The DOI matches its DataCite record; 18 of the 18 references it cites were found in a registry or an index.
Quantifying the Carbon Emissions of Machine Learning
Open accessAbstract
From an environmental standpoint, there are a few crucial aspects of training a neural network that have a major impact on the quantity of carbon that it emits. These factors include: the location of the server used for training and the energy grid that it uses, the length of the training procedure, and even the make and model of hardware on which the training takes place. In order to approximate these emissions, we present our Machine Learning Emissions Calculator, a tool for our community to better understand the environmental impact of training ML models. We accompany this tool with an explanation of the factors cited above, as well as concrete actions that individual practitioners and organizations can take to mitigate their carbon emissions.
Overview
Integrity signals
Registry
Cites
Statistics
Verification
Claims the paper makes
Claims other papers make about this one
Each is checked against this paper’s own text: does it say what the citing paper says it does?
In the citing paper
“the 'Machine Learning Emissions Calculator' for estimating the carbon footprint of GPU compute through specifying hardware type, hours used, cloud provider, and region [12], Anthony et al. [3]”
From Sustainable AI: AI for sustainability and the sustainability of AI (2021) · cites this paper as [12]
In the citing paper
“In Section 5, we discuss how while the estimation methods of these works provide some understanding of carbon and energy impacts, nuances in the estimation methods may make them inaccurate-particularly in experiments which utilize combined CPU and GPU workloads heavily.”
From Towards the Systematic Reporting of the Energy and Carbon Footprints of Machine Learning (2020) · cites this paper as Lacoste et al., 2019
In the citing paper
“Some GPUs have substantially higher efficiency in terms of floating point operations per second (FLOPS) per watt of power usage compared to others (Lacoste et al., 2019).”
From Carbontracker: Tracking and Predicting the Carbon Footprint of Training Deep Learning Models (2020) · cites this paper as Lacoste et al., 2019
In the citing paper
“Estimations were conducted using the Machine Learning Impact calculator presented in [66].”
From SatMAE: Pre-Training Transformers for Temporal and Multi-Spectral Satellite Imagery (2022) · cites this paper as [66]
In the citing paper
“For a model pre-trained and finetuned on fMoW Sentinel dataset, total emissions are estimated to be 109.44 kg CO 2 eq.”
From SatMAE: Pre-Training Transformers for Temporal and Multi-Spectral Satellite Imagery (2022) · cites this paper as [66]
In the citing paper
“Total emissions are estimated to be 136.8 kg CO 2 eq. of which 100 percent was directly offset by the cloud provider.”
From SatMAE: Pre-Training Transformers for Temporal and Multi-Spectral Satellite Imagery (2022) · cites this paper as [66]
Citations
Cites
Very Deep Convolutional Networks for Large-Scale Image Recognition
Simonyan, K., & Zisserman, A. (2014) · arXiv · doi:10.48550/arxiv.1409.1556 · Cited by 114,148 · cited 1×
“For a model such as VGG [18] or BERT [19], which are trained on multiple GPUs for several weeks, this can correspond to avoiding emitting several hundreds of kilograms of CO 2 eq by training on a server powered by hydroelectricity instead of fossil fuels.”
MethodBERT: Pre-training of Deep Bidirectional Transformers for Language Understanding
Devlin, J., Chang, M.-W., Lee, K., & Toutanova, K. (2018) · arXiv · doi:10.48550/arxiv.1810.04805 · Cited by 73,173
BERT: Pre-training of Deep Bidirectional Transformers for Language Understanding
Devlin, J., Chang, M.-W., Lee, K., & Toutanova, K. (2019) · Proceedings of the 2019 Conference of the North American Chapter of the Association for Computational Linguistics: Human Language Technologies, Volume 1 (Long and Short Papers) · doi:10.18653/v1/n19-1423 · Cited by 54,325 · cited 1×
“For a model such as VGG [18] or BERT [19], which are trained on multiple GPUs for several weeks, this can correspond to avoiding emitting several hundreds of kilograms of CO 2 eq by training on a server powered by hydroelectricity instead of fossil fuels.”
Method(2012) · Cited by 8,009
Universal Language Model Fine-tuning for Text Classification
Howard, J., & Ruder, S. (2018) · Proceedings of the 56th Annual Meeting of the Association for Computational Linguistics (Volume 1: Long Papers) · doi:10.18653/v1/p18-1031 · Cited by 4,100 · cited 1×
“Notably, recent research has shown that using pre-trained models with task-specific fine-tuning performs as well as training from scratch, while being more robust, for tasks in image recognition [7, 8] and NLP [9].”
BackgroundEnergy and Policy Considerations for Deep Learning in NLP
Strubell, E., Ganesh, A., & McCallum, A. (2019) · Proceedings of the 57th Annual Meeting of the Association for Computational Linguistics · doi:10.18653/v1/p19-1355 · Cited by 3,684 · cited 2×
“This trend has been the subject of recent studies aiming to evaluate the climate impact of AI, which have predominantly put the focus on the environmental cost of training large-scale models connected to grids powered by fossil fuels [1, 2].”
BackgroundConvolutional Neural Networks for Medical Image Analysis: Full Training or Fine Tuning?
