CleanSpark, Inc. (CLSK) Q4 2023 Earnings Report, Transcript and Summary

Good afternoon. My name is Krista, and I will be your conference operator today. At this time, I would like to welcome everyone to CleanSpark's Fiscal Full Year 2023 Financial Results Conference Call. [Operator Instructions] And at this time, I would like to turn the floor over to Isaac Holyoak, Chief Communications Officer. You may begin.

Thank you, Krista, and thank you for joining us today for our Fiscal Full Year Financial Results Call, covering the period October 1, 2022, through September 30, 2023. Our press release was issued about 30 minutes ago and is available on our website at www.cleanspark.com. So we expect to file our Form 10-K tomorrow. Today's call is also being webcast, and a replay and transcript will be available on our website. I'm here with Zach Bradford, our Chief Executive Officer; and Gary Vecchiarelli, our Chief Financial Officer. Keep in mind that some of the statements we make today are forward-looking and based on our best view of the world and our businesses, as we see them today. The statement and information provided remains subject to the risk factors disclosed in our most recently filed Annual Report. We will also discuss certain non-GAAP financial measures, concerning our performance during today's call. You can find the reconciliation of non-GAAP financial measures in our press release, which is available on our website. And with that, it is my pleasure to turn the call over to Zach.

Thank you, Isaac. It's a pleasure to be here with you all today. This past fiscal year has been one of substantial growth and success for CleanSpark. We've not only met but exceeded many of our strategic aims, and I'm proud to share these accomplishments with you. It was a record-breaking year for us financially and operationally. We posted revenue of over $168 million, a 28% increase over the last fiscal year. We expect to continue to follow this trajectory, and we are optimistic about revenue growth next year, even as we prepare for the halving. For example, during the last month, we posted our first single day revenue of almost $1 million. Like analysts across the traditional financial industry, we are bullish on Bitcoin. We believe that its utility as a payment system, investment opportunity and savings tool for the middle class, will continue to grow as more and more people take part in this next stage in the evolution of money. Operationally, our hashrate growth year-over-year has been exceptional. Highlighting our commitment to not just growing, but also scaling efficiently. In fact, we've achieved a remarkable milestone this year, by surpassing a total hashrate of 10 exahashes per second. A level reached only by a select few in our industry. Our hashrate has more than doubled since I spoke with you 1 year ago during our last fiscal year earnings call. What's more? Our hashrate is among the most efficient among publicly traded miners, thanks to one of the most advanced fleets in the industry. Our fleet-wide efficiency is now 26.4 joules per terahash, and we expect to continue to see sustained improvements in our efficiency with Sandersville coming online in the next few months. I'll talk more about this in a bit. But for now, let me emphasize that this growth is a testament to our team's hard work and the strategic investments, we've made in our technology and infrastructure. We've kept pace with the Bitcoin's network soaring hashrate and the results of our efforts were a record 6,902 Bitcoin mine for the first fiscal year, almost doubling what we mined last year. We mined that Bitcoin with an average gross margin of 42%. I'm also pleased to report that we are entering the new year with the strongest balance sheet in our history. As of this call, we have almost $170 million in liquidity, in cash and Bitcoin Holdings. I'd like to talk about our strategy around holding more Bitcoin on our balance sheet. We see Bitcoin as a smart treasury tool, and we expect Bitcoin to continue to increase in value over the near term. We plan to be opportunistic about how we use that Bitcoin for future growth opportunities. You'll notice that we grew our balance swiftly, over the last few months. We did this to strategically boost the balance, as we sense price improvement was on the horizon. We expect to continue to grow our Bitcoin treasury, but we have shifted back to a strategy of using our Bitcoin for expenses and holding the difference. We plan to maintain strategic flexibility and how we make capital decisions with our Bitcoin. As we head into halving, I expect us to lean into a balanced strategy of maximizing our holdings, while also using Bitcoin to fund our operations and therefore, avoid using equity to fund operating expenses. Seems like a long time ago now, but it was at the beginning of our 2023 fiscal year that we acquired our Sandersville location. More recently, we acquired our Dalton campus. We now have sites dotted across Georgia. Proudly, many of them are in rural communities, where we can have an outsized impact on the economic engine of the community and the power purchasing strategies of the respective utilities. We continue to maintain good relationships with our hosting provider in Upstate New York and see that partnership as part of our approach to a diverse operating framework through halving. The acquisitions of Sandersville and Dalton were pivotal in our growth strategy. These acquisitions have allowed us to expand our reach and capacity. In the case of Dalton, it has allowed us to further diversify our utility providers, as the city has its own independent power department. We are making significant progress on the expansion of Sandersville, and I will share more about Sandersville later in my remarks. It's exciting to see these plans materialize -- contributing to our goal of keeping