Ethereum co-founder Vitalik Buterin revealed on October 17, 2026, that Bitcoin mining costs are split roughly evenly between hardware and electricity, whereas Artificial Intelligence (AI) inference expenses are overwhelmingly dominated by compute capital expenditure. This financial divergence could allow AI firms to outbid cryptocurrency miners for access to power infrastructure.
Breakdown of Bitcoin Mining Costs
In a post shared on X, Buterin presented a detailed cost breakdown estimating total production expenses at approximately $93,700 per BTC. Direct power fed into mining machines accounts for $40,600 per coin (43.3%), while secondary electricity expenses add $2,400, bringing total energy costs to nearly 46%. Compute capital expenditure—covering hardware, cooling systems, power gear, and financing—totals $37,600 per coin, representing roughly 40% of expenses. The remaining 14% covers labor, facility rent, maintenance, insurance, bandwidth, and taxes.
This breakdown places production expenses significantly above JPMorgan's recent Bitcoin production cost estimate of $85,000. With total costs reaching $93,700 per coin, average miners face severe margin compression relative to market prices. The growing competition for power sites comes as analysts debate whether an AI mining pivot threatens Bitcoin security across energy-intensive facilities.
Why AI Operators Outbid Crypto Miners
Because AI inference relies heavily on expensive specialized chips, electricity accounts for a much smaller percentage of an AI firm's total operational spending. Consequently, AI data center operators can absorb higher electricity tariffs than traditional cryptocurrency miners, whose profit margins depend heavily on low energy rates.
Key takeaways from the cost analysis include:
- Cost Breakdown: Power accounts for 46% ($43,000 total) and compute capex accounts for 40% ($37,600) of Bitcoin production.
- Production Cost Disparity: Buterin's $93,700 estimate per BTC sits above JPMorgan's $85,000 benchmark.
- AI Infrastructure Shift: High hardware expenses allow AI data centers to pay premium rates for scarce power capacity.
Industry figures are actively attempting to bridge these competing business models. For example, BitMEX co-founder Arthur Hayes is constructing a Bitcoin-style AI mining network designed to compensate GPU hardware owners.
Why It Matters
This structural cost asymmetry creates long-term real estate and energy challenges for proof-of-work miners. As AI developers face data center construction delays, they are actively targeting grid connections currently leased by mining farms. If cheap capital favors AI expansion while cheap power remains the primary edge for miners, crypto operations risk being priced out of major power hubs unless they pivot to dual-use infrastructure.
