The final week of April 2026 marked a pivotal moment in the history of Bitcoin mining, with two transactions that signaled a major turning point for the industry.
On April 28 (local time), Nasdaq-listed mining firm Core Scientific officially announced it would convert its 300 MW Bitcoin mining facility in Pecos, Texas, into a 1.5 GW AI data center campus. The first batch of data halls is expected to go live in early 2027. To support this ambitious plan, the company completed the pricing and issuance of $3.3 billion in senior secured notes (7.750% coupon, maturing in 2031). Combined with a previously secured $1 billion credit facility from Morgan Stanley, total financing now exceeds $4 billion.
Around the same time, a subsidiary of leading mining firm Hut 8 priced $3.25 billion in senior secured notes (6.192% coupon, maturing in 2042) to fund the construction of a 245 MW AI data center at the River Bend campus in St. Francisville, Louisiana. This project has already signed a 15-year, $7 billion lease agreement with cloud provider Fluidstack, with lease payment guarantees provided by Alphabet, Google’s parent company. Fitch and S&P assigned a "BBB-" investment-grade rating to the bonds—an extremely rare occurrence in crypto mining history.
These transactions are not isolated events. Extending the timeline to the first quarter of 2026 reveals a broader picture: Bitcoin’s total network hashrate recorded its first quarterly decline since 2020, dropping by roughly 4% to 6%. Publicly traded mining firms sold over 32,000 BTC in a single quarter, a new all-time high. The total value of AI/HPC contracts announced by miners exceeded $7 billion.
Together, these signals point to a core theme: Bitcoin mining is undergoing a paradigm shift from "crypto mining machines" to "AI compute infrastructure providers."
From Halving Shock to Industry Pivot
To understand the scale and urgency of this transformation, we need to revisit the collapse of the industry’s profit model.
In April 2024, Bitcoin underwent its fourth halving, slashing block rewards from 6.25 BTC to 3.125 BTC—cutting miners’ per-block income overnight. Contrary to some expectations, the network hashrate did not drop sharply post-halving. Instead, it continued to climb, reaching a record high of about 1,160 EH/s by the end of 2025. The combination of rising hashrate and halved rewards meant each unit of hashrate earned fewer and fewer bitcoins.
The real profitability crisis erupted in Q4 2025. According to CoinShares’ Q1 2026 mining report, the weighted average cash cost to mine a single bitcoin for public miners surged to around $79,995, while the Bitcoin price hovered between $68,000 and $70,000. This meant a book loss of roughly $19,000 per BTC mined.
By early 2026, hashprice (revenue per unit of hashrate) dropped further to around $28–$30 per PH/s per day—a post-halving record low. CoinShares estimates that roughly 15% to 20% of global mining rigs were operating at a loss.
Under this financial pressure, miners fundamentally changed their capital allocation strategies. The wave of AI infrastructure deals seen in Q1 2026 was not a knee-jerk reaction, but a collective response to this profitability crisis.
The following timeline highlights the key milestones in this transformation:
| Date | Event |
|---|---|
| April 2024 | Bitcoin’s fourth halving, block reward drops to 3.125 BTC |
| Q4 2025 | Weighted average mining cash cost rises to ~$79,995 per BTC |
| End of 2025 | Network hashrate hits ~1,160 EH/s, an all-time high |
| Jan–Mar 2026 | Hashprice drops to $28–$30/PH/s/day, 15%–20% of miners unprofitable |
| Q1 2026 | Network hashrate declines 4%–6% from start of year, first Q1 drop in six years |
| Q1 2026 | Public miners collectively sell over 32,000 BTC, a new quarterly record |
| March 2026 | Core Scientific secures $1 billion credit facility from Morgan Stanley |
| April 21–23, 2026 | Core Scientific prices $3.3 billion in senior secured notes |
| April 27–28, 2026 | Hut 8 prices $3.25 billion in investment-grade secured notes |
| April 28, 2026 | Core Scientific officially announces Pecos 1.5GW AI data center plan |
Breaking Down the Three Pillars of Miner Transformation
Fundamental Differences in Financing Structures
The two financings by Core Scientific and Hut 8 represent distinct capital paths—an important lens for understanding industry divergence.
Core Scientific’s $3.3 billion bond is a high-yield ("junk") bond, reflecting higher credit risk and financing costs. The company secured a 12-year, $10 billion hosting agreement with AI cloud provider CoreWeave as a repayment foundation. During the transition, Core Scientific continued to sell its bitcoin holdings to supplement cash flow—its BTC balance dropped from 2,537 at the end of 2025 to under 1,000, and the company expects to sell most of its remaining bitcoin in 2026.
