Ethereum consumes 7.87 gigawatt-hours of electricity per year and burns roughly 33 kilowatt-hours for every $1 million of market value, making it the second most energy-efficient proof-of-stake network that Cambridge researchers bothered to measure. Solana, by contrast, devours 283 kWh per $1 million of market cap, about 8.5 times the energy intensity of Ethereum. The findings from the Cambridge Centre for Alternative Finance offer the most granular look yet at where the network stands nearly four years after the Merge.
The numbers arrive at an opportune moment. Institutional allocators still treat “crypto” as a monolith when ESG scoring comes up, and policymakers drafting electricity-disclosure rules have mostly relied on outdated proof-of-work estimates. This study gives both groups actual data to argue over instead of vibes.
The Methodology: Wall Sockets and Client Software
Cambridge did not estimate Ethereum’s energy use by extrapolating from hashrate or staking rewards. The researchers measured electricity draw at the wall across 20 combinations of Ethereum’s main execution and consensus clients. A home-grade setup pulled about 18 watts; a beefier workstation drew around 153 watts. Weighting these by the network’s mix of residential and professionally hosted nodes, Cambridge landed on an average of 105 watts per node.
The research team counted 8,522 discoverable full nodes. Roughly 64% of them run in cloud or enterprise data centers, while 36% sit on residential connections. That split matters because data-center grids tend to report cleaner energy mixes than the average suburban outlet, but also because professional hosting concentrates more hardware in fewer geographic locations, creating subtle centralization vectors that sustainability reports rarely discuss.
From those per-node wattage figures and node counts, Cambridge arrived at 7.87 GWh for the whole network annually. To put that in context, the average American household consumes about 10.5 megawatt-hours per year. Ethereum’s entire validator set uses roughly the electricity of 749 U.S. homes. Before the Merge in September 2022, the network’s mining rigs were burning something like 21 terawatt-hours, so the 99.9% reduction estimate you have probably seen cited elsewhere holds up.
How Ethereum Stacks Up Against Other PoS Chains
Raw gigawatt-hours only tell part of the story. A network that secures $240 billion in market cap should logically use more electricity than one securing $2 billion, all else equal. Cambridge adjusted for this by computing kilowatt-hours per $1 million of market value, an energy-intensity metric that lets observers compare chains on a level playing field.
On that basis, BNB Chain came out on top with the lowest energy intensity, although the study did not publish its exact figure. Ethereum ranked second at 33 kWh per $1 million. Solana, despite its reputation for speed and low fees, landed at 283 kWh per $1 million, roughly 8.5 times Ethereum’s intensity. The five PoS networks in the comparison consumed about 38 GWh combined, meaning Solana’s 13.48 GWh accounted for more than a third of the total.
Why the gap? Solana’s architecture prioritizes throughput. Its validators run on high-spec hardware (often bare-metal servers with 256 GB of RAM and NVMe SSDs) to keep up with the network’s 400-millisecond block times. Ethereum validators, by contrast, can run on a Raspberry Pi in theory and a modest desktop in practice. The trade-off is throughput: Ethereum’s base layer processes about 15 to 30 transactions per second without Layer 2 rollups, while Solana’s mainnet has handled bursts above 2,000 TPS.
Neither design is “right.” Solana’s approach sacrifices hardware accessibility for speed; Ethereum’s approach sacrifices base-layer speed for decentralization and lower energy overhead. The Cambridge data simply quantifies what that choice costs in kilowatt-hours.
Grid Mix and Residual Emissions
Ethereum’s post-Merge emissions are now driven almost entirely by where its nodes plug in. Cambridge estimated that 56.4% of the network’s electricity comes from renewable and nuclear sources, while 43.6% still flows from fossil fuels. That ratio roughly mirrors the global electricity grid, which is neither a surprise nor a vindication. Validators cluster in jurisdictions with cheap power, reliable internet, and crypto-friendly regulation (Germany, the United States, Finland, Singapore), and those regions’ grid mixes vary wildly.