Tajbakhsh, N., Shin, J. Y., Gurudu, S. R., Hurst, R. T., Kendall, C. B., Gotway, M. B., & Liang, J. (2016) · IEEE Transactions on Medical Imaging · doi:10.1109/tmi.2016.2535302 · Cited by 3,277 · cited 1×
“Notably, recent research has shown that using pre-trained models with task-specific fine-tuning performs as well as training from scratch, while being more robust, for tasks in image recognition [7, 8] and NLP [9].”
BackgroundHyperband: a novel bandit-based approach to hyperparameter optimization
Li, L., Jamieson, K., DeSalvo, G., Rostamizadeh, A., & Talwalkar, A. (2017) · Journal of Machine Learning Research · Cited by 930
Fine-tuned Language Models for Text Classification.Preprint
Howard, J., & Ruder, S. (2018) · arXiv (Cornell University) · Cited by 357
Food image recognition using deep convolutional network with pre-training and fine-tuning
Yanai, ., & Kawano, Y. (2015) · IEEE International Conference on Multimedia & Expo Workshops (ICMEW) · doi:10.1109/icmew.2015.7169816 · Cited by 339 · cited 1×
“Notably, recent research has shown that using pre-trained models with task-specific fine-tuning performs as well as training from scratch, while being more robust, for tasks in image recognition [7, 8] and NLP [9].”
Background
Cited by
Sorted by citations
Kirillov, A., Mintun, E., Ravi, N., Mao, H., Rolland, C., Gustafson, L., Xiao, T., Whitehead, S., Berg, A. C., Lo, W.-Y., et al. (2023) · 2023 IEEE/CVF International Conference on Computer Vision (ICCV) · doi:10.1109/iccv51070.2023.00371 · Cited by 11,953
Large language models in medicine
Thirunavukarasu, A. J., Ting, D. S. J., Elangovan, K., Gutierrez, L., Tan, T. F., & Ting, D. S. W. (2023) · Nature Medicine · doi:10.1038/s41591-023-02448-8 · Cited by 3,944
“This energy-intensive infrastructure has a substantial carbon footprint, and considerable investment is committed to improving hardware and software efficiency to minimize the environmental costs of development 33–36 .”
BackgroundOn the Opportunities and Risks of Foundation Models
Bommasani, R., Hudson, D. A., Adeli, E., Altman, R., Arora, S., von Arx, S., Bernstein, M. S., Bohg, J., Bosselut, A., Brunskill, E., et al. (2021) · arXiv · doi:10.48550/arxiv.2108.07258 · Cited by 2,676
Schwartz, R., Dodge, J., Smith, N. A., & Etzioni, O. (2020) · Communications of the ACM · doi:10.1145/3381831 · Cited by 1,647
Sustainable AI: AI for sustainability and the sustainability of AI
van Wynsberghe, A. (2021) · AI and Ethics · doi:10.1007/s43681-021-00043-6 · Cited by 939
“With tools like the ‘machine learning emissions calculator’ [12] and the ‘experiment-impact-tracker’ [10] this should no longer be the case.”
MethodA high-resolution canopy height model of the Earth
Lang, N., Jetz, W., Schindler, K., & Wegner, J. D. (2023) · Nature Ecology & Evolution · doi:10.1038/s41559-023-02206-6 · Cited by 810
“59).”
Energy and Policy Considerations for Modern Deep Learning Research
Strubell, E., Ganesh, A., & McCallum, A. (2020) · Proceedings of the AAAI Conference on Artificial Intelligence · doi:10.1609/aaai.v34i09.7123 · Cited by 774
Díaz-Rodríguez, N., Del Ser, J., Coeckelbergh, M., López de Prado, M., Herrera-Viedma, E., & Herrera, F. (2023) · Information Fusion · doi:10.1016/j.inffus.2023.101896 · Cited by 800
Multitask Prompted Training Enables Zero-Shot Task Generalization
Sanh, V., Webson, A., Raffel, C., Bach, S. H., Sutawika, L., Alyafeai, Z., Chaffin, A., Stiegler, A., Scao, T. L., Raja, A., et al. (2021) · arXiv · doi:10.48550/arxiv.2110.08207 · Cited by 612
Green Algorithms: Quantifying the Carbon Footprint of Computation
Lannelongue, L., Grealey, J., & Inouye, M. (2021) · Advanced Science · doi:10.1002/advs.202100707 · Cited by 712
Cowls, J., Tsamados, A., Taddeo, M., & Floridi, L. (2023) · AI & SOCIETY · doi:10.1007/s00146-021-01294-x · Cited by 552
“However, depending on where and how energy is sourced, stored and delivered, the rise of compute-intensive AI research can have significant, negative environmental effects (Lacoste et al. 2019).”
BackgroundMethodResultInfluentialCowls, J., Tsamados, A., Taddeo, M., & Floridi, L. (2021) · doi:10.2139/ssrn.3804983 · Cited by 552
The Ethics of Artificial Intelligence
Floridi, L. (2023) · doi:10.1093/oso/9780198883098.001.0001 · Cited by 682
Carbon Emissions and Large Neural Network Training
Patterson, D., Gonzalez, J., Le, Q., Liang, C., Munguia, L.-M., Rothchild, D., So, D., Texier, M., & Dean, J. (2021) · arXiv · doi:10.48550/arxiv.2104.10350 · Cited by 486
“Without access to such a dashboard, use the ML Emissions Calculator [Lac19] or Green Algorithms tool [Lan20] that estimate the CO 2 e mix by region (see Figure 6 below) 11 .”