our status as one of America's leading Bitcoin miners. Earlier this year, during summer, our expansion in Washington went live, marking a critical step in our strategic efforts to diversify and strengthen our operational footprint. This facility not only enhances our production capabilities, but also exemplifies our commitment to sustainable and responsible mining practices. The expansion added 50 megawatts of operational infrastructure, for a total of 86 megawatts. Which is currently capable of around 6.5 Bitcoin a day. In line with our growth strategy, we've aggressively expanded our machine inventory. Specifically targeting the most efficient miners. This year, we continued our proven track record of buying state-of-the-art machine, equipment at market setting prices. The result is that we have increased our operational efficiency and output. These machines are not just an investment in hardware, but in the future of CleanSpark. We've been particularly strategic in our acquisitions, purchasing over 80,000 machines, including 57,500 S19 XPs. Those XPs have since all been delivered and many of them are waiting to be racked and energized at our Sandersville facility. We also have an additional 22,000 S21s with delivery expected to begin in January 2024. As we look beyond this quarter, post Sandersville expansion and into halving, we plan to aggressively focus on incremental gains in performance and outperform our peers with these gains. We also plan to continue to strategically upgrade our machines and optimize our software ensuring that we stay ahead and at the forefront of the Bitcoin mining industry. These enhancements improve our efficiency and profitability. I want to take a moment to define what we mean, when we discuss efficiency of our operations. We measure efficiency at multiple levels. The most important though, are energy efficiency and uptime efficiency. Energy efficiency is the amount of energy it takes to produce a Bitcoin. Staying at the forefront of energy efficiency, gives us a meaningful advantage. Even now, we are producing up to 37% more Bitcoin, with the same amount of energy at some of our less efficient peers. Uptime efficiency refers to our ability to maximize the available shelf space and power at our sites. We limit the time in which a miner is off-line or a rack is unused. We have increased uptime by pursuing the following initiatives: building out our internal repair capabilities. Max optimizing our software stack and training our teams to rapidly address down miners. We also maintain a small safety stock of extra miners at each site to allow for rapid slots of miners. If a minor needs to be taken off the shelf for repairs, we put a new miner in its place, instead of leaving an empty slot. Earlier this week, we achieved 99.87% uptime, across all of our wholly owned sites, meaning that 88,710 of 88,825 available slots were hashing. Efficiency is crucial, as we head towards halving. We use software to make our machines more efficient. For example, we can under clock our machines, whereby we can choose to reduce their computing power, by an average of 27%. While reducing their energy consumption by almost half. This is the equivalent of mining 3/4 of a Bitcoin at a little more than half the energy consumption and cost, as it would take under normal conditions. To put another way, this is like mining 1.26 Bitcoins with the same amount of energy that it would normally cost to mine just 1 Bitcoin. We also use software to optimize or overclock our machines. Generally speaking, overclocking means that the machine consumes more energy to increase hashrate. We can actually increase the machine hashrate by about 12% with no increase in energy consumption at all. To build on my earlier math, it's like mining 1.12 Bitcoin, at the energy consumption levels of what you would normally see, when you mine just 1 Bitcoin. In an immersion cooling environment, we can push those efficiency gains even further. We expect to see our efficiency continue to improve into halving due to these initiatives and the deployment of the new S21 Bitcoin mining machines, the most efficient miner currently available in the market. Once installed, we expect our hashrate to exceed 20x a hash per second. Timing has always been a key part of our strategy, and I believe we have done it better than anyone. We will continue to time our strategic purchasing and growth to seek out the best returns for our shareholders. Subsequent to the fiscal year-end, we've continued our expansion, acquiring new machines that align with our long-term strategies and growth goals. This ongoing investment in our infrastructure is key to keeping our competitive edge and driving sustainable growth. Our recent acquisitions and machine purchases, made strategically during the bear market, have positioned us exceptionally well for the upcoming Bitcoin halving. We are scaling up our operations in anticipation of the next Bitcoin halving, ensuring that our infrastructure is robust and efficient. Our team is working tirelessly to enhance our operational capacity, and we are excited about the potential this expansion brings. Principally, we are excited about achieving scale. We expect our corporate overhead to remain flat. Going forward, while substantially growing our operations. Internal analysis suggests that the financial effects of this are large. It's on par with lowering our all-in cost by almost a full $0.01 per kilowatt hour. For example, we can increase machine inventory, power under management and hashrate without having to make proportional increases in overhead, both in terms of office space and headcount. Our principal cost driver is the cost of our energy. We expect to continue to expertly manage this cost, while keeping overhead down, thus driving more value to our shareholders. In terms of our energy acquisition