{"status":"ok","message-type":"work","message-version":"1.0.0","message":{"indexed":{"date-parts":[[2026,5,23]],"date-time":"2026-05-23T17:05:43Z","timestamp":1779555943303,"version":"3.53.1"},"reference-count":71,"publisher":"Elsevier BV","license":[{"start":{"date-parts":[[2027,1,1]],"date-time":"2027-01-01T00:00:00Z","timestamp":1798761600000},"content-version":"tdm","delay-in-days":0,"URL":"https:\/\/www.elsevier.com\/tdm\/userlicense\/1.0\/"},{"start":{"date-parts":[[2027,1,1]],"date-time":"2027-01-01T00:00:00Z","timestamp":1798761600000},"content-version":"tdm","delay-in-days":0,"URL":"https:\/\/www.elsevier.com\/legal\/tdmrep-license"},{"start":{"date-parts":[[2026,5,22]],"date-time":"2026-05-22T00:00:00Z","timestamp":1779408000000},"content-version":"vor","delay-in-days":0,"URL":"http:\/\/creativecommons.org\/licenses\/by-nc\/4.0\/"}],"funder":[{"DOI":"10.13039\/501100003030","name":"Government of Catalonia Agency for Administration of University and Research Grants","doi-asserted-by":"publisher","id":[{"id":"10.13039\/501100003030","id-type":"DOI","asserted-by":"publisher"}]},{"DOI":"10.13039\/501100011033","name":"Agencia Estatal de Investigaci\u00f3n","doi-asserted-by":"publisher","id":[{"id":"10.13039\/501100011033","id-type":"DOI","asserted-by":"publisher"}]}],"content-domain":{"domain":["elsevier.com","sciencedirect.com"],"crossmark-restriction":true},"short-container-title":["Computer Standards & Interfaces"],"published-print":{"date-parts":[[2027,1]]},"DOI":"10.1016\/j.csi.2026.104170","type":"journal-article","created":{"date-parts":[[2026,5,21]],"date-time":"2026-05-21T16:21:44Z","timestamp":1779380504000},"page":"104170","update-policy":"https:\/\/doi.org\/10.1016\/elsevier_cm_policy","source":"Crossref","is-referenced-by-count":0,"special_numbering":"C","title":["Estimating Deep Learning energy consumption based on model architecture and training environment"],"prefix":"10.1016","volume":"99","author":[{"ORCID":"https:\/\/orcid.org\/0000-0003-4979-414X","authenticated-orcid":false,"given":"Santiago","family":"del Rey","sequence":"first","affiliation":[],"role":[{"vocabulary":"crossref","role":"author"}]},{"ORCID":"https:\/\/orcid.org\/0000-0002-1615-355X","authenticated-orcid":false,"given":"Lu\u00eds","family":"Cruz","sequence":"additional","affiliation":[],"role":[{"vocabulary":"crossref","role":"author"}]},{"ORCID":"https:\/\/orcid.org\/0000-0001-9733-8830","authenticated-orcid":false,"given":"Xavier","family":"Franch","sequence":"additional","affiliation":[],"role":[{"vocabulary":"crossref","role":"author"}]},{"ORCID":"https:\/\/orcid.org\/0000-0001-9928-133X","authenticated-orcid":false,"given":"Silverio","family":"Mart\u00ednez-Fern\u00e1ndez","sequence":"additional","affiliation":[],"role":[{"vocabulary":"crossref","role":"author"}]}],"member":"78","reference":[{"key":"10.1016\/j.csi.2026.104170_b1","series-title":"The rapid adoption of generative AI","author":"Bick","year":"2024"},{"key":"10.1016\/j.csi.2026.104170_b2","series-title":"2024 50th Euromicro Conference on Software Engineering and Advanced Applications","first-page":"106","article-title":"The state of generative AI adoption from software practitioners\u2019 perspective: An empirical study","author":"Simaremare","year":"2024"},{"issue":"3","key":"10.1016\/j.csi.2026.104170_b3","doi-asserted-by":"crossref","DOI":"10.3390\/app15031344","article-title":"AI-driven innovations in software engineering: A review of current practices and future directions","volume":"15","author":"Alenezi","year":"2025","journal-title":"Appl. Sci."