What happens when three publicly traded Bitcoin miners release their September numbers on nearly the same day and the results pull in different directions? One company mines fewer coins and sells more than it produces. Another watches cloud mining shrink while its own machines do better. The third simply holds the line after months of decline and quietly turns waste heat into warmth for thousands of homes. These are not abstract industry stats. They are concrete operational snapshots that reveal how miners are adapting to thinner margins, shifting business models, and the constant pressure to stay relevant in a market that rarely stands still.
September Production Numbers Reveal a Clear Split
CleanSpark produced 529 Bitcoin in September. That figure sits almost 11 percent below the 593 coins it mined the month before. The drop reversed a modest uptick that had appeared in August and left the company’s month-end Bitcoin treasury at 13,530 coins, down from 13,703. BitFuFu came in at 167 Bitcoin, a four percent decline from 174. Canaan, by contrast, reported exactly 44 Bitcoin—identical to its August total and the first time in three months that production did not fall.
These numbers arrived through the companies’ own operational updates. They matter because production is the purest measure of how effectively a miner converts electricity and hardware into digital assets. When output slips, investors start asking whether the cause is higher difficulty, machine downtime, power constraints, or deliberate strategy. In my view, the September data points to a mix of all three factors, and each firm handled the situation differently.
CleanSpark’s Sales Exceeded What It Mined
CleanSpark did more than report lower production. It also disclosed sales of 2 Bitcoin on the open market and the delivery of another 700 Bitcoin through the exercise of call options. The average price realized across those transactions was $72,928. Add the two figures together and the company moved 702 Bitcoin while only mining 529. The net result was a treasury reduction of 173 coins—the exact amount by which sales and deliveries outpaced production.
That kind of imbalance is unusual in a single month. Most miners prefer to accumulate when prices are soft and sell selectively when liquidity is needed. CleanSpark’s choice suggests either a deliberate treasury-management decision or contractual obligations that matured in September. Either way, the headline production number alone does not tell the full story. The company’s balance sheet moved more than its mining floor did.
Alongside the numbers, CleanSpark announced it would stop publishing monthly mining updates. Future production and treasury figures will appear only in quarterly reports. The change coincides with the firm’s expanding focus on data-center operations, including a large campus under construction in Georgia. For anyone tracking the stock, the shift means less frequent visibility into the core mining business. Shares were trading near $8.76 at the time of the update and remained more than 24 percent lower year to date.
Financing the Georgia Data Center Push
CleanSpark closed a $2.276 billion senior secured notes offering to fund construction at its Sandersville site. An earlier proposal had floated notes worth roughly $2.227 billion due in 2031. Proceeds are earmarked for building costs, reimbursement of prior equity investments, and the creation of debt-service reserves. The notes carry a first-priority lien on most assets of the issuing subsidiary and the property company that owns the Georgia campus. Importantly, they do not convert into common shares.
The customer agreement behind the project was signed earlier. It covers 175 megawatts of capacity for twenty years with an investment-grade technology firm whose name has not been disclosed. CleanSpark estimates contracted revenue of $6.6 billion over the initial term. That figure dwarfs the company’s recent quarterly revenue of $138 million, which itself was down more than 30 percent from the prior year. The quarterly net loss stood at $239 million. In other words, the mining side continues to face pressure while the data-center side is being positioned as the longer-term growth engine.
I’ve found that miners who successfully diversify into high-performance computing or hosting often trade at different multiples once the market believes the transition is real. CleanSpark is clearly betting that the Georgia campus will change the narrative. Whether monthly production updates disappear because the mining business is becoming secondary or simply because management wants cleaner quarterly storytelling remains an open question. Either interpretation carries weight for shareholders.
BitFuFu’s Mixed Internal Picture
BitFuFu’s overall September production of 167 Bitcoin masked two opposing trends. Self-mining rose more than 11 percent to 98 Bitcoin from 88 the previous month. Cloud mining, however, dropped nearly 20 percent to 69 Bitcoin from 86. Average daily output stayed flat at 5.6 Bitcoin, which means the decline was steady rather than concentrated in a few offline days.
Cloud mining revenue depends on customers who rent hashrate. When those customers reduce their spend or when power issues hit hosted machines, the numbers move quickly. Self-mining, by contrast, is under the company’s direct control. The fact that BitFuFu’s own machines performed better while the hosted fleet struggled points to operational or demand-side problems in the cloud segment rather than a broad hashrate collapse.