strategy, we've been meticulous and strategic. Our goal has always been to strike the right balance between cost effectiveness and sustainability. We've adopted innovative approaches to energy consumption, focusing on investing in communities that have access to low-cost energy sources. Most of these energy sources are renewable or low carbon, like Nuclear Energy. Access to Nuclear Energy is one of the principal reasons we've made Georgia, our home. There is abundant energy in Georgia. And is a net exporter of energy. We are thrilled to be able to work with rural communities throughout Georgia, to keep that energy in the community. This approach not only aligns with our commitment to environmental stewardship, but it also ensures long-term operational efficiency and resilience. We consume a lot of energy as a Bitcoin miner, and we are committed to being wide stewards of that energy. We view our consumption as essential to securing and maintaining a robust decentralized mining. Despite substantial growth over the last year, we continue to see most of our energy mix being derived from clean energy sources, showing our dedication to responsible mining practices. As we look to the future, the expansion of our Sandersville facility is a key focus. Once completed, Sandersville will be the largest Bitcoin mining data center in Georgia and we believe in the entire Southeast. The site itself will be capable of producing over 6 exahash per second and will hold about 45,000 machines. For those of you who have followed along with our monthly updates on YouTube, you'll notice that the site has seen sustained progress. Most of the subsurface infrastructures like conduit, trenching and concrete is in place, while the teams are making substantial progress on our racking business and installing racking. We expect the site to be complete in the coming weeks, ready for full energization shortly after that. Sandersville holds a special place in our heart and in the community, and we look forward to introducing the site to you, when it's completed and energized early next year. I was recently in Sandersville where we have 4 different trades working side by side in this rural community, working full-time jobs that they would not otherwise have access to. The economic impact of growth is sometimes unappreciated in our current macro environment. But for a town like Sandersville, growth like this is crucial for the town's ongoing success. After the expansion is complete, many of those trades will go on to other opportunities, but a robust workforce of CleanSpark employees will remain. While the city benefits from millions of dollars a year in tax revenue and margins on the electricity we buy from the community. After Sandersville, we expect to need up to 72 megawatts of additional capacity to support our growth to over 20 exahash per second with the S21. We expect nearly half of this capacity to come through facility expansions at our existing sites. With that, I'm pleased to announce that we are making efforts to address the first 15 megawatts of this need by adding 15 megawatts of capacity at our Dalton campus. Once fully expanded, that campus will host an additional 4,320 latest generation S21 machines, capable of 850 petahashes per second. We also plan to pursue adding up to 27 megawatts of additional capacity at other existing sites. All these plans, including the Dalton expansion, are subject to further approval, by expanding at our own properties, we expect to gain 42 megawatts or approximately 60% of the power we need to exceed 20 exahash per second without halving to acquire or build new sites. We are currently reviewing additional opportunities that would more than cover the remaining 30-megawatt need. We expect to achieve 20 exahash per second in 2024, after the full delivery of the S21s we've recently ordered. We may also push beyond 20.4 exahash in 2024, but such growth will be done in a measured way as we approach halving. We plan to continue pursuing our M&A strategy on a post halving basis, at the right time and at the right value. In a way, these incremental additions now seems small after just talking about adding 150 megawatts, all at once in Sandersville. But this move is strategic to how we see ourselves growing, large mega sites can only be one part of the Bitcoin miners portfolio. Instead, we see the future of our campus growth to be more distributed, minimizing the risks of geographic concentration, while maximizing the strength of local knowledge. With that in mind, we expect to continue to expand throughout Georgia. Where we are local power experts. But we also expect to expand into new cities and diversify across multiple utilities. The nature of these expansions are likely to be more measured, as we near the halving in April. I want to take a moment to thank our incredible team at CleanSpark for dedication, innovation and hard work have been instrumental in achieving the milestones, we have discussed today. It's your efforts that drive our successes, and I'm proud to lead such an outstanding group of professionals. I'd like to end with a shout-out to 3 groups of people. First, thank you to the teams, who are working to build our largest site in Sandersville. You can't get that level of cooperation without clear leadership and commitment to the project. Second, I want to thank our burgeoning repair team, at our Norcross campus. The ability to make repairs in-house down to the chip level is a game changer. It means machines spend less time unplugged and more time hashing, directly contributing to the Bitcoin network and to shareholder value. Lastly, I want to thank the finance team at our corporate headquarters. Preparing fiscal year-end financial results over the holiday season is not on anyone's shortlist, not even a CPA shortlist. Thank you for your efforts. With that, over to Gary.