},{"key":"10.1016\/j.csi.2026.104170_b4","series-title":"AI and compute","author":"Dario","year":"2018"},{"issue":"7","key":"10.1016\/j.csi.2026.104170_b5","doi-asserted-by":"crossref","first-page":"18","DOI":"10.1109\/MC.2022.3148714","article-title":"The carbon footprint of machine learning training will plateau, then shrink","volume":"55","author":"Patterson","year":"2022","journal-title":"Computer"},{"issue":"12","key":"10.1016\/j.csi.2026.104170_b6","doi-asserted-by":"crossref","first-page":"1","DOI":"10.1007\/s11783-024-1918-y","article-title":"Revisit the environmental impact of artificial intelligence: The overlooked carbon emission source?","volume":"18","author":"Yu","year":"2024","journal-title":"Front. Environ. Sci. Eng."},{"key":"10.1016\/j.csi.2026.104170_b7","doi-asserted-by":"crossref","first-page":"54","DOI":"10.1145\/3381831","article-title":"Green AI","volume":"63","author":"Schwartz","year":"2020","journal-title":"Commun. ACM"},{"issue":"3","key":"10.1016\/j.csi.2026.104170_b8","first-page":"64:1","article-title":"Green IN artificial intelligence from a software perspective: State-of-the-art and green decalogue","volume":"57","author":"Guti\u00e9rrez","year":"2024","journal-title":"ACM Comput. Surv."},{"issue":"4","key":"10.1016\/j.csi.2026.104170_b9","doi-asserted-by":"crossref","first-page":"1915","DOI":"10.1007\/s11831-023-10032-z","article-title":"Review of lightweight deep convolutional neural networks","volume":"31","author":"Chen","year":"2024","journal-title":"Arch. Comput. Methods Eng."},{"issue":"3","key":"10.1016\/j.csi.2026.104170_b10","doi-asserted-by":"crossref","first-page":"1905","DOI":"10.1007\/s10462-022-10213-5","article-title":"A review of convolutional neural network architectures and their optimizations","volume":"56","author":"Cong","year":"2023","journal-title":"Artif. Intell. Rev."},{"key":"10.1016\/j.csi.2026.104170_b11","series-title":"ISO\/IEC TR 20226:2025","author":"ISO","year":"2025"},{"key":"10.1016\/j.csi.2026.104170_b12","series-title":"Software carbon intensity (SCI) specification","author":"Green Software Foundation","year":"2025"},{"key":"10.1016\/j.csi.2026.104170_b13","series-title":"Comit\u00e9 CTN 71\/SC 42 inteligencia artificial y big data","author":"Spanish Association for Standardization","year":"2025"},{"key":"10.1016\/j.csi.2026.104170_b14","series-title":"Home - Resource-efficient Software","author":"DIN","year":"2025"},{"key":"10.1016\/j.csi.2026.104170_b15","series-title":"Carbon emissions and large neural network training","author":"Patterson","year":"2021"},{"issue":"253","key":"10.1016\/j.csi.2026.104170_b16","first-page":"1","article-title":"Estimating the carbon footprint of BLOOM, a 176B parameter language model","volume":"24","author":"Luccioni","year":"2023","journal-title":"J. Mach. Learn. Res."},{"key":"10.1016\/j.csi.2026.104170_b17","series-title":"Counting carbon: A survey of factors influencing the emissions of machine learning","author":"Luccioni","year":"2023"},{"key":"10.1016\/j.csi.2026.104170_b18","series-title":"2025 IEEE International Symposium on High Performance Computer Architecture","first-page":"1201","article-title":"MLPerf power: Benchmarking the energy efficiency of machine learning systems from \u03bcWatts to MWatts for sustainable AI","author":"Tschand","year":"2025"},{"issue":"4","key":"10.1016\/j.csi.2026.104170_b19","article-title":"A systematic review of Green AI","volume":"13","author":"Verdecchia","year":"2023","journal-title":"WIREs Data Min. Knowl. Discov."},{"key":"10.1016\/j.csi.2026.104170_b20","series-title":"IEEE Big Data","first-page":"3759","article-title":"Deep learning in the automotive industry: Applications and tools","author":"Luckow","year":"2016"},{"key":"10.1016\/j.csi.2026.104170_b21","series-title":"High Performance Computing. ISC High Performance 2022 International Workshops","first-page":"108","article-title":"Precise energy consumption measurements of heterogeneous artificial intelligence workloads","author":"Caspart","year":"2022"},{"key":"10.1016\/j.csi.2026.104170_b22","series-title":"BDCloud-SocialCom-SustainCom","first-page":"477","article-title":"Evaluating