The company also flagged uncertainty around a proposed acquisition of a majority stake in an Ethiopian operator that runs an 80-megawatt facility. BitFuFu had already paid $4.1 million and deployed approximately 13,700 Antminer S19-series machines. Closing conditions remained unmet. On September 28 the local electricity supplier issued a notice threatening to seize collateralized equipment unless a disputed payment was received within ten days. Management is now evaluating termination, arbitration, recovery of funds, and the physical return of its machines. It explicitly warned that the deal may never close.
Ethiopia has reduced power deliveries to miners to as little as 23 percent of contracted amounts. That backdrop makes any large-scale expansion in the country riskier than it looked a year ago. BitFuFu’s shares were changing hands near $1.27 and had lost more than half their value since the start of the year. The combination of softer cloud results and a stalled African deal leaves the company with fewer near-term catalysts than many investors had hoped for.
Canaan Holds Steady and Turns Heat Into Value
Canaan’s 44 Bitcoin matched its August total exactly and interrupted a three-month streak of declining output. Stability itself can be newsworthy when difficulty keeps rising and older machines lose efficiency. The company also reported that its 8-megawatt Nordic district heating project is now fully operational. Recovered computing heat is being delivered to roughly 2,800 homes. A similar greenhouse heat-recovery initiative in Canada is approaching planned deployment, pending final site inspections.
Our 8 MW Nordic district heating project is now fully operational, delivering recovered computing heat to approximately 2,800 homes.
That statement from the chairman and CEO is more than a sustainability talking point. Heat recovery turns a cost center—cooling—into a second revenue or cost-offset stream. In colder climates the economics can be compelling once the infrastructure is in place. Canaan is among the first hardware makers to push this model at commercial scale.
On the treasury side, Canaan converted 3,952 ETH and 54.5 Bitcoin into approximately $13.9 million in cash. Part of those proceeds funded a share-repurchase program. The company spent $5.4 million buying back 13.6 million American depositary shares in late August, bringing year-to-date repurchases to about 16.4 million ADSs. Using digital assets to retire equity is a relatively uncommon move among miners and signals management’s view that the stock is undervalued relative to the underlying holdings.
Canaan’s Bitcoin balance rose modestly to 1,875 coins from 1,868. BitFuFu’s balance also edged higher, to 1,388 from 1,373. CleanSpark was the only one of the three to end the month with fewer coins than it started. The divergence in treasury policy is almost as interesting as the production numbers themselves.
Why Production Varied Across the Three Firms
Several forces shape monthly Bitcoin output. Network difficulty adjusts every two weeks and has spent most of 2025 and 2026 trending higher. Older-generation machines become less competitive as the bar rises. Power costs and curtailment events can idle entire facilities for days. Machine deployment schedules and maintenance windows also matter. CleanSpark’s drop looks consistent with a combination of difficulty pressure and possible operational adjustments as the company reallocates capital toward data centers. BitFuFu’s cloud decline appears more demand- or power-related. Canaan’s flat result suggests its installed base is stable and relatively efficient for its size.
Perhaps the most interesting aspect is how differently each firm is responding. CleanSpark is pivoting hard toward long-term contracted computing capacity. BitFuFu is still trying to expand geographically while managing a cloud business that is proving volatile. Canaan is extracting secondary value from the same kilowatt-hours that produce Bitcoin and returning capital to shareholders through buybacks. None of these strategies is inherently superior; each reflects a different reading of where the industry is heading.
- CleanSpark: lower production, heavy option-related sales, shift to quarterly reporting, large data-center financing
- BitFuFu: mixed self- versus cloud-mining results, stalled Ethiopian acquisition, soft share price
- Canaan: stable production, operational heating project, treasury conversion into share repurchases
Treasury Management Choices Speak Volumes
Holding Bitcoin on the balance sheet is both a bet on future price appreciation and a liquidity decision. Selling or delivering coins reduces exposure but can free capital for growth projects or share buybacks. CleanSpark’s decision to move more coins than it mined suggests either contractual necessity or a deliberate reduction in inventory. BitFuFu and Canaan both added a small number of coins while still finding ways to generate cash—BitFuFu through operations, Canaan through selective conversion of both Bitcoin and Ethereum holdings.
In my experience, the market rewards consistency in treasury policy more than it rewards any single month’s production figure. Investors want to know whether management intends to accumulate through cycles or to treat Bitcoin as working capital that can be monetized when needed. The September updates give clearer signals on that front than many previous months.
Data Centers Versus Pure Mining
CleanSpark’s Georgia project is the clearest illustration of the industry’s broader shift. Pure-play Bitcoin mining remains highly cyclical. Long-term leases with investment-grade customers for high-performance computing or AI-related workloads offer more predictable cash flows. The $6.6 billion contracted revenue estimate over twenty years is the kind of visibility traditional miners have rarely enjoyed. Whether the transition can be executed without distracting from the remaining mining operations will determine how the stock trades over the next several years.