BackgroundMethodInfluentialChemBERTa: Large-Scale Self-Supervised Pretraining for Molecular Property PredictionPreprint
Chithrananda, S., Grand, G., & Ramsundar, B. (2020) · arXiv (Cornell University) · doi:10.48550/arxiv.2010.09885 · Cited by 1,181
“Nevertheless, machine learning—and particularly largescale pretraining of the form we undertake here—is a resource-intensive process that has a growing carbon footprint [36].”
BackgroundThe Carbon Footprint of Machine Learning Training Will Plateau, Then Shrink
Patterson, D., Gonzalez, J., Holzle, U., Le, Q., Liang, C., Munguia, L.-M., Rothchild, D., So, D. R., Texier, M., & Dean, J. (2022) · Computer · doi:10.1109/mc.2022.3148714 · Cited by 400
CODA-Prompt: COntinual Decomposed Attention-Based Prompting for Rehearsal-Free Continual Learning
Smith, J. S., Karlinsky, L., Gutta, V., Cascante-Bonilla, P., Kim, D., Arbelle, A., Panda, R., Feris, R., & Kira, Z. (2023) · IEEE Computer Society Conference on Computer Vision and Pattern Recognition/Proceedings - IEEE Computer Society Conference on Computer Vision and Pattern Recognition/Proceedings · doi:10.1109/cvpr52729.2023.01146 · Cited by 409
“This could incur high financial [26] and environmental [33] costs if the replacement is frequent.”
BackgroundSatMAE: Pre-Training Transformers for Temporal and Multi-Spectral Satellite Imagery
Cong, Y., Khanna, S., Meng, C., Liu, P., Rozi, E., He, Y., Burke, M., Lobell, D., & Ermon, S. (2022) · Advances in Neural Information Processing Systems 35 · doi:10.52202/068431-0015 · Cited by 303
polyBERT: a chemical language model to enable fully machine-driven ultrafast polymer informatics
Kuenneth, C., & Ramprasad, R. (2023) · Nature Communications · doi:10.1038/s41467-023-39868-6 · Cited by 326
The Stable Signature: Rooting Watermarks in Latent Diffusion Models
Fernandez, P., Couairon, G., Jégou, H., Douze, M., & Furon, T. (2023) · 2023 IEEE/CVF International Conference on Computer Vision (ICCV) · doi:10.1109/iccv51070.2023.02053 · Cited by 251
“[43].”
Embedding Propagation: Smoother Manifold for Few-Shot Classification
Rodríguez, P., Laradji, I., Drouin, A., & Lacoste, A. (2020) · Lecture notes in computer science · doi:10.1007/978-3-030-58574-7_8 · Cited by 227
Self-citationMeasuring the Carbon Intensity of AI in Cloud Instances
Dodge, J., Prewitt, T., Tachet des Combes, R., Odmark, E., Schwartz, R., Strubell, E., Luccioni, A. S., Smith, N. A., DeCario, N., & Buchanan, W. (2022) · 2022 ACM Conference on Fairness Accountability and Transparency · doi:10.1145/3531146.3533234 · Cited by 257
“This is a more straightforward endeavor compared to other stages both upstream and downstream from the training process, given that it is well-defined in time and its emissions can be measured in real-time with tools like Code Carbon [38] and Carbon Tracker [1] or estimated post-hoc using tools such as ML CO2 Impact T…”
BackgroundMethodSelf-citationTrustworthy ai: A computational perspective
Liu, H., Wang, Y., Fan, W., Liu, X., Li, Y., Jain, S., Liu, Y., Jain, A., & Tang, J. (2023) · ACM Transactions on Intelligent Systems and Technology · doi:10.1145/3546872 · Cited by 261
“MLEC [214] A tool for estimating the carbon emissions of training a ML model.”
MethodEstimating the Carbon Footprint of BLOOM, a 176B Parameter Language Model
Luccioni, A. S., Viguier, S., & Ligozat, A.-L. (2022) · arXiv · doi:10.48550/arxiv.2211.02001 · Cited by 378
“In order to attempt to estimate the carbon emissions incurred by deploying BLOOM, we ran the CodeCarbon tool [18] on a Google Cloud Platform (GCP) instance with 16 Nvidia A100 40GB GPUs, where BLOOM was deployed via an inference API, and tracked the energy usage of the instance over a period of approximately 18 days.”
MethodSelf-citationTackling climate change with machine learning
Rolnick, D., Donti, P. L., Kaack, L. H., Kochanski, K., Lacoste, A., Sankaran, K., Ross, A. S., Milojevic-Dupont, N., Jaques, N., Waldman-Brown, A., et al. (2023) · ACM Computing Surveys · doi:10.1145/3485128 · Cited by 336
Self-citationLim, T. (2024) · Artificial Intelligence Review · doi:10.1007/s10462-024-10708-3 · Cited by 180
Towards the Systematic Reporting of the Energy and Carbon Footprints of Machine Learning
Henderson, P., Hu, J., Romoff, J., Brunskill, E., Jurafsky, D., & Pineau, J. (2020) · arXiv · doi:10.48550/arxiv.2002.05651 · Cited by 393
“Lacoste et al. (2019) recently provided a website for estimating carbon emissions based on GPU type, experiment length, and cloud provider.”