Thank you, Zach. It's my pleasure to share some insights in our financial performance for the fiscal year ended September 30, 2023. As Zach mentioned, it's been a year of significant achievements and our financial results reflect this. Diving right into the numbers, our revenues for the year were $168.4 million, an increase of approximately $37 million or 28%. This increase was primarily driven by the increase in our Bitcoin production. We produced over 6,900 Bitcoin for the fiscal year, compared to 3,750 Bitcoin produced for fiscal year 2022 which represents an increase of 84%. However, Bitcoin price saw much volatility during the year with a low of approximately 15,500 in our fiscal first quarter, late last calendar year and a high of almost 32,000 in our fourth quarter. Additionally, global hashrate has increased significantly. While our hashrate has more than doubled in the recent fiscal year, the combination of Bitcoin prices and increased difficulty has resulted in a muted overall increase in revenues. Our gross profit for the year saw a decrease in 2023 of approximately $15.5 million or 17%. The largest contributor to this decline was the average Bitcoin price, which was significantly higher last fiscal year. Looking at our GAAP net loss. We saw a loss of approximately $137 million for the fiscal year 2023. It is important to note that our net loss includes noncash GAAP items, the largest of which is a charge, we took in the fourth quarter of $32.7 million, related to the acceleration of depreciation around our older miners. As Zach has discussed today, and on prior calls, we are preparing our fleet for the halving event next year. This acceleration of depreciation relates to older, less efficient machines that we have either taken out of service, as we replace them with more efficient and powerful XP machines, or units we currently have in use, but we do not anticipate using post halving. This led us to taking a conservative approach to write down these machines, as their estimated useful life to us was significantly reduced to either 0 or in some limited cases, 7 months, which is the estimated time from our year-end to the halving. Also included in our net loss number is approximately $4.4 million related to discontinued operations, the majority of which includes a full reserve related to assets held for sale. We expect that in future periods, there will be no further costs associated with our legacy Energy business, as we have now fully exited that business. We've included in our 2023 numbers any future costs related to the wind down of the entities. A total cost, which is estimated to be less than $1 million. Finally, with respect to adjusted EBITDA, we did see a decrease year-over-year for reasons very similar to those, which affected our gross margins. However, I want to point out that we did have some fourth quarter costs, which were excluded as favorable adjustments to EBITDA, the largest of which is related to noncash impairment of Bitcoin. As we have discussed in prior calls, this impairment expense is required under current GAAP accounting. For the entire year, we recognized a total $7.1 million impairment expense, of which $6 million is directly attributable to the fact that we increased our Bitcoin balance significantly during the fourth quarter. To help put this in perspective, we held 529 Bitcoin, as of June 30 and finished our year-end with 2,240 Bitcoin, as of September 30 which is over 300% increase in Bitcoin Holdings just in the fourth quarter alone. While we remain subject to the accounting rules around impairment of our Bitcoin, we look forward to the FASB adopting the proposed mark-to-market rules, which will better reflect the true economics of our operations. I want to take a moment to discuss our power costs for the entire year. Our power cost for fiscal year '23 were $0.048 per kilowatt hour. As you will see here, our wholesale electricity costs alone were $0.036 per kilowatt hour, and we had an additional $0.12 per