Other miners have explored similar paths. Some host third-party machines; others build dedicated AI clusters. The common thread is the recognition that electricity infrastructure and cooling capacity have value beyond hashing Bitcoin. CleanSpark is simply moving faster and at larger scale than most peers.
Geographic Risk and Power Constraints
BitFuFu’s Ethiopian experience is a reminder that low-cost power is never risk-free. When a utility can cut deliveries to a fraction of contracted levels and then threaten seizure of collateral, the theoretical advantage of cheap electricity evaporates. Miners have learned this lesson in multiple jurisdictions over the past five years. Diversification of power sources and legal structures has become as important as the efficiency of the machines themselves.
Canaan’s Nordic heating project sits at the opposite end of the spectrum. By turning waste heat into a local utility service, the company creates a stakeholder relationship that may protect the operation during periods of energy scarcity. Homes that rely on the heat have an incentive to keep the miners running. That kind of symbiotic arrangement is rare and potentially durable.
What the Numbers Mean for the Broader Sector
September’s mixed results do not signal an industry-wide collapse. They do illustrate that individual company performance can diverge sharply even when network difficulty and Bitcoin price are the same for everyone. Machine age, power contracts, customer mix, and capital allocation decisions all matter. Investors who treat the sector as a homogeneous group risk missing important differences in strategy and execution.
The move by CleanSpark to quarterly reporting may also influence how other firms communicate. Monthly updates create a steady stream of headlines and short-term trading catalysts. Quarterly numbers force the market to focus on longer trends. Whether that reduces volatility or simply concentrates it around earnings dates remains to be seen.
Share-price performance year to date has been challenging for both CleanSpark and BitFuFu. Canaan’s buyback program is one response to that reality. Using digital assets to repurchase equity is a form of capital return that pure-equity investors can appreciate even if they are less focused on the underlying Bitcoin holdings.
Looking Ahead to the Next Reporting Cycle
CleanSpark’s next production figures will not appear until the full quarter is complete. That longer interval will make it harder to detect short-term operational hiccups but may also reduce the noise that monthly swings create. BitFuFu will need to clarify the status of its Ethiopian machines and whether the cloud-mining decline is temporary or structural. Canaan will be watched for further progress on the Canadian heat-recovery project and for any additional treasury conversions.
Difficulty is expected to remain elevated. Power costs in many regions are still under pressure. The companies that can either secure long-term contracted demand outside of pure mining or find secondary value streams for their energy use will likely navigate the next phase more comfortably than those that remain fully exposed to Bitcoin price and hashrate competition alone.
I’ve watched this sector long enough to know that one month’s numbers rarely define a company’s trajectory. Still, the September updates from these three firms offer a useful snapshot of the choices currently on the table: sell more than you mine and fund a pivot, struggle with geographic expansion while cloud demand softens, or keep production flat while turning heat into both revenue and goodwill. Each path carries its own risks and potential rewards.
Practical Takeaways for Observers
Production is only one lens. Treasury movements, capital-structure decisions, and secondary uses of energy often matter more for medium-term valuation. CleanSpark’s financing and reporting change signal a strategic shift that monthly mining numbers will no longer fully capture. BitFuFu’s cloud and acquisition challenges highlight the operational and political risks that accompany rapid geographic expansion. Canaan’s heating project and buyback program show how a hardware-centric firm can create additional value without relying solely on higher Bitcoin prices.
None of these stories is finished. The next few quarters will reveal whether CleanSpark’s data-center revenue materializes at the scale projected, whether BitFuFu stabilizes its cloud business or recovers its African machines, and whether Canaan can scale heat recovery beyond the Nordic pilot. Until then, the September figures remain the most concrete evidence of how three different approaches are performing under the same market conditions.
The industry continues to mature. Some participants are becoming infrastructure companies that happen to mine Bitcoin. Others remain focused on maximizing hashrate and managing treasury through cycles. A few are finding ways to monetize the by-products of computation itself. September simply made those differences more visible than usual. For anyone following the space, that visibility is useful—even when the numbers themselves are mixed.
In the end, Bitcoin mining is still a race for efficiency, reliable power, and smart capital allocation. The three companies examined here each ran that race differently in September. CleanSpark prioritized balance-sheet moves and future contracted revenue. BitFuFu managed internal divergence between self-owned and customer-hosted machines while navigating an uncertain acquisition. Canaan held production steady and demonstrated that waste heat can become a genuine product. Those contrasting results are worth watching closely as the year progresses.