BackgroundArtificial intelligence for carbon emissions using system of systems theory
Gaur, L., Afaq, A., Arora, G. K., & Khan, N. (2023) · Ecological Informatics · doi:10.1016/j.ecoinf.2023.102165 · Cited by 187
Delanoë, P., Tchuente, D., & Colin, G. (2023) · Journal of Environmental Management · doi:10.1016/j.jenvman.2023.117261 · Cited by 175
TABBIE: Pretrained Representations of Tabular Data
Iida, H., Thai, D., Manjunatha, V., & Iyyer, M. (2021) · Conference of the North American Chapter of the Association for Computational Linguistics: Human Language Technologies (NAACL-HLT) · doi:10.18653/v1/2021.naacl-main.270 · Cited by 151
The Rise of Artificial Intelligence in Educational Measurement: Opportunities and Ethical Challenges
Bulut, O., & Beiting-Parrish, M. (2024) · Chinese/English Journal of Educational Measurement and Evaluation · doi:10.59863/miql7785 · Cited by 272
“This focus on the emissions from AI has prompted Lacoste et al. (2019) to develop a tool quantifying emissions by cloud computing provider, hardware, and region.”
Ueda, D., Walston, S. L., Fujita, S., Fushimi, Y., Tsuboyama, T., Kamagata, K., Yamada, A., Yanagawa, M., Ito, R., Fujima, N., et al. (2024) · Diagnostic and Interventional Imaging · doi:10.1016/j.diii.2024.06.002 · Cited by 146
Al-Quraan, M. M., Mohjazi, L., Bariah, L., Centeno, A., Zoha, A., Arshad, K., Assaleh, K. T., Muhaidat, S., Debbah, M. A., & Imran, M. A. (2023) · IEEE Transactions on Emerging Topics in Computational Intelligence · doi:10.1109/tetci.2023.3251404 · Cited by 159
Sponge Examples: Energy-Latency Attacks on Neural Networks
Shumailov, I., Zhao, Y., Bates, D., Papernot, N., Mullins, R., & Anderson, R. (2021) · 2021 IEEE European Symposium on Security and Privacy (EuroS&P) · doi:10.1109/eurosp51992.2021.00024 · Cited by 125
Carbontracker: Tracking and Predicting the Carbon Footprint of Training Deep Learning Models
Anthony, L. F. W., Kanding, B., & Selvan, R. (2020) · arXiv · doi:10.48550/arxiv.2007.03051 · Cited by 255
“Lacoste et al. (2019) provided the Machine Learning Emissions Calculator that relies on self-reporting.”
BackgroundTowards Geospatial Foundation Models via Continual Pretraining
Mendieta, M., Han, B., Shi, X., Zhu, Y., & Chen, C. (2023) · IEEE/CVF International Conference on Computer Vision (ICCV) · doi:10.1109/iccv51070.2023.01541 · Cited by 157
“We also report the training time in hours on a V100 GPU, as well as the carbon impact estimations(2) in kg CO2 equivalent [24].”
MethodFMix: Enhancing Mixed Sample Data AugmentationPreprint
Harris, E., Marcu, A., Painter, M., Niranjan, M., Prügel‐Bennett, A., & Hare, J. S. (2020) · arXiv (Cornell University) · doi:10.48550/arxiv.2002.12047 · Cited by 185
Efficient Methods for Natural Language Processing: A Survey
Treviso, M., Lee, J.-U., Ji, T., van Aken, B., Cao, Q., Ciosici, M. R., Hassid, M., Heafield, K., Hooker, S., Raffel, C., et al. (2023) · Transactions of the Association for Computational Linguistics · doi:10.1162/tacl_a_00577 · Cited by 156
“Another way is to use software tools such as MLCO2 (Luccioni et al., 2019).”
MethodEstimating the environmental impact of Generative-AI services using an LCA-based methodology
Berthelot, A., Caron, E., Jay, M., & Lefèvre, L. (2024) · Procedia CIRP · doi:10.1016/j.procir.2024.01.098 · Cited by 103
“Methodologies on how to measure or estimate it have been explored in several approaches [9, 14, 15].”
RETRACTED: Estimation of transport CO2 emissions using machine learning algorithm
Li, S., Tong, Z., & Haroon, M. S. (2024) · Transportation Research Part D Transport and Environment · doi:10.1016/j.trd.2024.104276 · Cited by 100
RetractedEvaluating the carbon footprint of NLP methods: a survey and analysis of existing tools
Bannour, N., Ghannay, S., Névéol, A., & Ligozat, A.-L. (2021) · Proceedings of the Second Workshop on Simple and Efficient Natural Language Processing · doi:10.18653/v1/2021.sustainlp-1.2 · Cited by 98
“The CO 2 equivalent emissions obtained from ML CO2 Impact and Green Algorithms are higher than those returned by other tools.”
InfluentialArtificial intelligence and climate change: ethical issues
Nordgren, A. (2023) · Journal of Information, Communication and Ethics in Society · doi:10.1108/jices-11-2021-0106 · Cited by 94
“For example, when using cloud computing, data centres could be moved to countries or regions providing a more fossil-free energymix (Lacoste et al., 2019).”
