Bitcoin mining is experiencing its most severe economic crisis since China's 2021 mining ban. On February 5, 2026, hashprice — the industry's core profitability metric measuring daily revenue per unit of hashrate — fell to an all-time low of approximately $0.03 per TH/s (or below $30/PH/s), as Bi...
Bitcoin mining is experiencing its most severe economic crisis since China's 2021 mining ban. On February 5, 2026, hashprice — the industry's core profitability metric measuring daily revenue per unit of hashrate — fell to an all-time low of approximately $0.03 per TH/s (or below $30/PH/s), as Bitcoin's price slid under $70,000 for the first time since November 2024. The subsequent mining difficulty adjustment on February 8–9 saw an 11% downward correction, the largest single drop since the 2021 China exodus, confirming a capitulation-style shakeout across the sector.
But this is not merely a cyclical downturn. The 2024 halving's delayed economic consequences have converged with a structural force that no previous mining cycle has faced: the artificial intelligence infrastructure boom. AI data center operators, flush with capital from hyperscaler contracts worth billions of dollars, are now outbidding Bitcoin miners for every available megawatt of power capacity. The result is a two-front war where miners face collapsing revenues on one side and escalating opportunity costs on the other — and an increasing number are choosing to surrender.
This report examines the economic mechanics of the mining crisis, maps the industry's accelerating pivot toward AI/HPC infrastructure, quantifies the network security implications of hashrate concentration and volatility, and assesses whether Bitcoin's security budget model remains viable in a post-halving, AI-competitive energy market.
The hashprice index fell to approximately $0.034/TH/s on February 5, 2026, according to data reported by Bloomberg and Hashrate Index[^1][^2]. This represents:
To contextualize the scale of this deterioration: hashprice stood near $3.50/TH/s in 2017. The current reading represents a 99% decline over eight years — a compression that reflects both the exponential growth in network hashrate and the halving schedule's relentless reduction in block rewards.
Mining hardware prices have followed hashprice into historic territory. The Luxor ASIC Price Index shows used-generation machines trading at scrap-metal valuations, while even next-generation sub-15 J/TH units have seen 40–60% markdowns from mid-2025 peaks[^3]. The secondary hardware market, once a reliable indicator of miner sentiment, now signals broad-based capitulation.
The April 2024 halving cut the block reward from 6.25 BTC to 3.125 BTC. While Bitcoin's subsequent rally to $126,000 temporarily masked the revenue impact, the 45% price decline from that peak into February 2026 has exposed the full severity of post-halving economics:
Energy Requirements per Bitcoin:
Breakeven Electricity Costs (at ~$69,000 BTC):
Daily Network Mining Revenue:
ROI on New Hardware:
These economics have created a binary outcome set for operators: either secure electricity at below $0.06/kWh with latest-generation hardware, or face inevitable margin compression into unprofitability.
What distinguishes this mining downturn from every prior cycle is the emergence of a competing, higher-value use case for mining's core asset: power infrastructure. The artificial intelligence boom has created insatiable demand for electricity and physical data center capacity — precisely the assets that Bitcoin miners have spent a decade accumulating.
The scale of the pivot is staggering:
Bitfarms → Keel Infrastructure: On February 6, 2026, Bitfarms announced a full exit from Bitcoin mining over two years, rebranding as Keel Infrastructure and redomiciling from Canada to Delaware. The company declared, "We are no longer a Bitcoin company." Its Washington State facility is being retrofitted from an 18 MW mining farm to an Nvidia GB300 GPU cluster with liquid cooling, backed by a fully funded $128 million contract. Shares surged 16% on the news[^9][^10].
Hut 8: Secured a $7 billion Google-backed deal to develop AI data center capacity, representing one of the largest single contracts in the mining industry's history[^11].
TeraWulf: Google acquired a 14% stake in the company, becoming its largest shareholder. TeraWulf has secured AI/HPC hosting contracts valued at approximately $6.7 billion[^12].
Core Scientific: Signed a $3.5 billion deal with CoreWeave for 200 MW of HPC hosting capacity, effectively monetizing power assets at multiples of their Bitcoin mining value[^13].
The economic logic is straightforward and devastating for Bitcoin maximalism: a megawatt of power capacity committed to a 15-year, fixed-rate AI lease with an investment-grade counterparty (Google, Microsoft, CoreWeave) generates predictable, high-margin revenue. The same megawatt dedicated to Bitcoin mining generates volatile, declining revenue subject to halving-cycle compression and hashrate competition. Sector-wide data center capital expenditure increased 400% between March 2025 and February 2026[^14].