In contrast, Hut 8’s $3.25 billion bond received a "BBB-" investment-grade rating, resulting in significantly lower financing costs. This rating was primarily supported by Alphabet’s lease payment guarantees and the predictability of cash flows locked in by a 15-year fixed lease. The bond uses a project finance structure, with no recourse to parent company Hut 8 Corp., isolating project financial risk at the subsidiary level.
A side-by-side comparison:
| Dimension | Core Scientific | Hut 8 |
|---|---|---|
| Issuance Size | $3.3B notes + $1B credit | ~$3.25B notes |
| Bond Type | Senior secured (high-yield) | Senior secured (investment-grade) |
| Coupon Rate | 7.750% | 6.192% |
| Maturity | 2031 (~5 years) | 2042 (~16 years) |
| Credit Rating | Not investment-grade | BBB- (Fitch/S&P) |
| Underwriters | — | Goldman Sachs, J.P. Morgan, Morgan Stanley |
| Data Center Capacity | 1.5 GW (1 GW leasable) | 245 MW |
| Key Client | CoreWeave (12 yrs/$10B) | Fluidstack (15 yrs/$7B, Google guarantee) |
| Recourse Structure | Corporate-level obligation | No recourse to parent |
| BTC Holdings Change | 2,537 down to <1,000 | ~$1.4B in cash and BTC reserves |
The joint underwriting of Hut 8’s debt by top-tier investment banks—Goldman Sachs, J.P. Morgan, and Morgan Stanley—signals that traditional finance is now backing miners’ AI transformation with mainstream financial instruments.
Structural Drivers Behind Hashrate Decline
In Q1 2026, the Bitcoin network hashrate fell from about 1,066 EH/s (30-day average) at the start of the year to around 1,004 EH/s—a 5.8% drop, according to Hashrate Index, largely due to mass shutdowns of older hardware.
Zooming out, the year-to-date decline of roughly 4% marks the first Q1 drop since 2020, ending five years of consecutive double-digit growth.
Three main factors drove this decline: First, persistently low hashprice forced mid- and low-end miners out of the network. Second, some miners proactively shifted power resources from Bitcoin mining to AI hosting. Third, the late-January 2026 winter storm "Fern" caused widespread outages in Texas and other mining hubs, with some miners contractually required to curtail usage under demand response agreements.
Notably, despite the short-term hashrate dip, CoinShares still forecasts a potential rebound to 1.8 ZH/s by year-end 2026—but this is highly dependent on the future BTC price. Analysts believe that unless BTC recovers above $100,000, high-cost miners will continue to exit at an accelerated pace.
Scale and Rationale of Miner BTC Sales
In Q1 2026, public miners sold a record 32,000 BTC—up about 42% from Q4 2025.
There are two main motivations: operational sales (to pay for electricity, equipment upgrades, and routine costs, which is standard industry practice) and strategic sales (deliberately reducing BTC exposure on the balance sheet to fund AI infrastructure investment).
For example, Riot Platforms sold 3,778 BTC in Q1 2026, far exceeding its mining output of just 1,473 BTC for the quarter—indicating that it sold both new production and inventory. Its BTC holdings fell from 19,233 a year earlier to 15,680, with the company citing ongoing capital expenditures as the reason.
Bitdeer took an even more aggressive approach: as of April 2026, it maintained a zero BTC balance, matching weekly production and sales.
Core Scientific sold $175 million worth of bitcoin in a single transaction in March 2026 and stated it expects to sell most of its remaining holdings throughout the year.
Importantly, this wave of miner selling coincided with a surge in institutional buying—public companies net purchased over $2.5 billion in BTC in a single week. The simultaneous amplification of buying and selling is reshaping Bitcoin’s market supply-demand dynamics, with institutional demand offsetting the impact of miner sales.
Dissecting Market Sentiment: Consensus and Contention
Opinions on the collective pivot of miners toward AI are sharply divided among market participants.
Repricing Valuations and Improved Cash Flow
This camp, dominant among Wall Street analysts, argues that AI data centers provide long-term, fixed dollar-denominated contract revenue—a stark contrast to the extreme volatility of Bitcoin mining. Contracts typically span 10 to 15 years, with highly predictable income and hosting business operating margins reaching 80% to 90%.
All 11 analysts covering Core Scientific rate it a "buy," with a median 12-month target price of $26.48. CORZ shares have surged over 200% in the past year, even as Bitcoin fell about 11%, with the stock’s decoupling from BTC interpreted as the market rewarding the transformation narrative.