BackgroundMethodRobust deep learning-based semantic organ segmentation in hyperspectral images
Seidlitz, S., Sellner, J., Odenthal, J., Özdemir, B., Studier‐Fischer, A., Knödler, S., Ayala, L. A., Adler, T. J., Kenngott, H. G., Tizabi, M. D., et al. (2022) · Medical Image Analysis · doi:10.1016/j.media.2022.102488 · Cited by 98
An experimental comparison of software-based power meters: focus on CPU and GPU
Jay, M., Ostapenco, V., Lefèvre, L., Trystram, D., Orgerie, A., & Fichel, B. (2023) · IEEE/ACM International Symposium on Cluster, Cloud and Grid Computing (CCGrid) · doi:10.1109/ccgrid57682.2023.00020 · Cited by 92
“In this paper, we propose a comparison of nine software-based power meters: Carbon Tracker [37], Code Carbon [38], Energy Scope [42], Experiment Impact Tracker [39], Perf [46], Power API [40], Green Algorithms [36], ML CO2 Impact [35], and Scaphandre [41].”
InfluentialA Survey on Variational Autoencoders from a Green AI Perspective
Asperti, A., Evangelista, D., & Piccolomini, E. L. (2021) · SN Computer Science · doi:10.1007/s42979-021-00702-9 · Cited by 100
“…AI research (referred to as Red AI, in this con) targets accuracy rather than efficiency, exploiting massive computational power, and resulting in rapidly escalating costs; this trend is not sustainable for various reasons, it is environmentally unfriendly [42], socially not inclusive and inefficient.”
BackgroundData-Centric Green AI An Exploratory Empirical Study
Verdecchia, R., Cruz, L., Sallou, J., Lin, M., Wickenden, J., & Hotellier, E. (2022) · International Conference on ICT for Sustainability (ICT4S) · doi:10.1109/ict4s55073.2022.00015 · Cited by 108
“Previous studies observed that AI-related tasks are particularly energy-greedy [2], [3].”
BackgroundData augmentation instead of explicit regularizationPreprint
Hernández-García, Á., & Peter König, B. (2018) · arXiv (Cornell University) · doi:10.48550/arxiv.1806.03852 · Cited by 176
LDMVFI: Video Frame Interpolation with Latent Diffusion Models
Danier, D., Zhang, F., & Bull, D. (2024) · Proceedings of the AAAI Conference on Artificial Intelligence · doi:10.1609/aaai.v38i2.27912 · Cited by 108
“Secondly, the two-stage training strategy, large model size and slow inference speed of LDMVFI mean that large-scale training and evaluation processes can consume significant amounts of energy, leading to increased carbon footprint (Lacoste et al. 2019).”
BackgroundArtificial intelligence and the climate emergency: Opportunities, challenges, and recommendations
Taddeo, M., Tsamados, A., Cowls, J., & Floridi, L. (2021) · One Earth · doi:10.1016/j.oneear.2021.05.018 · Cited by 100
Toward Green AI: A Methodological Survey of the Scientific Literature
Barbierato, E., & Gatti, A. (2024) · IEEE Access · doi:10.1109/access.2024.3360705 · Cited by 100
“Lacoste et al. [90] use factors such as hardware energy consumption, computation location, emissions associated with the region, and any offsets purchased by the provider to mitigate environmental impact.”
BackgroundBevers, N., Sikora, E. J., & Hardy, N. B. (2022) · Computers and Electronics in Agriculture · doi:10.1016/j.compag.2022.107449 · Cited by 84
Data-Centric Green Artificial Intelligence: A Survey
Salehi, S., & Schmeink, A. (2023) · IEEE Transactions on Artificial Intelligence · doi:10.1109/tai.2023.3315272 · Cited by 90
“Lacoste et al. [18] also identified the key factors affecting the carbon emissions associated with neural network training: the server’s location and the energy grid it relies on, the duration of the training process, and even the specific hardware used for training.”
BackgroundMethodNewsCLIPpings: Automatic Generation of Out-of-Context Multimodal Media
Luo, G., Darrell, T., & Rohrbach, A. (2021) · Conference on Empirical Methods in Natural Language Processing (EMNLP) · doi:10.18653/v1/2021.emnlp-main.545 · Cited by 95
“Estima-tions were conducted using the MachineLearning Impact calculator presented in Lacoste et al. (2019).”
MethodLiao, M., Lan, K., & Yao, Y. (2021) · Journal of Industrial Ecology · doi:10.1111/jiec.13214 · Cited by 84
“For example, previous studies showed that computational time has direct correlationswith the energy and carbon footprints of AI andmachine learning, and thus training time should always be reported (Dhar, 2020; Henderson et al., 2020; Lacoste et al., 2019; Strubell et al., 2020).”