The Nvidia Blackwell (B200/GB200) GPU architecture is sold out through mid-2026, creating a "haves and have-nots" dynamic based on 2024 order history. Miners who placed early GPU orders are positioned to pivot; those who doubled down on ASICs face stranded assets.
In late January 2026, Winter Storm Fern struck the central and southern United States, forcing emergency power curtailments across the ERCOT (Texas), PJM, and SPP grids. The impact on Bitcoin's hashrate was immediate and severe:
This episode exposed a critical vulnerability in Bitcoin's security model: geographic concentration. The United States now hosts an estimated 40–50% of global Bitcoin hashrate, and the majority of that capacity is concentrated in Texas and a handful of other states with deregulated energy markets. A single severe weather event was able to temporarily remove roughly one-third of the network's total computational power.
While no chain disruptions or double-spend attempts were observed — and the theoretical cost of a 51% attack remains prohibitively high even at reduced hashrate levels — the episode demonstrated that Bitcoin's security is increasingly dependent on the operational continuity of a shrinking number of large, geographically concentrated industrial operators. The network's difficulty adjustment mechanism functioned as designed, providing automatic relief to remaining miners, but the 11% single-adjustment magnitude signals a level of hashrate volatility that was previously associated only with nation-state-level disruptions.
Galaxy Digital's 2024 prediction that the mining sector would undergo aggressive consolidation has proven prescient. The industry is now bifurcating into two distinct categories:
Category 1: Hybrid Infrastructure Companies — These firms are converting portions of their power capacity to AI/HPC while maintaining reduced Bitcoin mining operations. This group includes MARA Holdings (targeting 75 EH/s with 1.7 GW of captive capacity), Hut 8 (post-Google deal), TeraWulf (post-Google investment), and Core Scientific (post-CoreWeave deal). Their strategy is to use AI revenue as a stable base while maintaining optionality on Bitcoin upside.
Category 2: Pure-Play Miners Under Pressure — Operators who lack the capital, contracts, or infrastructure quality to pivot face a grim calculus. With hashprice below $30/PH/s, only those with electricity costs below $0.05/kWh and latest-generation hardware (<20 J/TH) can maintain positive operating margins. Everyone else faces a choice between selling assets to Category 1 firms or winding down operations.
Industry analysts project that the current seven publicly traded mining companies with market capitalizations around $1 billion will consolidate to approximately four within three to four years[^18]. The M&A pipeline is already active: Riot Platforms' attempted acquisition of Bitfarms in 2024–2025 and Hut 8's merger with US Bitcoin Corp. established the template. The Bitfarms-to-Keel transformation represents the latest — and most dramatic — expression of this consolidation thesis.
The mining crisis illuminates a fundamental question that the Bitcoin community has deferred for over a decade: can transaction fees alone sustain network security as block rewards diminish?
The data is sobering. Bitcoin's annualized fee revenue stands at approximately $115 million — a figure that must be weighed against the roughly $54–72 billion in annual infrastructure costs required to secure the network (including mining issuance subsidy, electricity, hardware depreciation, and operational overhead)[^19]. Fees currently represent less than 2% of miners' total economic incentive to secure the chain.
Each halving compresses this subsidy further. By the 2028 halving, the block reward will fall to 1.5625 BTC — and the industry will have 18 months fewer of post-halving runway before margin pressure becomes acute again.
The AI pivot introduces a new variable into this equation: if the most well-capitalized, efficient mining operations redirect their power infrastructure to AI data centers, the remaining hashrate will be concentrated among operators with lower margins, less geographic diversity, and weaker balance sheets. This is not merely a business concern — it is a network security concern, because the quality and resilience of the remaining hashrate base will decline even if the raw EH/s figure recovers through difficulty adjustments.
Bitcoin's difficulty adjustment mechanism ensures the network will always function, but it does not guarantee that the network will function securely. The difference between 1,000 EH/s distributed across dozens of independent operators on three continents and 1,000 EH/s concentrated among four publicly traded companies in Texas is profound — even though the headline hashrate number is identical.
The Bitcoin mining industry is undergoing a metamorphosis that has no precedent in the network's 17-year history. This is not a cyclical downturn that will resolve when Bitcoin's price recovers — though price recovery would certainly alleviate short-term margin pressure. It is a structural transformation driven by the collision of two forces: the halving schedule's inexorable compression of mining economics, and the emergence of AI as a higher-value competitor for the same physical infrastructure.
The firms that survive will not be mining companies. They will be power infrastructure companies that happen to mine Bitcoin when the economics are favorable. Bitfarms' declaration that it is "no longer a Bitcoin company" is not capitulation — it is the honest articulation of a new economic reality that the rest of the industry will eventually echo.
