With rewards shrinking post-halving, a growing number of bitcoin miners are pivoting into or including artificial intelligence infrastructure to the portfolio. These firms are repurposing their vast data centers, once used to the famous Proof of Work (PoW) calculations, to host AI workloads, lured by significantly higher margins and booming demand from cloud and generative model providers.
According to a recent report published by American consulting firm RSM, the profitability of AI hosting can surpass crypto mining by 60% to 100% depending on electricity costs and contract terms. This is because the infrastructure needs of generative AI are colossal. Companies like OpenAI and Anthropic require powerful GPUs, stable power supply, and massive cooling, all features already embedded in many mining sites.
This July, Core Scientific, one of the largest bitcoin miners in the U.S., announced a US$ 3.5 billion, 12-year deal with CoreWeave to host AI computing workloads in its Texas and North Dakota facilities. The agreement will redirect 200 MW of capacity away from bitcoin mining and into GPU cloud hosting, marking one of the largest crypto-to-AI infrastructure pivots to be registered until now. The deal will generate up to US$ 300 million in revenues annually, compared to under $100 million from bitcoin mining.
This is not just a survival play. It’s an emerging business model. Miners are realizing that their assets, like access to cheap power and cooling-ready data centers, are scarce and valuable among the AI boom. The average ROI timeline for a mining facility converted to AI hosting has decreased from 5 years to under 3, still according to RSM.
But the shift its not going to happen without friction. AI workloads require different hardware (primarily Nvidia H100 or A100 GPUs), more robust networking, and complex orchestration. This means miners need capital, new partnerships, and expertise in GPU cloud services, areas where traditional cloud providers still dominate. Still, firms like Core Scientific and Hut 8 are raising capital or merging to fund the transition, often betting on hybrid models that keep some crypto exposure while scaling AI hosting capabilities.
In essence, the line between bitcoin mining and AI infrastructure is beginning to blur. As the digital economy evolves, miners are repositioning themselves as the powerhouses behind the next wave of computing. With AI demand soaring and crypto returns under pressure, the move isn’t just strategic, it may be a matter of survival for these companies.