While the Bitcoin mining industry spends 2026 in what analysts have called its first hashrate bear market, one NASDAQ-listed miner just sprinted the other way. BitFuFu (NASDAQ: FUFU) reported on Sept. 3 that its total hashrate under management jumped 45.1% in a single month, reaching 20.6 EH/s in August — and its Bitcoin production surged 55.4% to 174 BTC.
By Michael Nguyen | September 4, 2026
The Hook: Growing While the Industry Shrinks
Hashrate is the total computing power securing the Bitcoin network — think of it as the number of workers competing to process transactions and earn new coins. For most of this year, that power has been drifting lower as struggling operators shut down machines, a trend industry executives have openly described as a hashrate bear market. Against that backdrop, BitFuFu’s August numbers read like they come from a different industry entirely.
- Bitcoin production: 174 BTC in August, up 55.4% from 112 BTC in July.
- Daily production: 5.6 BTC per day on average, versus 3.6 in July.
- Hashrate under management: 20.6 EH/s, up 45.1% from 14.2 EH/s.
- Power capacity: 344 MW, up 34.9% from 255 MW.
- Fleet efficiency: 16.7 J/TH, improved from 18.0 J/TH in July.
Where the Growth Came From
The expansion was not one big machine purchase — it was mostly other people’s machines. BitFuFu’s self-owned hashrate edged up 5.6% to 3.8 EH/s, while hashrate from third-party suppliers and hosting customers rocketed 58.5% to 16.8 EH/s. In plain terms: BitFuFu runs the facility and the service, and clients bring or rent the hardware. That is the cloud-mining model, and it rebounded hard — cloud mining produced 86 BTC in August, more than doubling from 40 BTC in July, nearly matching the 88 BTC the company mined for itself.
“As the additional hashrate capacity secured in June and July came online, our total hashrate under management surpassed 20 EH/s, driving a substantial recovery in Bitcoin production and a strong rebound in our cloud mining business,” said Leo Lu, BitFuFu’s Chairman and CEO, in the Sept. 3 announcement. He added that the growth was achieved “while further improving our fleet efficiency to 16.7 J/TH, positioning BitFuFu among the most efficient operators in the industry.”
Efficiency deserves a quick explainer, because it decides who survives in mining. The 16.7 J/TH figure means the fleet burns 16.7 joules of energy per terahash of computation — like miles-per-gallon for mining machines. Lower is better. When the Bitcoin price is under pressure, efficient operators can still mine profitably while power-hungry competitors sell or unplug.
The Treasury: 1,373 BTC and Counting
BitFuFu’s bitcoin balance grew from 1,314 BTC at the end of July to 1,373 BTC at the end of August. Two footnotes matter for investors reading that headline number: it includes 44 BTC pledged as collateral for loans and miner procurement payables, and it excludes coins produced by cloud-mining customers, which belong to those clients, not the company.
Holding mined bitcoin rather than selling it daily is a bet on future prices — the same corporate-treasury strategy pursued by larger holders, with the same risks. If Bitcoin rallies, the stash compounds the operational gains. If prices slide, an unhedged treasury can turn a good operating month into a bad financial one.
The Core Conflict: Expansion Into a Headwind
Here is the tension worth watching. The wider mining sector has been consolidating, with hashrate falling and marginal operators exiting — conditions that, ironically, reward survivors with a larger slice of the block rewards. BitFuFu’s aggressive expansion means it is betting that its hosting-and-services model, its 344 MW of power under management, and its efficiency gains can outgrow the industry’s contraction. The CEO pointed to “a more constructive Bitcoin market environment” ahead, and Bitcoin traded near 79,400 USD as of the CoinGecko snapshot taken Sept. 4 at 17:55 UTC, up more than 5% on the day.
One more caution for retail readers: these are unaudited, self-reported operational metrics from a company press release, not audited financials. Production numbers are straightforward to verify on-chain, but revenue and profit depend on costs — power prices, hosting contracts, depreciation — that this update does not disclose.