For Bitcoin itself, the implications are more complex. The network will continue to function — the difficulty adjustment mechanism guarantees this. But the quality of its security is evolving in ways that merit close attention: increasing geographic concentration, rising dependence on a shrinking number of well-capitalized operators, and a widening gap between organic fee revenue and the total cost of maintaining the current security posture. The question is no longer whether Bitcoin mining is sustainable as a standalone business. The question is whether Bitcoin's security model is sustainable in a world where its miners have better things to do with their electricity.
[^1]: Bloomberg, "Bitcoin Mining Revenue Gauge Falls to Record Low During Selloff," February 5, 2026. https://www.bloomberg.com/news/articles/2026-02-05/bitcoin-mining-revenue-gauge-falls-to-record-low-during-selloff
[^2]: TheMinerMag, "Bitcoin Hashprice Hits Record Low as BTC Falls Below $70,000," February 5, 2026. https://theminermag.com/news/2026-02-05/bitcoin-hashprice-record-low
[^3]: PR Newswire / Hashrate Index, "Bitcoin Price Drop Sends Mining Hardware Prices to Historic Lows as Hashprice Hits New Bottom," February 2026. https://www.prnewswire.com/news-releases/bitcoin-price-drop-sends-mining-hardware-prices-to-historic-lows-as-hashprice-hits-new-bottom-302685668.html
[^4]: CompareForexBrokers, "The Real Cost of Bitcoin Mining in 2026," February 2026. https://www.compareforexbrokers.com/us/bitcoin-mining/
[^5]: EZ Blockchain, "Is Bitcoin Mining Still Profitable In 2026?" https://ezblockchain.net/article/will-crypto-mining-stay-profitable-in-2026/
[^6]: CCN, "Bitcoin Mining ROI Soars to 1,000 Days | Hash Revenue Down 35%," February 2026. https://www.ccn.com/education/crypto/bitcoin-mining-roi-1000-days-hash-revenue-down-35-survival-explained/
[^7]: BuyBitcoinWorldwide, "61 Bitcoin Energy Consumption Statistics (2026)." https://buybitcoinworldwide.com/bitcoin-mining-statistics/
[^8]: CCN, ibid.
[^9]: CoinDesk, "Bitfarms Says It's 'No Longer a Bitcoin Company' as It Moves to U.S. Under New Name," February 6, 2026. https://www.coindesk.com/business/2026/02/06/bitfarms-says-it-s-no-longer-a-bitcoin-company-doubling-down-on-ai-with-u-s-move
[^10]: Decrypt, "Bitfarms Will 'Wind Down' Bitcoin Mining and Pivot to AI After $46 Million Loss," February 2026. https://decrypt.co/348573/bitfarms-wind-down-bitcoin-mining-pivot-ai-after-46-million-loss
[^11]: Carbon Credits, "Hut 8 Pivots From Bitcoin to AI With $7B Google-Backed Deal to Power Data Centers." https://carboncredits.com/hut-8-pivots-from-bitcoin-to-ai-with-7b-google-backed-deal-to-power-data-centers/
[^12]: Bitcoin Magazine, "BTC Miner TeraWulf Stock Jumps 25% On AI News." https://bitcoinmagazine.com/bitcoin-mining/btc-miner-terawulf-stock-jumps
[^13]: Cointelegraph, "Bitcoin Mining 2026: AI Pivot, Profitability Pressure & Consolidation." https://cointelegraph.com/news/bitcoin-mining-outlook-2026-ai-profitability-consolidation
[^14]: insights4vc, "Bitcoin Mining's AI Pivot: 2026 Thesis Update." https://insights4vc.substack.com/p/bitcoin-minings-ai-pivot-2026-thesis
[^15]: KuCoin, "Bitcoin Hashrate Drop 2026." https://www.kucoin.com/news/articles/bitcoin-hashrate-drop-2026-impact-on-mining-difficulty-crypto-btc-network-security
[^16]: CryptoNomist, "Bitcoin Mining: 11% Difficulty Drop, Biggest Since 2021," February 9, 2026. https://en.cryptonomist.ch/2026/02/09/bitcoin-mining-difficulty-drop-2026/
[^17]: The Coin Republic, "Bitcoin Mining Difficulty Posts Largest Drop Since 2021 China Ban," February 13, 2026. https://www.thecoinrepublic.com/2026/02/13/bitcoin-mining-difficulty-posts-largest-drop-since-2021-china-ban/
[^18]: Cointelegraph, ibid.
[^19]: Maze2 SA, "Economic Value Distribution in Blockchain Ecosystems," October 2025. https://github.com/Ricosworks1/blockchain-payment-flow-analysis