Ethereum Power Use Falls to 7.87 GWh in Cambridge Assessment

Cambridge Measures Ethereum at 7.87 GWh

Cambridge put Ethereum’s annual electricity consumption at 7.87 GWh and its ongoing power demand at 0.90 MW. Its estimate of 2.37 kilotonnes of carbon dioxide equivalent emissions is about 99.98% below the pre-Merge baseline it used for comparison.

Those numbers are far below proof-of-work Ethereum’s old profile. Before the Merge on September 15, 2022, Cambridge put the network’s power demand near 2.4 GW. Under proof of work, miners used hardware and electricity to compete for blocks. Under proof of stake, validators put ETH at risk as the security mechanism, removing that ongoing mining race.

The change does not mean the network has no physical footprint. Cambridge said 64% of the measured full-node infrastructure sits in cloud or data-center environments, while 36% is residential. The result is a smaller, more conventional IT-infrastructure footprint with a different set of concentration and reporting questions.

Ethereum’s own energy documentation reaches a similar conclusion while warning against a popular shortcut: energy per transaction. The site says that figure can be misleading because block-validation energy is not directly determined by the number of transactions in a block, and because layer-2 throughput complicates any single-chain denominator.

The gap between the old mining narrative and the new node narrative has become increasingly important as the network takes on more institutional roles. Daily Crypto Briefs recently covered how the Ethereum Foundation used AI agents to find protocol bugs, a reminder that operational scrutiny now reaches security and infrastructure as well as energy.

Proof of Stake Changed Ethereum’s Energy Equation

The Merge did not make Ethereum independent of electricity. It changed what electricity secures. Proof-of-work systems make energy expenditure part of the competition to add blocks. Ethereum’s proof-of-stake system instead relies on economic penalties for validators that act dishonestly or fail to meet the protocol’s rules.

That distinction helps explain the scale of the reduction. A proof-of-work network can attract more mining hardware when token prices and block rewards make it profitable. In Ethereum’s proof-of-stake design, adding validators does not create the same hardware arms race, although full nodes still need to run software, store data and communicate with peers.

Cambridge’s 2026 work also reflects a more granular methodology than broad country comparisons. The researchers measured client configurations, assigned node locations and considered the electricity mix of the grids where those nodes operate. The study said sustainable sources represented 56.4% of the measured mix, above the global average cited in the report.

That makes the remaining footprint more sensitive to geography. A node drawing from a cleaner grid has a different carbon effect from an otherwise similar node on a more fossil-heavy grid. It also means that an expansion in hosted infrastructure can change the estimate even without a fundamental change to Ethereum’s consensus rules.

This is a different energy story from Bitcoin mining, where power supply remains a central operating cost and a direct competitive input. The distinction has become sharper as miners pursue data-center contracts and AI workloads, a shift Daily Crypto Briefs tracked in its report on the Bitcoin miners’ AI funding gap.

Ethereum Node Geography Becomes the Next Climate Test

The new report points to infrastructure concentration as the issue to watch. Cambridge said the United States hosted 31% of the full nodes it counted, followed by Germany at 16%, Finland at 8% and France at 6%. Those four countries accounted for about 62% of the observed nodes.

No single number settles the decentralization question. Node counts do not fully capture validator concentration, client diversity or how quickly operators can move between hosts. But a geography and hosting map gives a more useful starting point than comparing today’s Ethereum with the mining-heavy network it replaced.

The same is true for climate disclosures. Cambridge’s earlier work explains that electricity use is only one part of an environmental assessment, while Ethereum’s documentation notes that estimates can vary with the underlying assumptions. Readers should treat annual figures as transparent estimates that need periodic revision, not as a fixed protocol property.

For institutions, the practical gain is comparability. A smaller energy footprint can simplify sustainability due diligence, but it does not remove the need to examine custody, software resilience, cloud dependence and the rules governing activity on the chain. The Ethereum Foundation’s recent staffing reset has kept those operational questions in view alongside the network’s technical roadmap.

Crypto risk appetite remains fragile. Alternative.me’s Crypto Fear and Greed Index registered 26, or Fear, on July 12.

The next signal is whether Cambridge publishes updated node and grid data as hosting patterns evolve. For now, the core result is narrow but consequential: Ethereum’s consensus system is no longer an industrial mining load, and the debate over its footprint has moved to the less visible infrastructure that keeps proof of stake online.