Bitcoin's network hashrate sits at 962 EH/s as of June 11, 2026 — down 17% from its October 2025 all-time high of 1,160 EH/s. The next difficulty adjustment, estimated for June 14, is projected to decrease mining difficulty by approximately 9%, from 138.96 T to 123.88 T. It would mark the fourth ...
"We are no longer a Bitcoin company. We are an infrastructure-first owner and developer for HPC/AI data centers across North America." — Keel Infrastructure (formerly Bitfarms), Corporate Repositioning Statement, April 2026
Bitcoin's network hashrate sits at 962 EH/s as of June 11, 2026 — down 17% from its October 2025 all-time high of 1,160 EH/s. The next difficulty adjustment, estimated for June 14, is projected to decrease mining difficulty by approximately 9%, from 138.96 T to 123.88 T. It would mark the fourth downward adjustment in a compressed window, a pattern not observed since July 2022.
The proximate cause is not a network failure. It is an economic migration. Publicly listed Bitcoin miners have announced over $70 billion in AI and high-performance computing (HPC) contracts, and CoinShares projects that some operators will derive up to 70% of 2026 revenue from non-mining workloads. The weighted average cash cost to produce one bitcoin among listed miners reached $79,995 in Q4 2025, according to the CoinShares Q1 2026 Mining Report, while spot BTC hovers near $67,000 — a per-coin loss of roughly $13,000 for the average public miner. Hash price collapsed to $28–30/PH/s/day in early 2026, a five-year low, rendering 15–20% of the global fleet cash-flow negative.
The result is a structural transformation of the mining industry's business model, with direct implications for Bitcoin's security budget, geographic hash distribution, and the long-term viability of proof-of-work economics.
Bitcoin mining difficulty peaked near 138.96 T in mid-2026 after a sustained climb through late 2025. The network's computational power — measured in exahashes per second — hit a record 1,160 EH/s in October 2025, briefly crossing the symbolic 1 ZH/s (1,000 EH/s) threshold in January 2026.
The reversal was swift. A combination of factors drove hashrate down:
Hashrate partially recovered to approximately 1,020 EH/s by May before settling at 962 EH/s in early June. The network is currently running 1.22 minutes slower than expected block times, confirming the hashrate shortfall that will trigger the projected 9% difficulty decrease on June 14.
Three consecutive negative difficulty adjustments have occurred in 2026 — the first such streak since July 2022, when Ethereum's merge uncertainty and energy cost spikes created similar conditions.
The math is straightforward. According to the CoinShares Q1 2026 Mining Report:
| Metric | Value | |--------|-------| | Weighted avg. cash production cost (Q4 2025) | ~$79,995/BTC | | Bitcoin spot price (June 2026) | ~$67,000 | | Per-coin loss (avg. public miner) | ~$13,000 | | Hash price (Q1 2026 low) | $28–30/PH/s/day | | Fleet unprofitability rate | 15–20% of global rigs | | Electricity threshold for profitability | Below $0.05/kWh |
Miners running mid-generation hardware — S19j Pro-class machines at approximately 29.5 J/TH — at average industrial electricity costs of $0.05/kWh are operating below breakeven. Only operators with access to stranded, subsidized, or state-backed power remain cash-flow positive on mining alone.
The contrast with AI infrastructure economics is stark. AI compute earns 5–10x more per megawatt-hour than Bitcoin mining at current prices, according to industry estimates cited in the Bitfinex Alpha report. AI infrastructure margins exceed 85% with multi-year revenue visibility, compared to mining margins that fluctuate daily with hash price and BTC spot.
The scale of committed AI/HPC revenue among former mining-only companies is substantial:
| Company | AI/HPC Contract Value | Details | |---------|----------------------|---------| | CoreWeave + Core Scientific | $10.2 billion | 12-year colocation agreement | | TeraWulf | $12.8 billion | Contracted HPC revenue pipeline | | Hut 8 | $7 billion | 15-year AI infrastructure lease | | Cipher Digital | Multi-billion | Agreement with Google-backed Fluidstack | | MARA Holdings | Undisclosed | 64% stake acquired in French AI firm Exaion |
Total announced AI/HPC contracts across the listed-miner peer group exceed $70 billion. The pipeline benefits from a structural tailwind: over $500 billion in hyperscaler AI infrastructure investment is expected globally in 2026, and miners already control two assets AI infrastructure needs — large blocks of contracted power and physical sites engineered for heavy compute loads.
