CleanSpark’s October 9 update reported unaudited September production of 529 bitcoin, average operating hashrate of 36.0 exahashes per second and operational hashrate of 50 EH/s. CleanSpark’s September update
A production total records an outcome. Hashrate measures computational speed. Efficiency describes the energy required for that work. Reading them together can identify questions worth asking, but turning one into a verdict about the entire business takes more evidence.
Start with the measurement window
CleanSpark defines operational hashrate as the highest combined rate historically achieved by installed, functional miners supported by energized infrastructure and capable of contributing to mining. Its August disclosure uses that definition alongside a separate monthly operating average. August update and definitions
Dividing September’s average by the historical maximum gives 72%, our calculation. Establishing uptime would require matching the fleet, its speeds and the measurement period.
Deployed miner counts answer another question. Converting a machine count into computational capacity requires knowing the machines’ speeds, as the hardware specification below illustrates.
For any operator, readers should preserve the company’s exact labels when comparing installed, deployed, energized and average operating capacity. Treating them as interchangeable makes a comparison look more precise than its inputs support.
Normalize production by time
August production was 593 bitcoin, with average daily output of 19.12 and average operating hashrate of 38.3 EH/s. September’s reported daily output was 17.64. August figures, September figures
Our calculations show declines of 10.8% in total production and 7.7% using the reported daily averages. September had 30 days against August’s 31. The daily comparison accounts for that calendar difference.
Network difficulty is one variable. Bitcoin adjusts its proof-of-work target every 2,016 blocks using the elapsed time of the preceding interval. A harder target requires more hash attempts on average to find a valid block. Consequently, the same computational effort can have different expected production across difficulty periods. Bitcoin’s technical guide to difficulty
Those protocol mechanics explain why difficulty belongs in the analysis. They do not establish how much of this particular monthly change it caused.
Pool arrangements add another layer. Bitcoin’s mining guide explains that pools distribute proceeds from block rewards and transaction fees under different reward systems. Readers comparing miners should check the payout method and fee treatment before assuming equal computing power must generate identical reported receipts. Bitcoin’s mining guide
Check what uptime actually counts
Mining-software provider Braiins distinguishes between system uptime, when a machine is reachable, and hashing uptime, when it is producing hashes. Its management documentation explains that a reachable machine can stop hashing because of software trouble, overheating or deliberate curtailment. Braiins’ uptime definitions
The economic meaning depends on the cause. Avoiding expensive electricity or participating in demand response can justify a pause, according to the same provider. A lower operating average therefore calls for site-level information about interruptions, settings and energy decisions. These are possible explanations to investigate, rather than findings about CleanSpark’s September operations.
Follow efficiency into the accounts
CleanSpark’s August disclosure labels its efficiency figure a peak measure for the deployed fleet. That label matters when comparing it with any measure averaged over a month. August fleet metrics
Energy efficiency is usually expressed in joules per terahash. BITMAIN’s S21 Pro specification provides a concrete example: 234 TH/s, 3,510 watts and 15 J/TH, with stated operating conditions and tolerances. Dividing watts by terahashes per second produces the efficiency figure. BITMAIN’s S21 Pro specification
For illustration, a hypothetical machine maintaining 100 TH/s at 15 J/TH would draw 1.5 kilowatts. Running for 24 hours would consume 36 kilowatt-hours, costing $1.80 at an assumed five cents per kilowatt-hour. This is an energy-only example, excluding other equipment and charges, and is not an estimate of CleanSpark’s costs.
Company accounts add the remaining layers. CleanSpark’s results for the quarter ended June 30 separately reported $85.48 million of cost of revenues excluding depreciation and amortization, plus $111.04 million of depreciation and amortization. Payroll and other expenses also had separate lines. Those are quarterly accounting categories, unsuitable for labeling September’s electricity bill. CleanSpark’s fiscal third-quarter results
The October release says monthly operational updates will end as the company moves to quarterly reporting. Reporting change
A reusable reading checklist is straightforward: record each metric’s definition, identify the period it covers, and keep the comparison on the same basis. Then examine the electricity assumptions and the expense categories omitted from any unit-cost figure. Those steps help readers move from a headline to questions the underlying disclosures can actually answer.