Bitcoin rose to around $86,000, but the mining industry saw only limited improvement in profitability.

CryptoSlate, a blockchain media outlet, reported on Sept. 21 that after the difficulty adjustment on Sept. 19, total revenue per unit of mining computing power was about 2.65 percent above the previous baseline.

The calculation reflects a bitcoin price of $84,751, the confirmed increase in difficulty and recent total hashprice. Hashprice was estimated at $40.31 per day per petahash per second (PH/s). The figure does not represent actual profit or an individual operator's earnings, but is closer to a network-wide theoretical gross revenue indicator.

A model as of Sept. 15 estimated that bitcoin would need to reach about $82,877 to offset the expected difficulty increase scheduled for Sept. 19. At block height 967,680 on Sept. 19, difficulty rose 4.1634 percent, to 13.2757 trillion from 12.7451 trillion. The current bitcoin price is about 2.26 percent above the baseline presented by the model at the time.

As bitcoin rose more than mining difficulty, the mining industry got some relief. Compared with the Sept. 15 model inputs, the price rose about 7.07 percent and difficulty rose about 4.16 percent, improving the price-to-difficulty ratio by about 2.79 percent. Even after reflecting the recent decline in average fees, the theoretical hashprice, meaning expected revenue per unit hashrate, rose about 2.65 percent from the previous baseline.

But fee income played little role as a buffer for miners. Fees generated across 144 blocks from 967,828 to 967,971 totalled 2.04858206 BTC, or an average 0.01422626 BTC per block. Fees accounted for about 0.45 percent of total rewards of 452 BTC. That means mining revenue still depends entirely on block subsidies and the bitcoin price.

Market attention is shifting to the next difficulty adjustment. The current estimate points to a decline of about 2.48 percent. The figure is from an early stage, with 14.43 percent of the difficulty adjustment cycle completed. Average block time was 625.3 seconds, about 25 seconds longer than the 10-minute target. If this pace holds, the next adjustment is expected around Oct. 3.

It is difficult to view this as an immediate signal of mining rigs shutting down or hashrate leaving the network. Block production has large probabilistic fluctuations, making it hard to draw conclusions on actual changes in network computing power from short observation periods. The current estimate is based on early block-production speed and is not a direct count of operating mining machines.

Over the past month, estimated hashrate has fluctuated between about 826.1 exahashes per second (EH/s) and 1.053 zettahashes per second (ZH/s), with the current estimate at about 937.5 EH/s. No sustained and clear collapse trend across the network has been confirmed in that range. It cannot rule out changes by individual operators, but it does not amount to a signal strong enough to claim widespread shutdowns.

Ultimately, three factors will shape future mining conditions. They are whether bitcoin stays above the existing model baseline, whether transaction fees grow as an additional revenue source beyond subsidies, and whether the early-stage estimate of a difficulty decline holds after more block samples accumulate. If slow block production continues, the difficulty-decline signal becomes more credible. If block times speed up again, the current estimate could shrink or reverse.

Mining profitability could also still vary widely depending on each operator's cost structure. Equipment efficiency, power contracts, financing and labour costs differ, meaning the same network revenue level can produce different outcomes across operators. For now, bitcoin's rebound has theoretically offset the confirmed increase in difficulty, but it remains unclear whether the improvement will last because of low fees and an immature estimate for the next difficulty adjustment.