainable that constant dilution is.
What we do is we actually fund the business through our revenues, and then we buy back the long-dated exposure on Bitcoin with a call option. So, we are able to both fund the business but maintain the upside exposure on the Bitcoin for anywhere from six to 11 months on those instruments. This enables a lot of strategic capital efficiency. Let's play it out in kind of three different scenarios. Let's say Bitcoin price stays flat. Okay, Bitcoin price stays flat, we've continued to fund the business. The option expires effectively worthless, but we've funded the business, and we haven't incurred the dilution that had no gain in the Bitcoin price. Say, the Bitcoin price goes down. We've now monetized our Bitcoin at significantly higher levels, funding our business and our, our growth.
The option would expire worthless, but we funded the Bitcoin, or we funded the business, and we monetize those Bitcoins at higher levels, preventing us from losing the value from selling the Bitcoin at a lower price in a worse market. Option three, Bitcoin price goes up. We've now funded the business. We've now profited and made upside from the long-dated call option that we've bought in the Bitcoin price, and the options function in an exponential manner, so just buying one call option is going to actually improve in value on a percentage basis, significantly more than the Bitcoin itself, so we're able to very cost effectively run our business, maintain the upside exposure to Bitcoin that all investors want, and protect us and operate, I think, a little bit more efficient in all three different market scenarios: Bitcoin sideways, Bitcoin down, Bitcoin up.
Summary, this means that we are incredibly well-positioned as a company for growth in twenty twenty-four and beyond. We are making continued improvements to our operating profile via our fleet upgrades, our portfolio expansion with the Stronghold acquisition, the Sharon acquisitions, and even today, we had an announcement where we are, we've brought on a few more highly skilled people, and we are making changes on the organizational structure, which are helping us to continue to improve our operating performance, provide greater management and accountability throughout our organization, and also lay the foundation for integrating both power generation and HPC and AI. We leverage low-cost power, very efficient hardware, and operational excellence, and strong financial positioning in order to generate industry-leading yields, whether that be per exahash or gross mining margin, or return on invested capital.
We have a continued focus on diversifying beyond just Bitcoin mining, both geographically, but also through new technologies and new revenue streams, which don't take away from Bitcoin mining, but provide accretive and compelling synergies which make us better Bitcoin miners, and lastly, we have an incredibly talented team of operators with a very strong track record of over seven years, including navigating two halvings. We have one of the best operational teams out there, and, you know, our operational metrics over the last four or five years, I think, will also confirm exactly that. With that, I'd like to welcome our CFO, Jeff Lucas, onto the stage to answer any questions that we might get from the audience.
Actually, two questions. In terms of the trajectory of your power growth, are these fully deployed, meaning you have built out the infrastructure, or these are just the power is in place, but you still need to implement the build out of the infrastructure, which could theoretically take longer? And the second question is, you have two slides that talk about how you connect to the PJM grid, right? There's one that PJM serves as a redundancy, which I get it for the HPC operation. But the one, the slide before, you talk about that will literally double your capacity in, for example, Bitcoin mining.
Mm-hmm.
In that scenario, will you be subject to curtailment?
So yeah, two great questions, and let me just kind of flip back to those slides so I can speak to them earlier. So, there's a, the first question is about where we are right now with our power portfolio and where we're going. We're right now here at this roughly 300 megawatts. If you include the Stronghold portion, you have this 142 portion right here. You include the expansion that we're doing here at Stronghold, and it would be this 165, which adds up to 307.
What we're also doing right now is we're building out a site in Yguazu, Paraguay, which has got 100 megawatts. That's going to be scheduled to be online by the end of this year, 200 megawatts for the following year, which is going to add up to. It's going to add one hundred out of this one hundred and thirty delta. The additional power capacity that adds up on this 440 is a small expansion at our Washington site by about 3 megawatts, a small expansion at our Argentina site from 54 to 62 megawatts, as well as an additional expansion at our Baie-Comeau facility by about 11 megawatts. So between the sites that we currently have active, it's about 310.
If you include what's currently active at the Stronghold, it'd be 310 plus 142, for a total of 452. If you include what's currently being built out for the end of the year, it would be this 440 plus the 142. For all of next year, we're looking at this 955. So it's a mixture of stuff that's already built out, and still needs to be constructed. But when we're talking about this 952, this is power that's secured, right? So the biggest bottleneck to growth in our industry is always power, and it's timeline to energization. There's always two different, you know, factors with that. One, you've got to actually get the regulatory approval, which is actually the most timely part of the process, and two, you've got to actually build out the construction.
So, with this 955 , we have this fully secured. A big portion of this is built out or is currently being built out. And we've got a proven team of operators and builders who have been doing this for seven years. We know exactly how to build substations, how to build data centers, and how to operate them. On the two different scenarios that I've laid out here, when you have a Bitcoin mining data center that's currently just tied to the power plant, there would be no curtailment or economic demand response on this portion, but there would be on this portion, right? So about half of the site, you'd have economic trading around the energy itself.
In order to enable this structure, you have to do what's called islanding the power plants, so there's no connection anymore between the power plant and the grid. So we wouldn't have the ability to sell the power into the grid, but we have the ability to buy power either when it's profitable to do so, we have the ability to hedge power, we have the ability to trade power, we have the ability to do all sorts of different structures around the trading the power and doing the economic demand response, which helps us bring that cost of power lower. The scenario that we have here with the HPC and AI, this is very, very similar, but we wouldn't get economic demand response. So in order to facilitate having the redundancy, we can't commit that we're going to turn off the power at any point in time.
But what we'd still have the ability to do is we'd still have the ability to buy power on the Day-A head Market. We'd still have the ability to avoid those expensive peaks, where the power price is very high and would be unprofitable to operate. And we'd still have the ability to, if we did buy the power on the day ahead or the week ahead or whatever market it is that we're operating on, and the power price peaks, we'd still have the ability to sell back the power that we previously purchased.
Unfortunately, it looks like we're out of time. But, Ben and Jeff, I want to thank you so much for being here today.
Okay, great.
Hopefully, everyone has a productive rest of the conference.