If you want to obsess over carbon rather than raw energy, the 43.6% fossil share is the number to watch. Ethereum’s 7.87 GWh at 43.6% fossil intensity produces a carbon footprint that is negligible compared to the network’s pre-Merge days, but it is not zero. A validator operator in West Virginia (coal-heavy grid) emits more than one in Iceland (geothermal-heavy grid) for the same software work. The Cambridge study stopped short of calculating a network-wide CO2 figure, likely because the error bars on geographic node distribution and grid-emission factors would have been enormous.
What This Means for Institutional Allocators
ESG-constrained capital has historically avoided crypto exposure because the asset class was synonymous with Bitcoin mining. That narrative is outdated for Ethereum but sticky. The Cambridge study gives compliance officers and ESG analysts a citable, peer-adjacent source to distinguish Ethereum from proof-of-work chains.
Consider the practical implications. A pension fund evaluating a spot Ethereum ETF can now point to a 33 kWh/$1M intensity figure and compare it to other portfolio holdings. The average kilowatt-hour per dollar of market cap for the S&P 500 is not published in a single place, but rough estimates for energy-intensive industrials run into the hundreds of kWh per $1M of enterprise value. Ethereum, by this metric, is less energy-intensive than running a steel mill or operating a data center for cloud computing (the irony being that many Ethereum validators run in those same cloud data centers).
For allocators who care about this stuff, the Cambridge data is a permission slip. For allocators who do not, it changes nothing. But the direction of regulatory travel (EU taxonomy, SEC climate-disclosure rules, UK Stewardship Code updates) suggests more capital will have to care in the coming years.
The Ethereum Foundation’s recent treasury management has drawn scrutiny for its ETH sales, but sustainability reporting may help the Foundation reframe its narrative around the network’s environmental credentials rather than its token-sale schedule.
Policy Implications and What Comes Next
Lawmakers drafting blockchain-energy-disclosure rules have historically reached for Bitcoin mining stats because those are the biggest, scariest numbers available. The Cambridge study gives legislators a more nuanced dataset. A blanket “crypto is bad for the climate” stance now requires ignoring peer-reviewed research, which is politically harder than ignoring Twitter threads.
The European Union’s MiCA regulation already distinguishes between consensus mechanisms for some disclosure purposes. The United States has no equivalent framework, but the Environmental Protection Agency and the Department of Energy have both floated blockchain-energy studies in recent years. Cambridge’s methodology (measure at the wall, weight by node distribution, adjust for market cap) could become a template for future regulatory assessments.
Ethereum’s core developers have discussed further efficiency improvements under the “Lean Ethereum” roadmap, which Vitalik Buterin outlined earlier this year. Proposals include reducing validator set sizes, pruning historical state, and adjusting attestation duties to lower per-node compute requirements. None of these changes would move the needle dramatically on a network already at 7.87 GWh, but they signal that efficiency remains a design priority.
Solana, for its part, has explored compressed transactions and other optimizations that could reduce hardware requirements for validators. Whether those changes materially close the 8.5x energy-intensity gap is an open question. The trade-off calculus (throughput vs. Decentralization vs. Energy) is not one that any network has definitively solved.
For readers tracking the broader market, Ethereum’s sustainability profile is one factor among many influencing its position relative to competitors. USDT recently flipped Ethereum for second place by market cap for the first time since 2018, a reminder that market-cap rankings shift on flows, not just fundamentals. Energy intensity is a long-term input, not a short-term price driver.
The Cambridge Centre for Alternative Finance plans to update its estimates periodically as Ethereum’s node count and client distribution evolve. The next major data point will likely arrive after the Pectra upgrade (expected later this year), which could shift the client mix and validator hardware profiles.
Until then, the headline numbers stand: 7.87 GWh annually, 33 kWh per $1 million of market cap, and an 8.5x efficiency advantage over Solana. Whether those figures change anyone’s allocation decision depends on how much weight ESG scoring carries in their process. The data, at least, is now on the table.