MethodThe AI Carbon Footprint and Responsibilities of AI Scientists
Tamburrini, G. (2022) · Philosophies · doi:10.3390/philosophies7010004 · Cited by 83
Ebrahimi, A., Mager, M., Oncevay, A., Chaudhary, V., Chiruzzo, L., Fan, A., Ortega, J. E., Ramos, R., Rios, A., Meza, I., et al. (2022) · Proceedings of the conference - Association for Computational Linguistics. Meeting · doi:10.18653/v1/2022.acl-long.435 · Cited by 76
Sirimewan, D., Bazli, M., Raman, S. N., Mohandes, S. R., Kineber, A. F., & Arashpour, M. (2024) · Journal of Environmental Management · doi:10.1016/j.jenvman.2023.119908 · Cited by 70
Artificial Intelligence in the NHS: Climate and Emissions✰,✰✰
Bloomfield, P. S., Clutton-Brock, P., Pencheon, E., Magnusson, J., & Karpathakis, K. (2021) · The Journal of Climate Change and Health · doi:10.1016/j.joclim.2021.100056 · Cited by 80
Pagallo, U., Ciani, J., & Durante, M. (2022) · Transforming Government People Process and Policy · doi:10.1108/tg-07-2021-0121 · Cited by 82
A first look into the carbon footprint of federated learningPreprint
Qiu, X., Parcollet, T., Fernandez-Marques, J., de Gusmao, P. P. B., Gao, Y., Beutel, D. J., Topal, T., Mathur, A., & Lane, N. D. (2021) · arXiv (Cornell University) · doi:10.48550/arxiv.2102.07627 · Cited by 88
Luccioni, A. S., Strubell, E., & Crawford, K. (2025) · Proceedings of the 2025 ACM Conference on Fairness, Accountability, and Transparency · doi:10.1145/3715275.3732007 · Cited by 71
“Papers have addressed carbon emissions from training large models [67, 69, 70, 120], water consumption for cooling servers [52, 80], and the mining of minerals for technical infrastructure [24].”
Self-citationA cognitive approach to human–AI complementarity in dynamic decision-making
González, C., & Heidari, H. (2025) · Nature Reviews Psychology · doi:10.1038/s44159-025-00499-x · Cited by 73
A Survey on Model Compression and Acceleration for Pretrained Language Models
Xu, C., & McAuley, J. (2023) · Proceedings of the AAAI Conference on Artificial Intelligence · doi:10.1609/aaai.v37i9.26255 · Cited by 85
A Closer Look at Rehearsal-Free Continual Learning *
Smith, J. S., Tian, J., Halbe, S., Hsu, Y.-C., & Kira, Z. (2023) · IEEE/CVF Conference on Computer Vision and Pattern Recognition Workshops (CVPRW) · doi:10.1109/cvprw59228.2023.00239 · Cited by 60
“This may lead to high financial [24] and environmental [32] costs after numerous replacements.”
BackgroundLearning Rich Representation of Keyphrases from Text
Kulkarni, M., Mahata, D., Arora, R., & Bhowmik, R. (2022) · Findings of the Association for Computational Linguistics: NAACL · doi:10.18653/v1/2022.findings-naacl.67 · Cited by 80
Tracking artificial intelligence in climate inventions with patent data
Verendel, V. (2023) · Nature Climate Change · doi:10.1038/s41558-022-01536-w · Cited by 61
A Survey on Green Deep LearningPreprint
Xu, J., Zhou, W., Fu, Z., Zhou, H., & Li, L. (2021) · arXiv (Cornell University) · doi:10.48550/arxiv.2111.05193 · Cited by 118
Revisit the environmental impact of artificial intelligence: the overlooked carbon emission source?
Yang, Y., Wang, J., Liu, Y., Yu, P., Wang, D., Zheng, P., & Zhang, M. (2024) · Frontiers of Environmental Science & Engineering · doi:10.1007/s11783-024-1918-y · Cited by 83
“Electricity consumption of training models was evaluated according to Lacoste et al. (2019).”
Towards Green Automated Machine Learning: Status Quo and Future Directions
Tornede, T., Tornede, A., Hanselle, J., Mohr, F., Wever, M., & Hüllermeier, E. (2023) · Journal of Artificial Intelligence Research · doi:10.1613/jair.1.14340 · Cited by 69
“Similarly, the Machine Learning Emissions Calculator(1) [37] is an easy-to-use online tool for estimating the CO2e footprint of a set of experiments.”
MethodExploring the Carbon Footprint of Hugging Face's ML Models: A Repository Mining Study
Castaño, J., Martínez‐Fernández, S., Franch, X., & Bogner, J. (2023) · Proceedings of the ... ACM-IEEE International Symposium on Empirical Software Engineering and Measurement/International Symposium on Empirical Software Engineering and Measurement · doi:10.1109/esem56168.2023.10304801 · Cited by 75
“’s ML workload calculator [19] can estimate the carbon footprint.”
BackgroundTöRF: Time-of-Flight Radiance Fields for Dynamic Scene View SynthesisPreprint
Attal, B., Laidlaw, E., Gokaslan, A., Kim, C., Richardt, C., Tompkin, J., & O’Toole, M. (2021) · arXiv (Cornell University) · doi:10.48550/arxiv.2109.15271 · Cited by 56
“02 kg CO2-equivalents per sequence for rendering (numbers generated by ML CO2 Impact [22]).”
