Bitcoin network hashrate sits at 890.9 EH/s as of July 24, 2026, near all-time highs. Bitcoin spot price: $62,042 — down 50.8% from its October 2025 peak of $126,000. The result is a hashprice of $30.88/PH/s/day, 37.2% below its October 2025 peak of $49.40 and 66% below July 2025 levels of roughl...
"We've made the strategic decision to exit Bitcoin mining operations entirely." — Paul Prager, CEO, TeraWulf
Bitcoin network hashrate sits at 890.9 EH/s as of July 24, 2026, near all-time highs. Bitcoin spot price: $62,042 — down 50.8% from its October 2025 peak of $126,000. The result is a hashprice of $30.88/PH/s/day, 37.2% below its October 2025 peak of $49.40 and 66% below July 2025 levels of roughly $63/PH/s/day.
This collapse in unit economics has triggered the most significant structural transformation in Bitcoin mining history. Public miners have announced more than $70 billion in aggregate AI and high-performance computing (HPC) contracts, liquidated over 15,000 BTC from treasury holdings, and begun reallocating power infrastructure away from proof-of-work. Mining stocks with secured HPC contracts now trade at EV/NTM sales multiples of 12.3x versus 5.9x for pure-play miners, according to CoinShares data. The market is repricing these companies as data center operators, not Bitcoin miners.
Hashprice — the daily revenue per petahash per second of deployed computing power — is the single most important metric for mining profitability. It captures the combined effect of Bitcoin price, network difficulty, block subsidy, and transaction fees into one number.
The trajectory since mid-2025:
| Date | Hashprice (USD/PH/s/Day) | Network Hashrate | BTC Price | |------|--------------------------|-----------------|-----------| | July 2025 | ~$63.00 | ~750 EH/s | ~$126,000 | | October 2025 | $49.40 | 1,160 EH/s (peak) | ~$126,000 | | December 2025 | ~$35.00 | 1,045 EH/s | ~$87,000 | | March 2026 | $28-$30 | 1,020 EH/s | ~$65,000 | | July 13, 2026 | $30.88 | 879 EH/s (7d SMA) | $62,042 |
The 7-day simple moving average hashrate of 879 EH/s on July 13 reflected a 6.4% weekly decline from 939 EH/s the prior week, according to Hashrate Index data. The 30-day SMA remained higher at 938 EH/s. Network difficulty adjusted downward 5% to 127.17 trillion on July 11, 2026, at block 957,600.
Transaction fees have collapsed to negligible levels. For the week ending July 13, fees totaled just 20 BTC (~$1.27 million), representing 0.69% of total block rewards. The 7-day average stood at 0.0247 BTC in fees per block. Weekly block rewards totaled approximately 2,914 BTC (~$184 million).
Bitcoin's year-to-date price performance: -29.1%.
The April 2024 halving cut the block subsidy from 6.25 BTC to 3.125 BTC. For miners, this mechanically doubled the cost of producing each bitcoin from mining operations, all else equal.
According to the CoinShares Q1 2026 Bitcoin Mining Report, the weighted average cash cost among publicly listed miners reached approximately $79,995 per BTC in Q4 2025. All-in costs — including depreciation, SG&A, and financing — ranged from $118,188 to $471,841 per BTC across major operators.
Electricity costs per bitcoin mined varied sharply by operator:
| Miner | Electricity Cost per BTC | Cash Cost per BTC | |-------|-------------------------|-------------------| | IREN | $34,325 | — | | CleanSpark (CLSK) | — | $71,188 | | Bitdeer (BTDR) | — | $87,144 | | Marathon (MARA) | — | $103,605 | | Core Scientific (CORZ) | — | $110,282 | | HIVE | $65,368 | — |
At a spot price of $62,042, every miner on this list is operating above cash cost on an all-in basis. Only the most efficient operators — those running sub-14 J/TH fleets with electricity below $0.04/kWh — generate positive operating margins from mining alone.
Hashrate Index data shows earnings stratified by fleet efficiency: machines under 14 J/TH generate $109 per megawatt-hour of energy consumed, while older 25-38 J/TH rigs yield only $41/MWh. At average U.S. industrial electricity rates of $70-$80/MWh, older-generation hardware is unprofitable even before accounting for cooling, labor, and overhead.
Marathon Digital (MARA) reported Q1 2026 revenue of $174.6 million, an 18.4% year-over-year decline that missed the consensus estimate of $192.7 million. Net loss: $1.26 billion, driven by a $1 billion unfavorable fair-value adjustment on bitcoin holdings and $45.9 million in restructuring charges. The company sold approximately $1.5 billion in bitcoin during the quarter, using proceeds to repurchase more than $1 billion of convertible debt at a discount. MARA reduced its workforce by 15%, projecting annualized savings of $12 million.
CleanSpark (CLSK) reported 42.6 EH/s average operating hashrate in June 2026, with peak fleet efficiency of 16.07 J/TH. The company produced 614 BTC in June and sold 429 BTC, maintaining treasury holdings of 13,924 BTC. CleanSpark's energy-efficient fleet positions it among the lowest-cost operators.
Riot Platforms (RIOT) deployed 42.5 EH/s with 36.4 EH/s average operating hashrate. Q1 production: 1,473 BTC. The company sold 3,778 BTC for $289.5 million during the quarter — more than double what it mined, drawing down treasury reserves.
