Cambridge estimates Ethereum uses ~7.87 GWh/year post-Merge
The model ranks ETH near the low end of PoS energy intensity at ~33 kWh per $1M of market value, behind BNB Chain.
The Cambridge Centre for Alternative Finance published new post-Merge estimates that put Ethereum’s annual electricity consumption at about 7.87 GWh. On a market-value-adjusted basis, the study pegged Ethereum at roughly 33 kWh per $1 million, ranking second-lowest among the proof-of-stake networks assessed.
Key Takeaways
- Ethereum’s post-Merge electricity consumption was estimated at about 7.87 GWh per year by the Cambridge Centre for Alternative Finance.
- Market-value-adjusted energy intensity came in around 33 kWh per $1 million, ranking second-lowest among the assessed PoS networks behind BNB Chain.
- Solana was modeled at roughly 13.48 GWh per year and about 283 kWh per $1 million of market value, around 8.5x Ethereum’s intensity.
- The estimate was built from “at-the-wall” node measurements and a model using ~8,522 discoverable full nodes with a 64% cloud/enterprise and 36% residential hosting split.
Cambridge Puts Ethereum at ~7.87 GWh/Year Post-Merge
Cambridge Centre for Alternative Finance estimates Ethereum consumes about 7.87 gigawatt-hours of electricity annually in its post-Merge proof-of-stake configuration. The same work frames Ethereum’s remaining footprint as a node-driven operational load rather than a mining-driven industrial one, consistent with the post-September 2022 shift away from proof-of-work.
For traders, the actionable part is not the raw GWh figure in isolation. It is the study’s attempt to standardize comparisons across chains using an energy-intensity metric scaled by market value, which is the format most likely to be repeated in ESG screens, policy decks, and institutional allocation narratives.
PoS Energy-Intensity Scoreboard: Ethereum vs BNB Chain and Solana
On Cambridge’s market-value-adjusted metric, Ethereum is estimated at roughly 33 kilowatt-hours per $1 million of market value, the second-lowest among the PoS networks assessed, behind BNB Chain. That ranking matters because it positions ETH as “large cap but relatively low intensity” in the same framework used to compare peers.
The contrast is cleanest versus Solana in the figures provided. Cambridge modeled Solana at about 13.48 GWh per year, the highest electricity use among the PoS networks in the comparison set, and an energy intensity of roughly 283 kWh per $1 million of market value. That is around 8.5 times Ethereum’s intensity on the same yardstick.
Cambridge also put the combined electricity consumption of the networks included in its comparison at about 38 GWh. The excerpt does not enumerate the full list of chains in that set beyond Ethereum, BNB Chain, and Solana, which limits how far traders can take the “league table” beyond the headline pairwise comparisons.
Inside the Model: Node Power Draw, Client Mix, and Hosting Split
Cambridge anchored its Ethereum estimate in direct node measurements rather than top-down assumptions. “Cambridge measured how much electricity Ethereum nodes used at the wall across 20 combinations of the network’s main software clients.”
“It found that a typical home setup used about 18 watts, while a more powerful workstation used roughly 153 watts.” From there, “Using Ethereum’s mix of residential and professionally hosted nodes, the researchers estimated an average power draw of about 105 watts per node.”
The network-wide rollup depends on node counts and where those nodes run. “Cambridge counted around 8,522 discoverable full nodes, with 64% running in cloud or enterprise facilities and 36% on residential connections.” That structure implies future estimates can move even if Ethereum’s consensus rules do not, simply through shifts in discoverable node counts or a drift in hosting mix.
The study also estimated Ethereum’s electricity mix at about 56.4% renewable and nuclear versus 43.6% fossil fuels, and said remaining emissions are now driven mainly by the electricity grids supplying nodes.
Signals to Watch for Cambridge estimates Ethereum PoS energy
The first tell will be whether Cambridge publishes a full methodology link with the exact market-value snapshot used to compute kWh per $1 million. Without that timestamp, traders cannot cleanly map the intensity metric to a specific price regime.
Second, watch whether Ethereum’s discoverable full-node count materially diverges from the ~8,522 figure, and whether the 64% cloud/enterprise versus 36% residential split shifts. The model is sensitive to those inputs, so a hosting migration could change the footprint estimate without any protocol-level change.
Third, track whether policymakers or institutions start citing the ~7.87 GWh/year and ~33 kWh per $1 million figures in sustainability discussions. Narrative adoption is the transmission mechanism from a research estimate to real positioning.
Finally, any refreshed peer-chain estimates, especially Solana’s ~13.48 GWh/year and ~283 kWh per $1 million, could reshuffle the relative scoreboard and weaken or strengthen the ETH-versus-SOL contrast.
Why These Numbers Matter for ESG Narratives and Relative L1 Positioning
I treat this as a narrative catalyst with a credible methodology hook, not a protocol breakthrough. The Cambridge framing reinforces the post-Merge claim that Ethereum’s operational electricity footprint is low when scaled by market value, and that is the format ESG-sensitive allocators actually use when comparing large networks.
The threshold that matters is whether these specific figures become the default citations in institutional and policy contexts. If the ~33 kWh per $1 million metric sticks and peer estimates remain wide, the setup starts to look structural rather than narrative-driven, because it hardens ETH’s relative positioning against chains like Solana in sustainability screens that can influence capital access and reputational risk.
