20% of Bitcoin mining capacity idle; restarts may squeeze margins
About 20% of Bitcoin mining capacity — about 235 EH/s — was idle in August due to uneconomic rigs, curtailment, transit and maintenance. Restarting that hashrate would raise difficulty and lower earnings per unit.
Mining services firm Luxor estimated about 235 exahashes per second of Bitcoin mining capacity was idle in August. The firm classified offline machines as uneconomic, deliberately curtailed for grid or transmission charges, in transit, or under maintenance.
Luxor derived the figure by comparing roughly 1,150 EH/s of total net ASIC capacity with about 915 EH/s of activity implied by August’s average mining difficulty. The 235 EH/s gap represents equipment present but not contributing to block production during the month.
Revenue improved in August. Luxor reported the dollar-denominated hashprice rose 24.4%, from $31.63 to $39.33 per petahash per second per day. Bitcoin’s price rose about 24.5%, from $62,889 to $78,312 in the firm’s data.
Less efficient machines saw the largest relative improvement. Luxor’s measured fleet consuming 25–38 joules per terahash averaged about $45 per megawatt-hour in August, below the firm’s estimated $48/MWh network-average power cost. That tier exceeded the benchmark on 11 days.
Some operators power down machines for reasons other than insolvency. In Texas, miners reduce activity during the four coincident peak intervals that run through June to September to avoid high transmission charges; that seasonal window ends in September. Other machines were offline for transport, installation or repairs. Luxor also noted some capacity may have been reassigned to artificial-intelligence or high-performance computing workloads but did not quantify that share.
Bitcoin’s protocol retargets mining difficulty every 2,016 blocks, about every two weeks. If additional computing power shortens average block times over an adjustment period, difficulty rises. Higher difficulty reduces expected Bitcoin rewards per unit of computing power when block rewards and fees remain constant.
Blocks averaged 9 minutes and 34 seconds in August, faster than the 10-minute target. The September 5 difficulty adjustment increased 1.31%, reflecting rising hashrate during the measurement period.
Signals that will show how much capacity was temporarily offline include a sustained change in smoothed hashrate, subsequent difficulty adjustments and public disclosures from mining operators about curtailment and restarts. Those data are needed to separate machines offline for grid or price reasons from equipment out of service for longer-term financial distress.
Articles by this author
