The Ethereum co-founder says Bitcoin mining splits its bills between hardware and power, while AI inference mostly pays for chips
Vitalik Buterin has a new side project: comparing electricity bills. In an October 17, 2026 post on X, the Ethereum co-founder contrasted how money gets spent in Bitcoin mining versus AI inference.
His core point is simple. Bitcoin mining costs are divided between hardware and electricity, while AI inference spends mostly on hardware.
The numbers behind the split #
Estimates tied to the comparison put the cost of mining one Bitcoin at roughly $93,700. Of that, about $37,600, or 40%, goes to compute capital expenditure, the upfront cost of buying mining machines.
Electricity accounts for approximately $40,600, or 43.3%. Put another way, the bill for the rigs and the bill for the power that runs them are close to a coin flip.
AI inference tells a very different story. Inference is the stage where a trained model actually answers questions, generates images or writes code for users.
The estimated cost there runs about $26,260 per GPU-year. Compute capex makes up roughly $19,900 of that, or 76%.
Electricity? Somewhere between $1,050 and $1,260, which works out to just 4% to 4.8% of the total. For an AI operator, the power bill is closer to a rounding error than a strategic concern.
The gap gets wider at the facility level. Building and operating a Bitcoin mining site is estimated at $700,000 to $1 million per megawatt.
The news moving money, markets, and the world—before your day starts.
Daily. Free. Join 34,000+ readers across crypto, finance, and policy.
AI infrastructure comes in at around $20 million per megawatt.
Why the cost structure matters more than the total #
Buterin’s analysis focuses on how costs are composed, not on which sector spends more or burns more energy overall. It does not attempt to equate absolute spending or total energy use between the two industries.
Bitcoin miners remain highly sensitive to ongoing power costs, since electricity eats such a large share of what it takes to produce each coin. AI operators’ economics hinge on hardware efficiency and utilization, meaning how much useful work each expensive GPU performs over its life.
Background: two industries chasing the same power #
Bitcoin mining and AI data centers increasingly compete for similar resources: grid connections, cheap power and physical space.
Mining is built around a proof-of-work system, where machines race to solve puzzles and the winner earns newly issued Bitcoin. Electricity is a structural input to that process, which helps explain why it shows up as such a large slice of the cost estimate.
Ethereum, the network Buterin co-founded, moved away from proof-of-work years ago. So when he weighs in on Bitcoin’s energy economics, he is speaking as someone whose own chain no longer pays that particular bill.
AI inference, by contrast, lives and dies on specialized chips. Those chips are costly, and the data centers that house them are far more expensive to build per megawatt, as the facility estimates show.
What this means for miners and investors #
If electricity makes up more than 40% of the cost to mine a coin, miners with access to cheap, stable energy hold a structural edge, while those exposed to volatile power markets carry real margin risk. Some miners could explore using their infrastructure for AI applications where capital efficiency may be greater. However, mining sites cost an estimated $700,000 to $1 million per megawatt, while AI infrastructure runs around $20 million per megawatt. Any miner eyeing AI would be signing up for a radically different capital commitment, swapping electricity risk for hardware and utilization risk.
For the AI side, with electricity at roughly 4% to 4.8% of inference costs, the pressure falls on keeping GPUs busy and efficient, since idle hardware at $19,900 a year in capex is an expensive thing to leave sitting around. Disclosure: This article was edited by Estefano Gomez. For more information on how we create and review content, see our