kilowatt hour of cost, related to transmission and distribution, profit margins to the cities we operate in and taxes. Management considers its all-in power costs, which include these fees and costs on top of the wholesale electricity numbers, when analyzing its operations. We believe that we have one of the lowest all-in power costs amongst miners in the industry. And when combined with our uptime, CleanSpark is one of the top miners in the industry. With respect to the fourth quarter, we did see higher cost of $0.051 per kilowatt hour, compared to the third quarter, which we saw cost of $0.041 per kilowatt hour. This was in line with our expectations, as we have seasonally higher temperatures in July and August, which leads to higher power costs. Additionally, our team did run miners during periods with higher than normal power costs. However, this was done intentionally as the increase in Bitcoin prices allowed us to still mine profitably during those periods. We continue to have an active Power Management strategy which takes into consideration the 3 variables of current Bitcoin price, difficulty and power costs. This strategy allows us to remain responsive to the needs of the grid, while remaining flexible and profitable. I also want to point out our cost to mine Bitcoin. Our average cost for fiscal year '23 was $12,473, for our wholly owned facilities and $15,797 at hosted facilities. This is important to note because given our world-class operations, the economics of running our own infrastructure, continues to be significantly better than hosting relationships. We will continue to have an infrastructure first strategy, while using hosting relationships when it makes sense. Additionally, I want to point out that hosting costs will typically be higher, as they include a profit-sharing component which reduces margins, compared to our wholly owned locations, especially as Bitcoin price increases. On a final note, I want to talk about our balance sheet and our liquidity position. When I look back on fiscal year 2023, I see that our company took advantage of the bear market to grow our hashrate by more than double and we also shored up our balance sheet during that time. As previously mentioned, we have increased our Bitcoin holdings over 300% in the fourth quarter alone. And that number as of today, is just shy of 2,600 Bitcoins, representing almost $100 million in value. We also have approximately $29 million in cash, as of September 30. And as of today, we have over $70 million of cash in the bank. Combined, this represents approximately $170 million in liquidity between our Bitcoin holdings and cash as of today. From my perspective, we have set the stage for a successful 2024. And are very healthy going into the halving with multiple tailwinds at our back. Not only do we have significant liquidity, but we have one of the industry's most efficient fleets at an average of 26.4 joules per terahash. We have a low cost to mine Bitcoin and we have 150 megawatts producing at least 6 exahash coming online prior to the halving. That 6 exahash, represents an additional 60% processing power above and beyond what we currently have plugged in. And it will drive our fleet efficiency to 25.2 joules per terahash. And of course, we are also excited about the potential Bitcoin spot ETF approval will bring as we view it as a positive catalyst for the industry and Bitcoin price. Again, our size, scale and strong balance sheet have set the stage for the halving event an event we know will not only create some challenges for miners, but also create opportunities for miners such as CleanSpark, as we have the ability desire and liquidity to once again take advantage of opportunities in the marketplace. Until the halving, we remain laser-focused on execution and look forward to bringing over 6 exahashes per second online in Sandersville. As we move towards exceeding 20 exahashes per second. With that, I'll turn the call back over to Isaac to open the floor for questions.

