Mining Firms Pivot Data Centers to Support AI Services
Providers are repurposing crypto-mining sites to secure multi-year AI cloud service contracts through 2028.
Updated on Oct. 23, 2026 in Data Centers
Live Poll
Is shifting data-center capacity from crypto mining to AI computing a sound long-term business strategy?
As of Q2 2026, major cryptocurrency infrastructure providers are aggressively repurposing their power-intensive data centers for artificial intelligence workloads. Companies including IREN, TeraWulf, and CoreWeave are transitioning away from mining equipment to lease capacity or provide managed GPU cloud services.
Why it matters
The migration of mining infrastructure to AI helps alleviate the acute shortage of high-density data center capacity and stable electrical power required by large-scale GPU deployments. This pivot allows mining firms to capture significantly higher margins in the AI sector while leveraging their existing grid-connected assets.
CoreWeave reported $2.575 billion in revenue for the quarter ended June 30, 2026, alongside $1.393 billion in depreciation and $640 million in interest expenses. This financial profile reflects the massive capital expenditure required to transition from mining to AI-ready cloud compute.
The players
IREN
A digital infrastructure company transitioning its data center capacity from Bitcoin mining to GPU-based cloud services.
TeraWulf
A data center developer focused on high-density infrastructure leasing for artificial intelligence and compute-heavy workloads.
CoreWeave
A specialized cloud service provider that scaled from crypto-mining origins to operating large-scale GPU clusters for AI.
Anthropic
An AI research and deployment company requiring large-scale computing power to train and run its frontier models.
Microsoft
A major cloud service provider and software entity integrating specialized GPU cloud services into its broader infrastructure.
The details
Data center providers are gutting specialized mining hardware to install high-performance GPU clusters which require significantly different cooling and power distribution layouts. By leveraging existing energy contracts and grid interconnections—the physical links connecting a facility to the electrical power system—these companies fast-track site deployment for AI developers. This involves upgrading facility electrical infrastructure to handle the distinct power draw characteristics of AI server racks versus specialized mining rigs.
Timeline
2016: CoreWeave mined its first Ethereum block.
Sept. 15, 2022: Ethereum ended its proof-of-work mining mechanism.
November 2025: IREN announced a GPU cloud services contract with Microsoft.
June 30, 2026: CoreWeave recorded quarterly financials.
Dec. 31, 2026: IREN targets the completion of its transition from mining to AI.
The Tech Race
This transition marks a strategic pivot away from the business model established after the 2022 Ethereum network upgrade. Mining firms are now competing to become primary AI infrastructure hosts to gain long-term service contracts with hyperscalers.
This conversion increases the total available compute power for AI developers, potentially accelerating the speed at which new machine learning models reach production. Commercial users may see improved availability for high-end GPU cloud instances as these repurposed facilities come online through 2028.
The takeaway
Mining firms are betting that high-margin AI compute contracts offer a more sustainable future than the volatile economics of cryptocurrency. Investors should monitor the gap between capital expenditures and realized revenue as these firms scale their AI capacity through 2028.
What happens next
IREN aims to finalize its transition by Dec. 31, 2026, while TeraWulf expects to bring its 401-megawatt Kentucky site to full capacity by early 2028.
Further reading
Learn more about the shifting demands of the Data Centers industry.
Source note: This article includes information reported by TokenPost.
Live Poll
Is shifting data-center capacity from crypto mining to AI computing a sound long-term business strategy?