the energy efficiency of deep convolutional neural networks on CPUs and GPUs","author":"Li","year":"2016"},{"key":"10.1016\/j.csi.2026.104170_b23","series-title":"2023 IEEE\/ACM 2nd International Conference on AI Engineering \u2013 Software Engineering for AI","first-page":"25","article-title":"Uncovering energy-efficient practices in deep learning training: Preliminary steps towards green AI","author":"Yarally","year":"2023"},{"key":"10.1016\/j.csi.2026.104170_b24","series-title":"2019 32nd International Conference on VLSI Design and 2019 18th International Conference on Embedded Systems","first-page":"215","article-title":"The ramifications of making deep neural networks compact","author":"Jha","year":"2019","ISSN":"https:\/\/id.crossref.org\/issn\/2380-6923","issn-type":"print"},{"issue":"3","key":"10.1016\/j.csi.2026.104170_b25","doi-asserted-by":"crossref","DOI":"10.3390\/s25030846","article-title":"Towards an energy consumption index for deep learning models: A comparative analysis of architectures, GPUs, and measurement tools","volume":"25","author":"Aquino-Br\u00edtez","year":"2025","journal-title":"Sensors"},{"key":"10.1016\/j.csi.2026.104170_b26","series-title":"Computer Vision \u2013 ECCV 2024 Workshops","first-page":"388","article-title":"Watt for what: Rethinking deep learning\u2019s energy-performance relationship","author":"Gowda","year":"2025"},{"key":"10.1016\/j.csi.2026.104170_b27","doi-asserted-by":"crossref","DOI":"10.1016\/j.csi.2024.103906","article-title":"Energy-efficient neural network training through runtime layer freezing, model quantization, and early stopping","volume":"92","author":"Reguero","year":"2025","journal-title":"Comput. Stand. Interfaces"},{"key":"10.1016\/j.csi.2026.104170_b28","series-title":"Quantifying the carbon emissions of machine learning","author":"Lacoste","year":"2019"},{"issue":"12","key":"10.1016\/j.csi.2026.104170_b29","doi-asserted-by":"crossref","DOI":"10.1002\/advs.202100707","article-title":"Green algorithms: quantifying the carbon footprint of computation","volume":"8","author":"Lannelongue","year":"2021","journal-title":"Adv. Sci."},{"issue":"4","key":"10.1016\/j.csi.2026.104170_b30","article-title":"Empirically-calibrated H100 node power models for accurate AI training energy estimation","volume":"2","author":"Newkirk","year":"2025","journal-title":"Environ. Res.: Energy"},{"key":"10.1016\/j.csi.2026.104170_b31","series-title":"Measuring the energy consumption and efficiency of deep neural networks: An empirical analysis and design recommendations","author":"Tripp","year":"2024"},{"issue":"3","key":"10.1016\/j.csi.2026.104170_b32","doi-asserted-by":"crossref","first-page":"14","DOI":"10.1145\/3743095.3743099","article-title":"Greening AI-enabled systems with software engineering: A research agenda for environmentally sustainable AI practices","volume":"50","author":"Cruz","year":"2025","journal-title":"SIGSOFT Softw. Eng. Notes"},{"key":"10.1016\/j.csi.2026.104170_b33","series-title":"The ML.ENERGY benchmark: Toward automated inference energy measurement and optimization","author":"Chung","year":"2025"},{"issue":"11","key":"10.1016\/j.csi.2026.104170_b34","doi-asserted-by":"crossref","DOI":"10.1088\/2515-7620\/acf81b","article-title":"How to estimate carbon footprint when training deep learning models? A guide and review","volume":"5","author":"Bouza","year":"2023","journal-title":"Environ. Res. Commun."},{"key":"10.1016\/j.csi.2026.104170_b35","doi-asserted-by":"crossref","first-page":"402","DOI":"10.1016\/j.future.2024.01.033","article-title":"Development and evaluation of a reference measurement model for assessing the resource and energy efficiency of software products and components\u2014Green Software Measurement Model (GSMM)","volume":"155","author":"Guldner","year":"2024","journal-title":"Future Gener. Comput. Syst."