Core Scientific to Exit Bitcoin Mining by Year-End

Core Scientific will complete its exit from bitcoin mining by year-end, management said on the company's third-quarter earnings call Thursday, formalizing a pivot toward high-density colocation that has already reshaped the firm's revenue mix and balance sheet. Chief Financial Officer Jim Nygaard told investors the company expects to "complete the wind-down during the fourth quarter," while Chief Executive Officer Adam Sullivan stated it plainly: "We expect to complete our exit from Bitcoin mining by year end." The comments accompanied a results release filed with the SEC, as reported by Bitcoin Magazine, that shows how far the shift has already progressed and what Core Scientific expects its business to look like once mining is fully gone from the books.

The Numbers Behind the Pivot

According to the earnings exhibit cited by Bitcoin Magazine, colocation revenue climbed nearly 60% sequentially to $215.4 million in the quarter ended Sept. 30. Over the same period, digital asset self-mining revenue fell to $12.3 million, down about 79% from $57.4 million a year earlier, while hosted mining for third parties contributed just $5.2 million. Of $232.9 million in total revenue, colocation now accounts for roughly 92%. The mining side of the business posted a $7.5 million gross loss in the quarter, in sharp contrast to colocation gross profit of $128.4 million, representing a 60% margin. Adjusted EBITDA more than tripled sequentially to $133.6 million, with the margin expanding to 57% from 25%.

Nygaard said Core Scientific had nearly 50% fewer miners online at the end of September than at year-end 2025, reflecting a steady process of repurposing power and floor space under long-term customer agreements for AI and high-performance computing workloads — a process the company's investor materials describe as continuing "as circumstances allow." The bitcoin balance on Core Scientific's books has shrunk in parallel: digital assets stood at just $2.7 million on Sept. 30, down from $49.7 million three months earlier and $222 million at year-end 2025, according to the earnings materials. Bitcoin Magazine notes that Core Scientific was still adding coins earlier in the year even as it sold into the pivot, and describes that chapter as effectively closed.

The wind-down has not been without cost. A $41.9 million loss on contract termination was booked in the second quarter, reflecting the decision to end a Block/Proto mining chip supply agreement, Bitcoin Magazine reported. Liquidity ended the third quarter at about $1.7 billion, including roughly $800 million in cash, after heavy capital spending on AMD-backed sites and the Polaris acquisition.

Scaling the Colocation Business

Core Scientific is billing for 485 megawatts of colocation capacity as of Sept. 30, up 90 MW in the quarter, representing more than $700 million in average annualized colocation GAAP revenue, according to the earnings materials. An additional roughly 100 MW under the company's CoreWeave buildout is in commissioning and is expected to start recognizing revenue over the next few months, Sullivan said on the call. Beyond CoreWeave's contracted 590 MW, Core Scientific has 530 MW under an AMD-backed agreement spanning five campuses, with initial delivery targeted for the first half of 2027 and full delivery by the end of 2028. Together, the CoreWeave and AMD relationships cover about 1.1 gigawatts. Once fully delivered and billing, management expects that combined portfolio alone to generate $1.8 billion in average annual run-rate revenue and $1.6 billion in net operating income.

Bitcoin Magazine's reporting situates Core Scientific's move within a broader industry pattern: peers among public bitcoin miners have been selling bitcoin and leasing campuses to AI tenants, with research notes flagging a widening funding gap between companies that can finance large data center builds and those that cannot. Sullivan framed the Core Scientific wind-down as simplifying the business toward "more attractive long-term growth opportunities," while Nygaard said starting 2027 with a "very clean P&L" — one without mining in the numbers — should give investors a clearer view of a true colocation run-rate.

What Comes Next

The near-term test, as Bitcoin Magazine frames it, is twofold: whether Core Scientific turns its last mining machines off on schedule during the fourth quarter, and whether the AMD financing package lands in time to keep that "clean P&L" growing.

Further reading / context