Core Scientific has approximately 350 MW energized under its CoreWeave contract, with a target of 590 MW by early 2027 and a total campus capacity of 1.5 GW planned for its Texas facility. HIVE Digital Technologies secured a $30 million contract deploying Nvidia GPUs for enterprise AI customers, reporting 219% year-over-year revenue growth.
The corporate identity shift is explicit. Several companies have formally abandoned the "Bitcoin miner" label:
Keel Infrastructure (formerly Bitfarms): Redomiciled from Canada to the U.S. in April 2026. Trades on Nasdaq and TSX under ticker KEEL. Declared itself "an infrastructure-first owner and developer for HPC/AI data centers."
Cipher Digital (formerly Cipher Mining): CEO Tyler Page stated during the Q1 2026 earnings call that he does not expect Bitcoin to be part of the company's story by 2030, projecting a mining exit "by end of 2027 at the latest, if not sooner." The company has halted further mining capital expenditure.
Hut 8: Separated its mining operations into American Bitcoin Corp via a spin-off, retaining the AI/HPC infrastructure under the Hut 8 parent.
Core Scientific: Bitcoin mining segment reported negative gross margins in Q1 2026. AI colocation revenue now accounts for 39% of total company revenue.
Riot Platforms: Q1 2026 revenue of $167.2 million, with $33.2 million (20%) from data center operations. CEO Jason Les described the quarter as an "inflection point" in the company's data center transition. The company initiated a 112 MW AI-dedicated data center expansion, partially funded by liquidating portions of its 15,680 BTC treasury.
MARA Holdings: Expanded into AI via its Exaion acquisition and a partnership with Starwood Capital Group to build a large-scale U.S. data center targeting AI and cloud computing customers. Energized hash rate reached 72.2 EH/s in Q1, a 33% year-over-year increase — but the company simultaneously authorized treasury BTC sales to fund diversification.
The pivot carries a secondary effect: miners are selling bitcoin to fund it. Publicly listed miners have reduced their collective treasuries by more than 15,000 BTC from peak holdings.
Notable sales include:
The liquidation adds sell pressure to spot markets at a time when BTC is already under compression from macro headwinds. This creates a feedback loop: lower BTC prices increase mining losses, which accelerate the AI pivot, which triggers more treasury sales, which further suppress prices.
The hashrate decline raises questions about Bitcoin's security budget — the total economic cost an attacker would need to bear to compromise the network via a 51% attack.
At 962 EH/s, the network remains more than twice as secure as 2023 levels (~400 EH/s). A 51% attack remains economically impractical at current scale. However, the directional trend matters more than the absolute number.
Key risk factors:
CoinShares attributes the network's residual resilience to "state-backed miners, private operators with cheap or stranded power, and ASIC manufacturers" who continue deploying regardless of margin conditions.
Markets are pricing the AI pivot favorably. Year-to-date 2026 mining stock performance diverges sharply from BTC itself:
| Company | YTD Performance | |---------|----------------| | TeraWulf | +85% | | Hut 8 | +67% | | Riot Platforms | +46% | | Core Scientific | +40% | | Applied Digital | +37% | | Bitcoin (BTC) | -20% |
The message from equity markets is clear: mining companies are being re-rated as AI infrastructure plays, not as bitcoin proxies. The stocks most aggressively pivoting toward AI have outperformed those maintaining mining-heavy profiles.
However, the pivot carries its own risks. Adam Sullivan, CEO of Core Scientific, warned during a recent interview: "I worry about the amount of people that are building purely on spec right now that are going to be delivered four years from now, like 2028. The industry could face some headwinds where we start to see some cracks in terms of demand."
The Bitcoin mining industry is undergoing a capital reallocation event without historical precedent in crypto. The halving cycle has always squeezed margins, but previous cycles saw miners wait for price recovery. This time, an alternative revenue source — AI compute — offers higher margins, longer contract duration, and less volatility. The rational economic response is to reallocate, and that is precisely what is occurring.
For Bitcoin's network, the near-term security picture remains adequate. 962 EH/s is more than sufficient to resist attacks. The longer-term question is whether Bitcoin's fee market can develop fast enough to replace the mining revenue that is being permanently redirected to AI workloads. Block rewards will halve again in 2028. If hash price remains suppressed and AI margins remain elevated, the current trickle of miner departures could become a structural drain.
Fred Thiel, CEO of Marathon Digital, framed the macro trajectory during a recent interview: "By 2028, you'll either be a power generator, be owned by one, or be partnered with one." The statement encapsulates the industry's new calculus. Bitcoin mining is becoming a byproduct of energy infrastructure companies, not a standalone business.
The difficulty adjustment on June 14 will not be the last downward move. The question is whether Bitcoin's economic model can adapt before the miners finish leaving.