BackgroundToward Sustainable Serverless Computing
Patros, P., Spillner, J., Papadopoulos, A. V., Varghese, B., RANA, O. F., & Dustdar, S. (2021) · IEEE Internet Computing · doi:10.1109/mic.2021.3093105 · Cited by 53
Reporting electricity consumption is essential for sustainable AI
Debus, C., Piraud, M., Streit, A., Theis, F. J., & Götz, M. (2023) · Nature Machine Intelligence · doi:10.1038/s42256-023-00750-1 · Cited by 60
ST-MFNet: A Spatio-Temporal Multi-Flow Network for Frame Interpolation
Danier, D., Zhang, F., & Bull, D. (2022) · IEEE Computer Society Conference on Computer Vision and Pattern Recognition/Proceedings - IEEE Computer Society Conference on Computer Vision and Pattern Recognition/Proceedings · doi:10.1109/cvpr52688.2022.00351 · Cited by 51
AVATAR: A Parallel Corpus for Java-Python Program Translation
Ahmad, W. U., Tushar, M. G. R., Chakraborty, S., & Chang, K.-W. (2023) · Findings of the Association for Computational Linguistics: ACL 2023 · doi:10.18653/v1/2023.findings-acl.143 · Cited by 46
The Foundation Model Transparency Index
Bommasani, R., Klyman, K., Longpre, S., Kapoor, S., Maslej, N., Xiong, B., Zhang, D., & Liang, P. (2023) · arXiv · doi:10.48550/arxiv.2310.12941 · Cited by 46
PHNNs: Lightweight Neural Networks via Parameterized Hypercomplex Convolutions
Grassucci, E., Zhang, A., & Comminiello, D. (2022) · IEEE Transactions on Neural Networks and Learning Systems · doi:10.1109/tnnls.2022.3226772 · Cited by 38
“Estimations were conducted using the MachineLearning Impact calculator presented in [57].”
MethodReguero, Á. D., Martínez‐Fernández, S., & Verdecchia, R. (2024) · Computer Standards & Interfaces · doi:10.1016/j.csi.2024.103906 · Cited by 45
Fusion of Complementary 2D and 3D Mesostructural Datasets Using Generative Adversarial Networks
Dahari, A., Kench, S., Squires, I., & Cooper, S. J. (2022) · Advanced Energy Materials · doi:10.1002/aenm.202202407 · Cited by 37
The MultiBERTs: BERT Reproductions for Robustness Analysis
Sellam, T., Yadlowsky, S., Wei, J., Saphra, N., D'Amour, A., Linzen, T., Bastings, J., Turc, I., Eisenstein, J., Das, D., et al. (2021) · arXiv · doi:10.48550/arxiv.2106.16163 · Cited by 37
“Using the calculations of Lacoste et al. (2019)2, we estimate this as about 250 kg CO2e for each of our 25 models.”
MethodMeasuring the environmental impacts of artificial intelligence compute and applications
OECD (2022) · OECD digital economy papers · doi:10.1787/7babf571-en · Cited by 96
The Cost of Training Machine Learning Models Over Distributed Data Sources
Guerra, E., Wilhelmi, F., Miozzo, M., & Dini, P. (2023) · IEEE Open Journal of the Communications Society · doi:10.1109/ojcoms.2023.3274394 · Cited by 44
“To achieve this goal, they use the machine learning emission calculator [36] and green algorithms [37].”
MethodA Diffusion Model Predicts 3D Shapes from 2D Microscopy Images
Waibel, D., Röell, E., Rieck, B., Giryes, R., & Marr, C. (2023) · IEEE International Symposium on Biomedical Imaging (ISBI) · doi:10.1109/isbi53787.2023.10230752 · Cited by 42
“Using the ML CO2 calculator [20] we estimate 81.”
MethodCCVS: Context-aware Controllable Video SynthesisPreprint
Moing, G. L., Ponce, J., & Schmid, C. (2021) · arXiv (Cornell University) · doi:10.48550/arxiv.2107.08037 · Cited by 34
“Methods sparing GPU-hours like ours are crucial to make AI less polluting [43, 61], and move from a “Red” to a “Green” AI [58].”
BackgroundBieńkowska, J. (2024) · Sustainable Development · doi:10.1002/sd.3312 · Cited by 32
The future of artificial intelligence in the context of industrial ecology
Donati, F., Dente, S. M. R., Li, C., Vilaysouk, X., Froemelt, A., Nishant, R., Liu, G., Tukker, A., & Hashimoto, S. (2022) · Journal of Industrial Ecology · doi:10.1111/jiec.13313 · Cited by 34
“For example, AImodels have known issues of high energy consumption, which are also projected to grow beyond 2%of theworld energy consumption (Lacoste et al., 2019; Lottick et al., 2019; Strubell et al., 2019).”
Fine allocation of sectoral carbon emissions at block scale and contribution of functional zones
Ma, X., Yang, J., Sun, D., Zhang, R., Xiao, X., & Xia, J. C. (2023) · Ecological Informatics · doi:10.1016/j.ecoinf.2023.102293 · Cited by 31
Large-Scale, Language-Agnostic Discourse Classification of Tweets During COVID-19
Gencoglu, O. (2020) · Machine Learning and Knowledge Extraction · doi:10.3390/make2040032 · Cited by 31
Localizing in-domain adaptation of transformer-based biomedical language models
Buonocore, T. M., Crema, C., Redolfi, A., Bellazzi, R., & Parimbelli, E. (2023) · Journal of Biomedical Informatics · doi:10.1016/j.jbi.2023.104431 · Cited by 37
“[21] results in a total of approximately 273 kgCO2eq produced, which is equivalent to 1100 km driven by an average internal combustion engine car.”
BackgroundTopicBERT for Energy Efficient Document Classification
Chaudhary, Y., Gupta, P., Saxena, K., Kulkarni, V., Runkler, T. A., & Schütze, H. (2020) · Findings of the Association for Computational Linguistics: EMNLP (Findings of EMNLP) · doi:10.18653/v1/2020.findings-emnlp.152 · Cited by 34
How to Adapt Your Pretrained Multilingual Model to 1600 Languages
Ebrahimi, A., & Kann, K. (2021) · Annual Meeting of the Association for Computational Linguistics · doi:10.18653/v1/2021.acl-long.351 · Cited by 46
“Experiments for NER and POS take around 1 and 2 hours respectively, totalling to 165 total training hours, and 21.38 kgCO2eq emitted (Lacoste et al., 2019).”