The Verdict: A Live Experiment in Counter-Cyclical Mining
For regular investors, BitFuFu’s August is a case study in two diverging strategies: shrink and survive, or scale into the downturn. If the hashrate bear market deepens, hosted capacity can leave as quickly as it arrived. If the market recovers, the company enters the upswing with 45% more capacity and one of the most efficient fleets in the sector. Watch the September production report and the next quarterly filing — that is where the expansion either converts into earnings or shows up as overreach.
The cryptocurrency market remains highly volatile. This article is for informational purposes only and does not constitute financial advice.
55.4% more btc while difficulty climbs is the stat people skip. everyone else mined less, they printed 174. someone timed those US site energizations perfectly
20.6 EH/s for only 174 BTC shows who the real boss is, difficulty. still, printing 55% more BTC while the sector shrinks is a legit flex, those energizations were timed perfectly
20.6 EH for 174 btc, difficulty is absolutely the boss here. and the growth is contracted capacity booked in june, gaslen called it below
wait for the related party footnote in the next 20-F. hosting revenue booked from your own parent at negotiated rates is the kind of thing auditors circle in red
45% hashrate growth while everyone else is unplugging machines. either brilliant timing on hosting contracts or the balance sheet is doing the heavy lifting. need to read the filings
the growth is contracted hosting capacity booked back in june when nobody else wanted rigs, not spot demand showing up. the 45% jump is real, it was just paid for months ago
the 45.1% jump is mostly new US sites coming online, not organic demand. still, deploying that fast in a hashrate bear market takes guts
market rewarding an actual production number instead of narrative for once. 55.4% more BTC month over month is a real metric, FUFU earning the pop
45% hashrate growth while everyone else unplugs machines. either BitFuFu got insane deals on hosted rigs or someone’s subsidizing the expansion
subsidized is my guess too, their singapore parent locks in hosted deals nobody else can touch. want the power cost per MWh in the next filing
the singapore parent angle is underrated. hosting deals priced like nobody else can touch them, that 45.1% hashrate jump has a cosigned balance sheet behind it
want the power cost disclosure too but fat chance. hosting margins through the parent is where the 45% jump actually comes from, self-mining is barely breakeven
spot on about the parent. bitfufu margin is a singapore accounting artifact more than us mining ops. 174 btc against that fleet, the filing matters more than the hashrate flex
they secured the capacity back in june and july before difficulty spiked, it’s literally in the leo lu quote. timing was everything here
174 BTC in a month is serious revenue at current prices. The real question is what they paid for the energy behind 20 EH/s
cloud mining rebound is carrying the numbers tho. self-mining margins are getting eaten alive by power costs everywhere else
exactly the right question. 174 btc is roughly 13.8m in revenue but 20.6 EH/s burns through that fast if power runs above 6 cents. next filing tells the story
the breakeven math at current difficulty sits somewhere near 5 to 6 cents per kwh. their US sites were signed below the market rate, so the margin exists, its locked not market
locked beats market right up until renewal. those june contracts roll over mid hash bear market and the entire margin story gets repriced in one filing
exactly, everyone flexing the 45 percent like its recurring revenue. hosted capacity rolls over, then FUFU holders meet the actual cost curve
locked rates only look smart till the renewal cliff. june pricing on 2026 power is the whole trade, next 20-F tells you if they hedged it
the 20-F cannot come soon enough. if those june rates were not hedged the renewal cliff resets the whole margin story and FUFU gives back the pop in a day
below 5 cents in the US sounds right for sites signed back in june. anyone paying spot power right now is mining at a loss
174 BTC from 20.6 EH/s feels kinda low no? difficulty at ATH will do that i guess. rip margins
174 split across 31 days is under 6 btc a day for the whole fleet. per EH thats rough, difficulty is chewing everyone equally
FUFU popping on an actual production print in a hashrate bear market. cloud mining contracts are the recurring revenue story here, self mining is barely breakeven across the sector