Ethereum co-founder Vitalik Buterin outlined contrasting cost structures between Bitcoin mining and AI inference, noting that the two industries increasingly compete for the same power resources despite spending money very differently.
According to Buterin, Bitcoin mining costs are split roughly evenly between hardware and electricity. A breakdown he shared showed approximately $93,700 per coin in total mining costs, with electricity accounting for $40,600 per coin, or 43.3% of the total, plus an additional $2,400 from other electricity expenses. Compute capital expenditure—covering hardware, cooling, and power infrastructure—runs about $37,600, representing roughly 40% of mining costs. Labor, facility rent, maintenance, insurance, taxes, and bandwidth make up the remaining approximately 14%.
The total estimate exceeds JPMorgan's recent Bitcoin production cost estimate of $85,000 and currently sits above Bitcoin's market price, meaning average miners would not fully cover their costs at present levels.
Contrasting Economics for AI Operations
AI inference operates under a fundamentally different model. Hardware costs dominate the economics, with electricity representing a much smaller share of expenses. This structure allows AI operators to absorb higher electricity prices than Bitcoin miners can tolerate, since power comprises a larger portion of mining's total cost burden.
The cost difference is already influencing competition for grid capacity. Many Bitcoin miners hold permitted grid access that AI developers cannot rapidly expand, as local data center restrictions tighten. Some industry observers question whether AI demand could displace mining from premium power locations.
Others are exploring hybrid approaches. BitMEX co-founder Arthur Hayes is developing a Bitcoin-style AI mining network that compensates GPU owners, attempting to merge elements of both models.
If these cost structures hold, Buterin's analysis suggests cheap power remains mining's competitive advantage, while cheap capital access favors AI operators. Future competition for energy sites may increasingly depend on which industry can finance hardware acquisition most rapidly.
