Assessing carbon emissions and energy consumption of US hyperscale data centers

  • Crawford, K. The Environmental Costs of Generative AI Are Skyrocketing and Mostly Secret. Nature 626693 (2024).

    CAS article Google Scholar

  • Erdenesanaa, DAI may soon need as much electricity as an entire country. New York Times (2023).

  • Patterson, D. et al. Carbon emissions and training on a large neural network. Preprint at https://doi.org/10.48550/arXiv.2104.10350 (2021).

  • Mytton, D. and Ashtine, M. Sources of data center energy estimates: a comprehensive review. Joule 62032 (2022).

    Google Scholar article

  • Shehabi, A. et al. 2024 U.S. Data Center Energy Consumption Report technology. representative. LBNL-2001637 (Lawrence Berkeley National Laboratory, 2024).

  • Han, T. et al. Design and regulate clean energy data centers. Nat. Reverend Clean Technol. 1373-374 (2025).

    Google Scholar article

  • Kamiya, G. & Bertoldi, P. Energy consumption in data centers and broadband communications networks in the EU (European Commission, Joint Research Centre, 2024).

  • Sun, K., Luo, N., Luo, X. and Hong, T. Prototype energy models for data centers. Energy construction. 231110603 (2021).

    Google Scholar article

  • Dayarathna, M., Wen, Y. and Fan, R. Data center energy consumption modeling: a survey. IEEE Communiqué. Survive. Tutorials 18732 (2015).

    Google Scholar article

  • Guidi, G. et al. The environmental burden of the US bitcoin mining boom. Nat. Common. 162970 (2025).

    CAS article Google Scholar

  • Stokel-Walker, C. The race for generative AI has a dirty secret. Wired (2024).

  • Strubell, E., Ganesh, A. and McCallum, A. Energy and policy considerations for deep learning in NLP. Preprint at https://doi.org/10.48550/arXiv.1906.02243 (2019).

  • Shehabi, A. et al. US Data Center Energy Consumption Report technology. representative. LBNL-1005775 (Lawrence Berkeley National Laboratory, 2016).

  • Ferreira, J., Callou, G., Josua, A. and Maciel, P. Estimating the environmental impact of data centers. In 2018 17th IEEE International Symposium on Networked Computing and Applications (NCA) 1 to 4 (IEEE, 2018).

  • How much electricity is consumed each year in the United States? (American EIA, 2024).

  • Data center energy needs could upend power grids and threaten the climate (Institute for Environmental and Energy Studies, 2025).

  • AB Siddik, M., Shehabi, A. & Marston, L. The environmental footprint of data centers in the United States. Approximately. Res. Lett. 16064017 (2021).

    Google Scholar article

  • Xiao, T., Fuso Nerini, F., Matthews, D., H., Tavoni, M. & You, F. Environmental impact and net zero emissions pathways for sustainable artificial intelligence servers in the United States. Nat. Sustain. 81541 (2025).

    Google Scholar article

  • Stoll, C., Klaaßen, L. & Gallersdörfer, U. The carbon footprint of Bitcoin. Joule 31647 (2019).

    Google Scholar article

  • National Energy Data System (NEDS): Fuel consumption by state. American EIA https://www.eia.gov/state/seds/sep_fuel/html/pdf/fuel_use_es.pdf (2023).

  • Brander, M. The most important GHG accounting concept you may not have heard of: the attribution-consequence distinction. Carbon management. 13337 https://doi.org/10.1080/17583004.2022.2088402 (2022).

  • Frequently asked questions about eGRID (U.S. EPA, 2024).

  • de Vries-Gao, A. The carbon and water footprints of data centers and what it could mean for artificial intelligence. Patterns 71 (2026).

    Google Scholar

  • Business purchases of clean energy increased 12% to a new record in 2023, according to BloombergNEF. BloombergNEF https://about.bnef.com/blog/corporate-clean-power-buying-grew-12-to-new-record-in-2023-according-to-bloombergnef/ (2024).

