According to Cointelegraph, Bitcoin mining companies are exploring diversification into data centers to enhance revenue. However, Phil Harvey, CEO of blockchain data center consulting firm Sabre56, argues that this shift to high-performance computing is impractical due to significant cost differences and logistical challenges.

Harvey highlighted that operating an AI or high-performance data center is far more expensive than running a crypto mining facility. He noted that a typical commercial mining operation costs between $300,000 and $350,000 per megawatt, whereas AI data centers require between $3 million and $5 million per megawatt—a 10-15x increase. Additionally, a mining operation with a gigawatt of available power can only convert approximately 200 megawatts for high-performance computing needs.

The CEO emphasized that only about 20% of a miner's portfolio could potentially support the necessary attributes like power, data, and land for AI. He also pointed out the significant physical space requirements, with crypto mining needing 1,000 square feet per megawatt compared to 5,000 square feet per megawatt for AI or HPC data centers. Furthermore, converting crypto mining facilities to AI and high-performance data centers involves substantial upfront costs, with over 90% of existing infrastructure needing replacement.

Bitcoin miners are seeking alternative revenue streams due to decreased block subsidies and reduced revenues, which hit an 11-month low in August. The idea of shifting to AI data centers and high-performance computing has been proposed, raising concerns about energy resource allocation between the emerging AI industry and crypto mining.

A report by VanEck suggested that Bitcoin mining companies could potentially earn $13.9 billion in annual revenues if they allocate 20% of their output to AI data processing and high-performance computing. However, the practical challenges and significant costs involved in such a transition make it a complex and potentially unfeasible solution for many mining companies.