arXiv preprint ·
Quantifying Ethereum Energy Consumption via Network Mapping
- Published
- Source
- arXiv preprint, posted before peer review
- Original
- Read it at arXiv
- Topics
Abstract
Ethereum's electricity use fell by about 99.95% after the move from proof of work to proof of stake. Service providers still need to report operational energy use, e.g. under the EU Markets in Crypto-Assets Regulation (MiCAR). Existing estimates either apply one typical wattage to every node or start from aggregated monitoring counts. Both ignore attributes that nodes already advertise on the peer-to-peer network: client software, ARM or x86 hardware, hosting location, and validator role. We crawl the consensus and execution layers, assign each peer a wattage from those attributes using published measurements, and estimate the remaining incomplete peers with a Random Forest. On 6,934 peers from two Nebula crawls (19 and 22 June 2026), reachable nodes sum to 415 kW, or 3.63 GWh if that draw were held for a year. The same Lighthouse+Nethermind x86 wattage on every peer yields 431 kW. Observed attributes lower the total by 3.9%, mainly because nodes at Hetzner and other non-AWS clouds draw less than that home-desktop figure. AWS accounts for 15.6% of watts from 12.2% of peers, and validator-flagged nodes for 31.4% of watts from 25.7% of peers. The 415 kW snapshot is about 46% of the Cambridge Centre for Alternative Finance (CCAF) estimate of about 0.90 MW. Both use about 60 W per node, so the gap is mostly how many nodes each estimate includes. Rules cover 3,110 peers and the forest the other 3,824. On held-out labeled peers with client, architecture, and OS hidden, the forest's mean absolute error against the rule wattage is 4.3 W. Twenty-four-hour measurements on a gaming desktop differ from the predictions. After subtracting a 33 W idle graphics card that Ethereum clients do not need, both differences fall to about 19%.
The authors’ abstract, as arXiv lists it. arXiv’s listings, abstracts included, are free for anyone to reuse.
Read it on ReadSearch
Each opens after you create an account, and brings you back to this paper.
More in Computers & Society
Lessons from Trauma-Informed Training on Technology-Facilitated Abuse for Gender-Based Violence Advocates
AI-Enabled Quality Assurance for Multiple-Choice Assessment Items
LearnAdapt Praxis: Controlled AI Assistance and Evidence Traces for Adult Workplace Learning
Modeling Deletion Requests in Machine Unlearning
ReLEAF: A Socio-Technical Framework Bridging Custodians and Researchers for Trustworthy Data Sharing
Get the few papers worth your time in Computers & Society every morning.
ReadSearch reads the new papers in your field each day and brings you the few worth reading, each with its key finding in a sentence.