Thank you, Gary, for that detailed financial overview. We'll now open the floor to questions from the analyst community. Operator, please provide instructions and manage the queue for the Q&A session. Thank you.

[Operator Instructions] Our first question comes from the line of Mark Colonnese from HC Wainwright.

I guess I am Mark now. First one for me is on your HODL balance. I appreciate all the color Zach that you provided on your go-forward strategy. But I was curious if you can walk through a little bit more about how you've been able to grow the balance so rapidly in recent months?

Mike, good to hear from you. When we approached it, we really -- we were expecting Bitcoin to go up, and we were proven correct on that. And we also looked at our cost to mine Bitcoin. And as Gary just talked about, it's a little over $12,000 of Bitcoin for us to mine. And so we made a choice. Just like MicroStrategy bought Bitcoin, we exercised the value of leverage. So instead of being like MicroStrategies and buying spot, they just disclosed today buying a lot of Bitcoin at over 36,000. We effectively funded some of our operations on a temporary basis from equity. And as a result, it has essentially allowed us to acquire Bitcoin at that cost of -- a little over $12,000 a Bitcoin. So it was a short-term strategic way to boost the balance and then we shifted back to, again, looking to maximize really the output from our operations by using Bitcoin for what we do. Now what we really expect is to continue to maintain very healthy margins. And on the go forward, what that looks like, I'm very bullish on what ETFs and what halving ultimately means for Bitcoin price appreciation. And I'm really seeing us HODL-ing the delta between what we need to keep the lights on and pay our power bills. And that difference, I see going into the HODL balance for future appreciation.

That's really helpful. I appreciate the additional color there. And second one for me, you also mentioned in your prepared remarks that construction to support the expansion at Sandersville is tracking the plan and expected to be completed in the coming weeks. Can you provide some more specifics as to when you expect to have the site fully energized and hashing at full capacity?

Yes. We're really excited about the progress, we're making. As I mentioned in the prepared remarks, we expect the construction side of that to be done in the next few weeks. Really, we expect it to be done before the end of the year. That really leaves the power coming online and getting the machines plugged in and hashing. What we expect, as of right now, is for us to begin energization in February. And we think it will take a few weeks to basically bring all the miners online, just like it did in Washington. Obviously, this is not like flipping a light switch. It's instead like turning on several hundred Walmarts. And with all that power that goes into it, we turn it on transformer by transformer. So it will come on by 2 megawatts at a time. And we expect to be able to turn on multiple transformers each and every day starting February.

Your next question comes from the line of Josh Siegler from Cantor Fitzgerald.

Congrats on all that you've achieved this year. I think it's a testament to your strong execution there. First of all, I'd love if we could dive a little bit more into the 15-megawatt expansion at Dalton. Do you have any parameters around the CapEx requirement for that first leg of infrastructure growth?

Yes, I've got a little bit around that. We expect it to actually be pretty low cost. And so, I would anticipate that our all-in CapEx needed to do that expansion is under $5 million. Obviously, we've already taken care of the miners. We've disclosed what that looks like. But from a facilities infrastructure side, we do expect it to be under that $5 million mark to finish the $15 million -- or 15-megawatt expansion.

Perfect. That's very helpful. And then on the energy side, I was wondering if you can give us any update on how energy prices have been trending throughout this winter in Georgia and any major trends or anything you'd like to highlight as we progress into '24?