BackgroundInfluentialOptimizing Reusable Knowledge for Continual Learning via MetalearningPreprint
Hurtado, J., Raymond-Sáez, A., & Soto, Á. (2021) · arXiv (Cornell University) · doi:10.48550/arxiv.2106.05390 · Cited by 29
RETRIEVE: Coreset Selection for Efficient and Robust Semi-Supervised LearningPreprint
Killamsetty, K., Zhao, X., Chen, F., & Iyer, R. (2021) · arXiv (Cornell University) · doi:10.48550/arxiv.2106.07760 · Cited by 34
“CO2 emissions were estimated based on the total compute time using the Machine Learning Impact calculator presented in [29].”
MethodArtificial Intelligence in Ecology: A Commentary on a Chatbot's Perspective
Haghighi, S. R., Saqalaksari, M. P., & Johnson, S. N. (2023) · Bulletin of the Ecological Society of America · doi:10.1002/bes2.2097 · Cited by 42
Cano-Ortiz, S., Iglesias, L. L., del Árbol, P. M. R., & Castro‐Fresno, D. (2023) · Developments in the Built Environment · doi:10.1016/j.dibe.2023.100315 · Cited by 29
The implications of handwritten text recognition for accessing the past at scale
Nockels, J., Gooding, P., & Terras, M. (2024) · Journal of Documentation · doi:10.1108/jd-09-2023-0183 · Cited by 41
The “Collections as ML Data” checklist for machine learning and cultural heritage
Lee, B. C. G. (2023) · Journal of the Association for Information Science and Technology · doi:10.1002/asi.24765 · Cited by 55
“…(2014, 2016); Friedman and Hendry (2012); Gebru et al. (2020); Gursoy and Kakadiaris (2022); Holland et al. (2018); Katell et al. (2019); Lacoste et al. (2019); Madaio et al. (2020); Mitchell et al. (2019); Pushkarna et al. (2022); Reisman et al. (2018); Ribeiro et al. (2020); Seifert…”
Single-Node Power Demand During AI Training: Measurements on an 8-GPU NVIDIA H100 System
Latif, I., Newkirk, A. C., Carbone, M. R., Munir, A., Lin, Y., Koomey, J., Yu, X., & Dong, Z. (2025) · IEEE Access · doi:10.1109/access.2025.3554728 · Cited by 50
“Academic researchers instead have to rely on what data they have available, with one common approach to multiply the manufacturer rated power of AI-optimized GPUs or servers [12] by the reported training time (typically in GPU-hours) of a given model [13], [14], [15].”
A Codesigned Integrated Photonic Electronic Neuron
Marinis, L. D., Catania, A., Castoldi, P., Contestabile, G., Bruschi, P., Piotto, M., & Andriolli, N. (2022) · IEEE Journal of Quantum Electronics · doi:10.1109/jqe.2022.3177793 · Cited by 27
Climate Change and Health: The next Challenge of Ethical AI
Fiske, A., Radhuber, I. M., Willem, T., Buyx, A., Celi, L. A., & McLennan, S. (2025) · The Lancet Global Health · doi:10.1016/s2214-109x(25)00124-x · Cited by 25
Metadata
- Authors
- Lacoste, A., Luccioni, A., Schmidt, V., & Dandres, T.
- Affiliations
- Polytechnique Montréal (CA)
- Venue
- arXiv
- Publisher
- arXiv
- Published
- 2019
- ISSN
- 2331-8422 (ISSN-L)
- arXiv
- 1910.09700
- OpenAlex
- W2981540061
- Cited by
- 408
- References
- 18
- Citations by year
- 2026: 13 · 2025: 55 · 2024: 71 · 2023: 139 · 2022: 48 · 2021: 64 · 2020: 16
- Type
- posted content
- Language
- en
- Open access
- green
- arxiv.org
- Locations
- arXiv (Cornell University) (submittedVersion)
- arXiv (Cornell University)
- Topics
- Physical Sciences › Engineering › Electrical and Electronic Engineering › Green IT and Sustainability
- Physical Sciences › Computer Science › Hardware and Architecture › Energy Efficiency in Computing
- Physical Sciences › Environmental Science › Environmental Engineering › Environmental Impact and Sustainability
- Keywords
- CO2 emissions, machine learning training, environmental impact of AI, neural network training, power grid, training duration, carbon mitigation
About these checks
Registry, retraction, Cites and statistics results are computed, not judged, with one exception: a printed reference that no index lists is read once by a model before it counts as a citation error. Claims are extracted by a model and kept only when their quote appears verbatim in the text; each cited source is then checked, with a model’s help, for whether it says what the claim relies on. Those verdicts can be wrong — read the claim’s quote and the cited source before relying on one.
Statistics and arithmetic are recomputed, not judged.
Confidence bands are shown only for fields calibrated on our annotated test set. None is calibrated yet, so no band is shown.
What counts as a claim ›How we measure confidence ›How we check this ›
Are you an author, or do you see a problem with this page? Request a review
