Roughly 235 EH/s of Bitcoin ASIC capacity is sitting outside active network production, according to Luxor’s August hashrate analysis. That sidelined hardware is large enough to matter as Texas miners approach the end of the summer transmission-charge season and operators reassess whether older machines can profitably return.

The figure should not be read as 235 EH/s of broken miners. Luxor’s analysis groups together ASICs that are uneconomic, deliberately curtailed, in transit or under maintenance. It estimates roughly 1,150 EH/s of net ASIC capacity against about 915 EH/s of activity implied by August’s average difficulty.

235 EH/s Is a Large Reserve of Potential Competition

Idle hashrate is different from retired hashrate. A miner that has been scrapped is unlikely to return. A modern ASIC switched off for a power-market event, moved between facilities or waiting on a repair can rejoin the network when conditions improve.

That distinction matters because Bitcoin difficulty responds to aggregate competition. If a meaningful portion of sidelined capacity returns, the same revenue opportunity that encouraged machines to restart can be diluted by the resulting increase in network difficulty.

BitcoinVersus.tech recently reported that Bitcoin difficulty reached 132.76T, increasing the premium on efficient hardware. Luxor’s estimate shows how much additional specialized equipment may still be waiting outside the active network.

August’s Hashprice Recovery Changed the Restart Math

Luxor says dollar-denominated hashprice increased 24.4% during August, from $31.63 to $39.33 per PH/s/day. Bitcoin’s price rose 24.5% over the same period, meaning the BTC rally supplied most of the revenue improvement.

But older equipment remained under pressure. Luxor estimates that machines in the 25–38 J/TH range generated roughly $45 per MWh on average during August, below its $48-per-MWh benchmark electricity cost. The fleet exceeded that benchmark on only 11 days during the month.

That helps explain why BitcoinVersus.tech’s recent look at PowerCompute’s S19 XP refresh matters beyond a simple miner upgrade. More efficient ASICs can keep operating through conditions that push older machines below their electricity breakeven point.

Texas Can Put Curtailed Hashrate Back Into Play

One portion of the idle fleet may be offline for reasons unrelated to hardware economics. Texas miners routinely curtail during summer power-market conditions, including the Four Coincident Peak mechanism used to determine transmission charges.

Luxor notes that the final 4CP peak is established in September. After that, the seasonal incentive to avoid those summer peaks disappears until the next cycle. Machines that were economically rational to curtail during summer can therefore become candidates to restart even without a major improvement in ASIC efficiency.

The relationship between mining and flexible grid load is already central to the Texas market. BitcoinVersus.tech has tracked Cipher Digital’s 3.2 GW ERCOT power queue and Bitdeer’s 35 MW Project Kati deployment, both of which show how mining capacity is increasingly inseparable from grid position and power strategy.

October Has Historically Added Difficulty

Luxor’s historical data gives the restart question a seasonal edge. Since 2022, October has delivered positive difficulty growth every year, averaging about 4.38% per adjustment epoch and roughly 10% across the month. Those historical figures do not guarantee the same outcome in 2026, but they identify a recurring period when additional hashrate has tended to arrive.

The September 5 difficulty adjustment had already risen 1.31% after August blocks averaged 9 minutes and 34 seconds. Faster-than-target block production is one mechanism through which returning hashrate eventually translates into a higher difficulty setting.

Not Every Idle ASIC Will Return

The 235 EH/s estimate is not a forecast that all sidelined capacity will switch on at once. Luxor does not provide a precise breakdown among uneconomic hardware, curtailment, transit and maintenance. Each category has a different path back to production.

Some inefficient miners may never clear their local electricity cost again. Some machines may be awaiting transformers, racks or repair parts. Others may be deliberately idle because their operators earn more by responding to power-market conditions than by hashing continuously.

That is why BitcoinVersus.tech’s earlier examination of Bitcoin testing estimated mining-cost levels near $85,000 and the current idle-capacity estimate describe different parts of the same system: revenue determines which machines can run, while efficiency, electricity contracts and grid incentives determine which machines actually do.

The Network Can Take Back a Hashprice Rally

For operators, the central risk is self-correcting competition. Higher Bitcoin prices can lift hashprice and make marginal ASICs profitable. Those machines return, network hashrate rises, difficulty follows, and revenue per unit of hashrate can be compressed again.

Luxor summarizes that dynamic directly: every EH that switches back on adds competition. With hundreds of exahashes potentially sidelined for several different reasons, the next mining cycle is not only about how much new hardware manufacturers can ship. It is also about how much existing hardware operators decide to turn back on.

Independent reporting on the estimate similarly notes that sustained hashrate, subsequent difficulty adjustments and operator disclosures will be needed to determine how much of the capacity actually returns.

BitcoinVersus.Tech Editor’s Note

Support independent technology reporting: BTC donations may be sent to 3C9o19EH5HSiwEPyCTmEKzxhNCbo2X6TTb.

Follow BitcoinVersus.tech on X/Twitter for Bitcoin mining, ASIC hardware, AI infrastructure, semiconductor and data-center reporting.

Disclaimer: BitcoinVersus.tech is not a financial advisor. This media platform reports on financial subjects purely for informational purposes.