Yes. How it's trending right now is it's trending meaningfully better than it did 1 year ago. So 1 year ago, if you recall, we were -- everybody was experiencing still quite a bit of pain, when it came to natural gas prices. So if you look at our first fiscal quarter last year, prices were trending from a wholesale basis around $0.05. So our all-in cost was floating right around that $0.06 range. As of right now, we're seeing those in the 3s. And we're really happy with the trajectory and the projections. So right now, with some of the projections that we've been provided by the power providers, we expect the balance of the year to kind of continue on trend between that $0.03 and $0.04 range. And then for '24 and beyond, we're actually really excited about what we see for '24. A lot of the utility companies did quite a bit of purchasing to shore up what this year would look like, with the natural gas issues that they encountered a year ago. And so we're looking at a largely -- at least we're projecting out as are the utilities, largely flat prices, that are going to be in the $0.03 range and oftentimes lower than that. So 2024 is looking really, really good in Georgia for power prices.

Your next question comes from the line of Greg Lewis from BTIG.

Zach, I was hoping you could talk a little bit more about some of the cost savings maybe we're seeing as kind of what brought in the machines to internal repairs and the improvement of uptime. Is there any kind of way you guys are able to quantify that benefit that the company is getting?

Yes. What we're seeing right now is with our internal repair team, we can actually turn between 600 and 1,000 machines every single month, without having to send to an outside source. And if you think of it in context of 2 ways, from savings on the repair side. One is when you send it out, you're losing meaningful time, sometimes weeks, before that machine ever comes back to you from another party. But also, what we found is, when you send it to a third party, you see markup not only on parts and materials, sometimes 100% markup on that. And these are often small dollars. It could be a $20 instead of a $40 part that has to get switched out. But then also, we're seeing meaningful differences in labor. Obviously, when you send it to a third party, they have their own labor costs, but they mark it up to make a profit. And so I don't necessarily have a specific percentage, but it's very, very meaningful. I'm going to kick it to Gary to see if he has more context on that.

Yes, I'm give you a little bit more color. We're still in the process of really ramping up our internal repair shop. And so really, those synergies have not been recognized fully in our financial results yet. So -- because we were previously outsourcing that. But we expect in future quarters that will meaningfully help the margins.

Just to quantify the other piece, from an uptime perspective, if you look at some of the independent reports that exist out there, there's the average uptime across the industry is floating around in the 80%, 85% range. Obviously, unless someone discloses that, we don't have all the details, but that's what it seems to be from an industry standard of uptime. And us, consistently achieving in the high 90% uptime. Part of that is what turns into our returns on that, which is ultimately measured in top line revenue versus some specific cost savings metric. So that's why I would look at it as we're probably 12% to 15% or more, profitable due to our uptime metrics because of what we have developed internally.

Yes, 100%. I was looking for a little bit more clarity on the you kind of called out the $72 million -- I'm sorry, the 72 megawatts of expansion potential -- just kind of in terms of the timing, you highlighted Dalton but beyond that, is that -- should we be thinking about that? Is that more of a first half event? Or is it with the halving coming, we're going to maybe look at doing that that's kind of maybe part of our second half '24 kind of growth build-out? Just a little more color around that. And then really, I guess, a 2-part question. How are you thinking about splitting off that are deploying that extra megawatts across the facilities beyond Dalton?

Yes. So we do expect it to be a first half event because our intention is to basically add megawatts so that as we receive the S21s, we can plug them in and get them hashing right away. So for example, the Dalton expansion should happen very rapidly. And then some of the follow-on expansion, I would still expect to happen in the first half of -- calendar half of 2024. And then from...

And similar to Dalton, it's going to be like minimal -- and similar to the Dalton it seems like these are all going to be a couple of million dollars here, a couple of million dollars there.

Yes, yes. We expect -- our intention is to make a fairly low-cost endeavor. We've gotten even better at building. And so we -- depending on what we build, whether it is air cooled or immersion and we are eyeing some additional immersion going forward, but you're talking about anywhere from 350,000 megawatt to around 500,000 megawatt, which I should add is substantially better than the immersion that we've built in the past. The markets improved substantially for that. And we're pretty excited about what Immersion seems to bring in the coming year.