  • Stricter reporting requirements coming for data centers. Knowledge of the data center https://www.datacenterknowledge.com/regulations/tougher-reporting-mandates-ahead-for-data-centers (2023).

  • Dietrich, J. US Mandates Crypto Energy Reporting: Will Data Centers Be Next? Uptime Institute Blog https://journal.uptimeinstitute.com/us-mandates-crypto-energy-reporting-will-data-centers-be-next/ (2024).

  • Patel, P. et al. Characterizing energy management opportunities for LLMs in the cloud. In Proc. 29th ACM International Conference on Architectural Support for Programming Languages ​​and Operating Systems (ASPLOS ’24) Flight. 3, 207-222 (Association for Computing Machinery, 2024).

  • Ekvall, T., -M. Tillman, A. & Molander, S. Normative ethics and methodology for life cycle analysis. J. Cleaner Prod. 131225 (2005).

    Google Scholar article

  • Nordenstam, L. Attributional or consequential assessments in a cyclical greenhouse gas management process – comparison of guidelines on the use and production of electricity and district heating. J. Cleaner Prod. 317128214 (2021).

    Google Scholar article

  • Monthly electrical energy (American EIA, 2024).

  • U.S. Electricity Production, Capacity, and Sales (American EIA, 2023).

  • Xing, J. et al. Carbon Responder: coordinate demand response for the data center fleet. Preprint at https://doi.org/10.48550/arXiv.2311.08589 (2023).

  • Jagannadharao, A., Beckage, N., Nafus, D., and Chamberlin, S. Time-shifting strategies for carbon-efficient long-term large language model training. System of innovations. Software Eng. 21517-531 (2025).

    Google Scholar article

  • Dodge, J. et al. Measuring the carbon intensity of AI in cloud instances. Preprint at https://doi.org/10.48550/arXiv.2206.05229 (2022).

  • Gagnon, P. & Cole, W. Planning the evolution of the electricity network with a long-term marginal emission rate. iScience 25, 103915 (2022).

  • Prettenhofer, P. & Louppe, G. Gradient-boosted regression trees in Scikit-Learn. In PyData 2014 (NumFOCUS, 2014).

  • Istrate, R. et al. The environmental sustainability of digital content consumption. Nat. Common. 153724 (2024).

    CAS article Google Scholar

  • Niewenhuis, D., Talluri, S., Iosup, A. & De Matteis, T. FootPrinter: quantifying the carbon footprint of data centers. In Companion to the 15th ACM/SPEC International Conference on Performance Engineering, ICPE ’24 Companion 189-195 (Association for Computing Machinery, 2024).

  • Sarkar, S. et al. SustainDC: Benchmarking for sustainable data center control. In Proc. AAAI Artificial Intelligence Conference 22322-22330 (ACM, 2024).

  • Masanet, E., Shehabi, A., Lei, N., Smith, S. and Koomey, J. Recalibrating global data center energy consumption estimates. Science 367984 (2020).

    CAS article Google Scholar

  • Luers, A. et al. Will AI accelerate or delay the race to net zero emissions? Nature 628718 (2024).

    CAS article Google Scholar

  • Masanet, E. and Lei, N. How much energy do data centers actually use (Aspen Global Change Institute, 2020).

  • AI will drive a 165% increase in data center energy demand by 2030 (Goldman Sachs, 2025).

  • Newkirk, AC et al. Techno-economic analysis of microgrids for AI data centers in the continental United States. Research place https://doi.org/10.21203/rs.3.rs-8272920/v1 (2025).

  • Lei, N. & Masanet, E. Climate- and technology-specific PUE and WUE estimates for US data centers using a hybrid statistical and thermodynamic approach. Resource. Keep. Recycle. 182106324 (2022).

    Google Scholar article

  • Guidi, G. et al. Hyperscale broadcasts: code and reproducibility materials. GitHub https://github.com/gianguidi/hyperscale-emissions (2026).

  • Guidi, G. et al. Hyperscale broadcasts: v0.3.1-round3-final. GitHub https://github.com/gianguidi/hyperscale-emissions/tree/v0.3.1-round3-final (2026).

  • Gn headline

    Logo