HIVE Digital posted fiscal Q1 2026 revenue of $45.6 million, with mining contributing $40.8 million (up 44.9% sequentially). HIVE mined 406 BTC in the quarter.
Across the sector, public miners reduced holdings by over 15,000 BTC from peak levels during Q4 2025 through Q1 2026, according to CoinShares. This selling pressure, combined with the need to fund operational pivots, represents a structural shift away from the "HODL" treasury strategies that defined the 2021-2024 era.
The mining industry's pivot toward AI and HPC infrastructure is no longer speculative — it is the dominant strategic theme among publicly listed operators. Aggregate announced AI/HPC contracts exceed $70 billion.
TeraWulf (WULF) has been the most aggressive. The company signed a 20-year lease with Anthropic for approximately 401 MW of critical IT load, with $19 billion in expected contracted revenue. In Q1 2026, TeraWulf generated $21 million in HPC revenue out of $34 million total — meaning AI already accounts for 62% of revenue. Management has stated its intention to exit Bitcoin mining operations entirely.
Hut 8 (HUT) secured a $7 billion Google-backed deal and signed a 15-year, $9.8 billion lease for a 352 MW Texas facility built to NVIDIA's reference architecture. Total contracted campus-level revenue: $19.6 billion. The company is building the Beacon Point campus with a second 352 MW phase.
Core Scientific (CORZ) sold $175 million worth of Bitcoin in March 2026 — 1,992 BTC — specifically to fund its infrastructure transition toward AI workloads.
The economic logic is straightforward. CoinShares projects that AI/HPC revenue could constitute 70% of listed miners' total revenue by end-2026, up from approximately 30% at the start of the year. AI workloads offer structurally higher and more stable returns per megawatt than Bitcoin mining. A megawatt dedicated to AI inference or training generates predictable, contracted revenue under multi-year leases; the same megawatt dedicated to Bitcoin mining generates variable revenue subject to hashprice fluctuations, difficulty adjustments, and BTC spot volatility.
The market has already repriced. Bitcoin's YTD price decline of approximately 46% contrasts sharply with mining stock performance:
| Stock | YTD Performance | Strategy | |-------|----------------|----------| | HUT 8 (HUT) | +363.3% | AI/HPC pivot (Google deal) | | TeraWulf (WULF) | +268.9% | Full AI pivot (Anthropic deal) | | IREN | +121.1% | AI/HPC expansion | | Riot Platforms (RIOT) | +59.9% | Partial AI pivot | | CleanSpark (CLSK) | +12.4% | Pure-play mining |
The divergence is instructive. Companies with secured HPC contracts trade at EV/NTM sales multiples of 12.3x, while pure-play miners trade at 5.9x. Investors are not buying hashrate; they are buying access to power infrastructure, grid interconnection agreements, and data center real estate — assets that happen to currently mine Bitcoin but are worth more running GPU clusters.
CleanSpark, which has remained the most committed to pure-play mining, has the weakest stock performance. This is despite running one of the most efficient mining fleets in the industry. The market is signaling that operational efficiency in mining is less valuable than strategic positioning in AI.
The structural migration of mining capacity toward AI raises questions about Bitcoin's long-term security model. If the most capitalized, best-connected operators redirect power to higher-margin AI workloads, the network becomes increasingly dependent on smaller, less efficient miners and geographic jurisdictions with low electricity costs.
The difficulty adjustment mechanism — Bitcoin's built-in self-correcting feature — will absorb departures by lowering difficulty, making mining more profitable for remaining participants. The 5% difficulty decrease on July 11 and the projected 16% decrease on July 26 demonstrate this mechanism in action. Year-to-date, 8 of the first 14 difficulty adjustments in 2026 have been negative, reflecting a cumulative 14.22% decline from the January 8 peak of 146.47 trillion.
However, the hashrate-to-security relationship is not linear. What matters for network security is the cost of acquiring 51% of hashrate, which depends on both the total hashrate and the geographic and organizational distribution of that hashrate. Concentration risk may increase if the remaining miners are fewer and more geographically clustered.
Transaction fees at 0.69% of block rewards offer no meaningful alternative revenue source at current levels. The long-term viability of Bitcoin's security model depends on either substantially higher BTC prices, dramatically increased transaction fee revenue (potentially from Layer 2 settlement activity), or both. The next halving in 2028 will cut the subsidy to 1.5625 BTC, further compressing this economic model.
The Bitcoin mining industry is undergoing an irreversible structural transformation. The combination of post-halving economics, a 50% BTC price decline from peak, and the availability of higher-margin AI workloads has created a rational economic incentive for miners to reallocate their most valuable asset — access to cheap, large-scale power — away from proof-of-work.
This is not a temporary dislocation. The multi-year, multi-billion-dollar AI contracts being signed by TeraWulf, Hut 8, and others lock in power allocation for decades. Once a 401 MW site is leased to Anthropic for 20 years, that capacity does not return to Bitcoin mining.
The difficulty adjustment will keep the network functional. Bitcoin does not need today's miners to survive. But the economic model that sustained the industry — mine BTC, hold BTC, benefit from price appreciation — is being replaced by a model where Bitcoin mining is, at best, a transitional use of power infrastructure until higher-value tenants arrive.
For the first time in Bitcoin's history, the companies best positioned to mine it are choosing not to.