Core Scientific reported a $347.2 million net loss in Q1 2026, swinging sharply from $576.3 million in net income a year earlier, as the bitcoin miner logged $266.5 million in non-cash impairment charges and a $30.8 million non-cash loss tied to warrant and contingent value right revaluations. The loss landed even as total revenue grew to $115.2 million from $79.5 million, and gross profit climbed to $30.1 million from $8.2 million — a divergence that points to the heart of the quarter: the deliberate pivot away from self-mining toward higher-margin AI and HPC colocation. Self-mining revenue collapsed to $30.1 million from $67.2 million, a 55% drop the company attributed to a 45% decline in bitcoin mined and an 18% drop in average BTC price. Colocation revenue did the opposite, jumping to roughly $77.5 million from $8.6 million. CORZ slipped 7% in after-hours trading on May 6 and was down another 6.25% pre-market on Thursday.
Why it matters
The stock is being repriced around a strategic transition, not the GAAP loss. CEO Adam Sullivan framed the quarter around "investing ahead of contracts" and accelerating ready-for-service dates across multiple sites — language that signals capex-heavy ramp into AI/HPC tenants rather than a mining-margin squeeze. Core Scientific has been stockpiling the capital to do it: a $3.3 billion private debt offering earlier this year, plus $1 billion in loans from JPMorgan and Morgan Stanley, alongside the freshly announced $421 million acquisition of Oklahoma-based miner Polaris DS, which adds 440 MW of contracted power through Oklahoma Gas & Electric. The impairment-driven loss, in other words, reflects the cost of the old book being written down while the new colocation build-out gets capitalized — a familiar pattern in mining-to-AI pivots.
Market impact
The colocation line is now larger than self-mining, a structural milestone. Core Scientific's total gross power pipeline has expanded to 4.5 GW, with planned 1.5 GW additions in Muskogee, Oklahoma, and Pecos, Texas, and a $233 million land-and-power deal in Texas supporting roughly 430 MW.
Frequently asked questions
-
Why did Core Scientific swing to a net loss in Q1 2026 despite higher revenue?
The $347.2 million Q1 net loss was driven by $266.5 million in non-cash impairment charges and a $30.8 million non-cash loss tied to changes in the fair value of warrants and contingent value rights, even as total revenue and gross profit both grew year over year.
-
How is Core Scientific's revenue mix shifting away from bitcoin mining?
Colocation revenue jumped to roughly $77.5 million in Q1 2026 from $8.6 million a year earlier, while self-mining revenue fell to $30.1 million from $67.2 million. Core Scientific attributed the self-mining decline to a 45% drop in bitcoin mined and an 18% drop in average BTC price, alongside a strategic shift to…
-
What is the Polaris DS acquisition and why does it matter for Core Scientific?
Core Scientific announced a $421 million deal to acquire Oklahoma-based bitcoin miner Polaris DS LLC, which is expected to give the company access to 440 megawatts of contracted power through Oklahoma Gas & Electric, supporting its AI and HPC colocation build-out.
-
How much total power capacity is Core Scientific developing for AI and HPC?
Core Scientific expanded its total gross power capacity pipeline to 4.5 GW, including planned 1.5 GW expansions at campuses in Muskogee, Oklahoma, and Pecos, Texas. A separate $233 million land-and-power deal in Texas is expected to support roughly 430 MW of gross power capacity.
-
How is Core Scientific funding its expansion into AI infrastructure?
Earlier in 2026, Core Scientific announced a $3.3 billion private debt offering and secured loans totaling $1 billion from JPMorgan and Morgan Stanley to fund the pivot toward AI and high-performance computing colocation.
TheBlock