Your next question comes from the line of Brian Dobson from Chardan Capital.

So you've done a tremendous job decreasing joules per terahash over the past 6 months. And you mentioned that new machines could bring that metric down to about 25 joules per terahash. What time frame are you thinking about for that further reduction?

Inside the first half of next year. So we're going to see a meaningful improvement into the 25 range, when Sandersville comes online next calendar quarter. And then with the 21s plugging in, which, as I mentioned, we do expect to be shelfing those in the first half of the year. That should -- will bring it down into the 24 joules per terahash.

Yes. No, that's impressive. And I guess, as you're thinking of just adding those new more efficient machines, what other kind of efficiencies can you wring out of the fleet? Is there more to be had on the software side of things?

Yes. When you look at what we -- those numbers we are using that doesn't include, that's based sticker value of what this machine can do. So what our focus is, is on that flexibility. So as I mentioned in the call, we could really make that 27% more efficient from an energy use, on about 2/3 of the fleet and also take everything up about 12%. And, so that's really the flexibility we think we can continue to wring out of it. In addition to, there's still a small portion of the fleet that is included in that number, that 24 number that we would consider to also upgrade, which would likely -- and I do expect we will execute upon that. We're just going to be very cautious in try and time that from a capital side that would bring it under 24 watts or joules per terahash.

Yes, very good. Thinking about the halving next year, you mentioned that, that should create some challenges for weaker players, obviously, and opportunities for yourself. How are you thinking about M&A as we head into '24?

Right now, we have really interesting opportunities already because there are weak players that know they're weak and that are looking for things. But I think it's all about timing it correctly. So we're incredibly interested in M&A. We've been very focused on site by site, and we continue currently, if you're asking me sitting here today, we're focusing on the site level. A few reasons, you avoid any baggage that often comes with entities and mergers. How I'm going to see that on a post halving basis though, I really think that there's going to be opportunities to not only merge with weaker players, but I think there will present itself opportunities to merge with equally or appear to as strong of players. And I think that, that's important, and that's why our focus really is going to be on the coming year, is really on maximizing our market value because when it comes time, when and if it comes time to have discussions like that in '24 and '25, we, of course, want to be the bigger, stronger, more valuable player in that discussion. So there's a lot of ways we're looking at it. And so that's why for the next 2 years, it's going to be all about maximizing value.

[Operator Instructions] We have no more questions in the queue at this time. I will now turn the call -- I'm sorry, we do have a question. Your next question comes from Reggie Smith from JPMorgan.

This is Charlie on for Reggie. Thinking about the remaining 72 megawatts needed to reach your 20 exahash target, it sounds like 30 megawatts will come from new sites. Can you talk a little bit about what you're prioritizing in that search? Would that be the cheapest power? Would that be maximizing uptime, would that be staying in Georgia? Anything there would be helpful.

Yes. Our first priority, of course, is low-cost power that is reliable. Those 2 go together in making sure that we have high uptime at low cost. We're, of course, interested in growing in Georgia, but if additional opportunities present themselves, we're certainly not going to take the borders of the state and call that a hard and fast line, certainly not. So it really is that focus on low-cost power reliable. And the other thing we're prioritizing is speed to deployment. We, of course, want to deploy those S21s, the time of receipt. We don't want to be sitting on those. It's always been our strategy to be able to plug things in as quick as we can after we get the miners in hand. So that's going to be the other piece that we prioritize is, is it going to be up and functional in the time lines that we want it to be.

We have no further questions at this time. I will turn the call back over to Isaac Holyoak for closing remarks.

Thank you, and thank you to everyone for joining today's call and to our analysts. Thank you for your questions. We appreciate your interest and your support in CleanSpark. As always, we're committed to driving value for our shareholders and advancing our position as a leader in responsible Bitcoin mining. Happy holidays from all of us here at CleanSpark.

Ladies and gentlemen, with that, we'll conclude today's conference. We thank you for attending. You may now disconnect your lines.