Hut 8’s stock soared roughly 478% over the past year. Arete Research initiated coverage with a "buy" rating and a street-high $136 target, BTIG raised its target to $90, and Benchmark reiterated its buy rating and $85 target. Piper Sandler called HUT the best-performing miner in its coverage universe.
Capex Black Holes and Delivery Risks
Not all voices are so optimistic. CleanSpark CEO Matt Schultz sounded a clear warning at the "Bitcoin 2026" conference: upgrading a mining site to an AI data center drives per-megawatt costs from about $500,000 to $10–12 million—a more than 20-fold increase. Staffing jumps from roughly one person per 10 MW to eight. More critically, top cloud providers impose extremely strict lease terms—delays of even a single day can trigger penalties equal to an entire year’s contract revenue.
Schultz cautioned the industry not to focus solely on the short-term stock price boosts from contract announcements, while ignoring the enormous challenges of actual delivery.
CoinShares’ research also notes that some hybrid miners have seen their all-in BTC production costs spike due to AI buildouts, while taking on massive debt. For example, IREN holds $3.7 billion in convertible note debt, and TeraWulf’s total debt stands at $5.7 billion—burdens that weigh heavily even before revenue materializes.
Long-Term Security of the Mining Network
A deeper debate centers on Bitcoin network security. A declining hashrate theoretically lowers the cost of attacking the network. If major miners continue shifting power from mining to AI hosting, future hashrate growth could be structurally constrained.
However, CoinShares offers a counterpoint: as high-cost miners exit, profitability for remaining miners may improve, since the difficulty adjustment mechanism automatically lowers mining difficulty. In addition, the geographic diversification of hashrate—driven by emerging hubs like Paraguay (4.3%), Ethiopia (2.5%), the UAE, and Oman (each about 3%)—could enhance network resilience.
Industry Impact Analysis: Three Transmission Effects
Impact on Bitcoin Network Security and Hashrate Distribution
The direct impact of hashrate decline on network security is twofold. In the short term, lower hashrate reduces the theoretical cost threshold for a 51% attack. In practice, however, the economic cost of acquiring a majority of network hashrate remains astronomical—even at the current ~1,000 EH/s level, the required investment in mining rigs and electricity is immense.
More noteworthy is the structural shift in geographic hashrate distribution. As of Q1 2026, the US leads with a 37.4% share (~375 EH/s), followed by Russia at 16.9%, and China at 12%. Together, these three countries account for about 65% of global hashrate, indicating continued concentration. Yet, the rapid rise of new mining regions is significant—Paraguay, leveraging cheap hydroelectric power ($0.033/kWh), has grown to 4.3%, with Ethiopia, the UAE, and Oman also expanding quickly.
If passed, the "Mined in America Act" proposed by Republican senators could further drive mining hardware manufacturing back to the US, challenging the current status quo where about 97% of specialized rigs are produced by China-linked firms.
Impact on Miner Asset Structure and Business Models
Miners’ balance sheets are undergoing a historic transformation. Previously, their core assets were bitcoin holdings and mining hardware. Going forward, a growing share of miner assets will be in the form of data center real estate, power contracts, and long-term hosting agreements.
On the revenue side, some leading miners already derive 30% to 39% of income from AI-related sources (Core Scientific stands at 39%), with projections reaching 70% by the end of 2026. This shift means these companies are transitioning from "bitcoin producers and holders" to "infrastructure operators focused on compute leasing, with mining as a secondary business."
Impact on Token Sell Pressure and Market Supply-Demand
Large-scale miner BTC sales create short-term additional sell pressure. Structurally, however, this selling is "front-loaded"—miners are liquidating accumulated inventory, not future production. Once inventory reaches target levels, ongoing sell pressure should decrease significantly.
More importantly, concentrated institutional buying is reshaping market supply-demand. Weekly net institutional purchases exceeding $2.5 billion now match or surpass quarterly miner sales (~$2.2 billion). The pricing power in the bitcoin market is shifting from miners to institutional investors.
Conclusion
In Q1 2026, Bitcoin mining stands at a crossroads. The first hashrate decline in six years, a record 32,000 BTC sold, and over $7 billion in combined AI infrastructure financing by Core Scientific and Hut 8—these are not isolated industry events, but interconnected facets of a structural transformation.
The shift from "crypto mining machines" to "AI compute infrastructure providers" is still in its early stages. Its ultimate success will hinge on three core variables: the actual delivery and operational efficiency of AI data centers, the future trajectory of bitcoin prices, and miners’ ability to balance aggressive expansion with financial discipline.
The outcome of this "century-defining pivot" will not only determine the fate of a cohort of public companies, but will also profoundly impact the security foundation of the Bitcoin network and the role of crypto mining in the broader digital economy.
