Core Scientific Plans to Exit Bitcoin Mining by Year-End as AI Growth Ramps Up

Core Scientific says it expects to fully leave Bitcoin mining by the end of 2026, positioning itself as a “pure-play” high-density colocation provider focused on artificial intelligence and high-performance computing. The shift marks the latest step in an industry-wide retooling of data-center capacity away from self-mining and toward AI infrastructure.

Speaking during a Q3 earnings call, Chief Financial Officer Jim Nygaard said Core Scientific plans to complete the wind-down during the fourth quarter and “enter 2027 as a pure-play, high-density colocation provider,” according to the webcast at event.choruscall.com. The company previously disclosed a timetable for ending hosted mining operations in a July quarterly filing, including a stated end date of Dec. 31 as it continued winding down its self-mining activity.

Key takeaways

• Core Scientific targets a complete exit from Bitcoin mining by the end of 2026, with a fourth-quarter wind-down planned and 2027 framed as a colocation-only year.


• In Q3, Bitcoin mining contributed only $17.5 million—7.5%—of total revenue of $232.9 million, down from the dominance of mining in earlier periods.


• Self-mining revenue fell sharply year-over-year in Q3, reflecting the company’s ongoing transition away from running its own mining operations.


• AI-focused hosting revenue and colocation services have become the core of Core Scientific’s business mix as its miner footprint shrinks.


• The company’s pivot is unfolding alongside broader political scrutiny of data-center expansion in the US, particularly around power demand and local impacts.

From mining to colocation: the 2026 endpoint

Core Scientific’s latest guidance tightens the timeline for its mining exit. While the July filing already indicated hosted mining would end by Dec. 31 as self-mining continued to be wound down, the earnings call comment provided what amounts to a targeted “completion” date for leaving Bitcoin mining altogether.

That matters because it clarifies the direction of the company’s capital allocation and operational planning. For investors and counterparties, a defined endpoint helps distinguish between transitional revenue streams (still tied to mining operations) and the long-term model the company intends to operate going forward—high-density colocation for compute-intensive workloads.

Nygaard also noted that Core Scientific had nearly 50% fewer miners online at the end of September than at the end of 2025, reinforcing that the mining portion of the business has been contracting well before the end-of-2026 target.

Q3 results show mining’s shrinking role

Core Scientific’s third-quarter results, published by the company at investors.corescientific.com, illustrate how quickly mining is losing economic weight relative to colocation.

According to those results:

• Self-mining revenue in Q3 fell to $12.3 million from $57.4 million a year earlier.


• Hosting revenue for other miners dropped to $5.2 million from $8.7 million.


• Bitcoin mining generated $17.5 million in Q3, accounting for 7.5% of total revenue of $232.9 million.


• Colocation services brought in $215.4 million during the same quarter.

In other words, even within the broader category of “Bitcoin mining,” the company is now operating in a relatively minor revenue lane compared with colocation. For market watchers, this provides a practical signal: the longer-term story is no longer about hashpower performance, but about compute capacity utilization, uptime, and demand from non-mining clients.

How the AI pivot developed

Core Scientific’s move toward AI hosting accelerated after it began shifting its data-center strategy in 2024. The transition took shape through hosting agreements that gradually expanded the role of high-performance compute beyond mining.

As described in earlier coverage, Core Scientific signed an initial hosting agreement with CoreWeave in March 2024 and followed with a larger deal announced in June 2024—developments that were framed as complementary to its Bitcoin mining operations at the time, including language about preserving mining capacity (see earlier reporting linked in the source: cointelegraph.com and cointelegraph.com).

By late 2025, reporting indicated the company was already targeting an eventual reduction of mining operations farther out on the calendar. Earlier coverage from The Wall Street Journal stated that Core Scientific was targeting to wind down its Bitcoin mining operations by 2028 (wsj.com). Core Scientific’s newer 2026 exit plan therefore suggests the timeline has moved up, aligning the end of mining closer to the near term.

Mining pressures and what comes with the transition

The shift has not been costless. Mining economics have continued to weigh on financial performance, contributing to impairments and losses tied to assets and mining-related agreements.

In the first quarter of 2026, Core Scientific recorded $266.5 million in mining-related asset impairments, citing factors including weaker Bitcoin prices, historically low hashprice, and declining equipment values in the secondary market. In the second quarter, the company terminated its mining equipment agreement with Block and Proto, recording a $41.9 million loss as part of the transition, according to information cited in the source.

This helps explain why the company is now driving toward a clean operational split. For mining firms, the challenge has been that even when data-center utilization improves, lingering exposure to mining-specific economics can distort earnings volatility. A defined exit supports a steadier underwriting of revenue—assuming colocation demand for AI workloads continues to scale.

US scrutiny of data-center expansion

The mining-to-AI story is unfolding at the same time as growing political and community scrutiny of data-center growth in the US. According to Reuters, opponents of data-center development were among the winners in the Democrats’ sweeping midterm gains, with the outlet pointing to concerns that include electricity costs and the strain large facilities can place on local resources (Reuters report: reuters.com).

While Core Scientific’s mining exit is a company decision, its ability to execute on a long-term colocation and AI hosting strategy depends on the broader environment for permitting, power procurement, and local support. Readers should watch not only Core Scientific’s operational metrics, but also whether policy and infrastructure constraints affect timelines or capacity expansions across the sector.

Going forward, the most important question is whether Core Scientific’s colocation model can fully absorb what mining once provided—both in revenue stability and in asset utilization—without new bottlenecks in power, permitting, or customer demand. With the 2026 mining exit now explicitly targeted, the next updates to key operating statistics and contract performance will be where the plan either consolidates or starts to show friction.

https://www.cryptobreaking.com/core-scientific-plans-to-exit-2/?utm_source=blogger%20&utm_medium=social_auto&utm_campaign=Core%20Scientific%20Plans%20to%20Exit%20Bitcoin%20Mining%20by%20Year-End%20as%20AI%20Growth%20Ramps%20Up%20