},{"key":"10.1016\/j.csi.2026.104170_b36","series-title":"2022 18th International Conference on Intelligent Environments","first-page":"1","article-title":"PowerJoular and joularjx: Multi-platform software power monitoring tools","author":"Noureddine","year":"2022"},{"key":"10.1016\/j.csi.2026.104170_b37","series-title":"EnergiBridge: Empowering software sustainability through cross-platform energy measurement","author":"Sallou","year":"2023"},{"key":"10.1016\/j.csi.2026.104170_b38","series-title":"2009 IEEE Conference on Computer Vision and Pattern Recognition","first-page":"248","article-title":"ImageNet: A large-scale hierarchical image database","author":"Deng","year":"2009"},{"key":"10.1016\/j.csi.2026.104170_b39","series-title":"Learning Multiple Layers of Features from Tiny Images","author":"Krizhevsky","year":"2009"},{"key":"10.1016\/j.csi.2026.104170_b40","series-title":"2012 IEEE Conference on Computer Vision and Pattern Recognition","first-page":"3498","article-title":"Cats and dogs","author":"Parkhi","year":"2012"},{"key":"10.1016\/j.csi.2026.104170_b41","doi-asserted-by":"crossref","unstructured":"G. Van Horn, O. Mac Aodha, Y. Song, Y. Cui, C. Sun, A. Shepard, H. Adam, P. Perona, S. Belongie, The INaturalist Species Classification and Detection Dataset, in: Proceedings of the IEEE Conference on Computer Vision and Pattern Recognition, CVPR, 2018, pp. 8769\u20138778.","DOI":"10.1109\/CVPR.2018.00914"},{"key":"10.1016\/j.csi.2026.104170_b42","series-title":"Computer Vision \u2013 ECCV 2014","first-page":"446","article-title":"Food-101 \u2013 mining discriminative components with random forests","author":"Bossard","year":"2014"},{"key":"10.1016\/j.csi.2026.104170_b43","first-page":"528","article-title":"The goal question metric approach","author":"Basili","year":"1994","journal-title":"Encycl. Softw. Eng."},{"key":"10.1016\/j.csi.2026.104170_b44","series-title":"The AI index 2022 annual report","author":"Zhang","year":"2022"},{"key":"10.1016\/j.csi.2026.104170_b45","series-title":"2022 International Joint Conference on Neural Networks","first-page":"1","article-title":"Compute trends across three eras of machine learning","author":"Sevilla","year":"2022","ISSN":"https:\/\/id.crossref.org\/issn\/2161-4407","issn-type":"print"},{"key":"10.1016\/j.csi.2026.104170_b46","series-title":"2023 IEEE High Performance Extreme Computing Conference","first-page":"1","article-title":"An analysis of energy requirement for computer vision algorithms","author":"Edelman","year":"2023"},{"key":"10.1016\/j.csi.2026.104170_b47","series-title":"Proceedings of the IEEE\/ACM 3rd International Conference on AI Engineering - Software Engineering for AI","first-page":"134","article-title":"Green AI: A preliminary empirical study on energy consumption in DL models across different runtime infrastructures","author":"Alizadeh","year":"2024"},{"issue":"8","key":"10.1016\/j.csi.2026.104170_b48","doi-asserted-by":"crossref","first-page":"211:1","DOI":"10.1145\/3680470","article-title":"Enhancing energy-awareness in deep learning through fine-grained energy measurement","volume":"33","author":"Rajput","year":"2024","journal-title":"ACM Trans. Softw. Eng. Methodol."},{"key":"10.1016\/j.csi.2026.104170_b49","series-title":"2021 IEEE International Conference on Big Data (Big Data)","first-page":"2040","article-title":"Recent efficiency gains in deep learning: Performance, power, and sustainability","author":"Hodak","year":"2021"},{"key":"10.1016\/j.csi.2026.104170_b50","series-title":"Caltech 101","author":"Li","year":"2022"},{"key":"10.1016\/j.csi.2026.104170_b51","unstructured":"A. Khosla, N. Jayadevaprakash, B. Yao, L. Fei-Fei, Novel Dataset for Fine-Grained Image Categorization, in: First Workshop on Fine-Grained Visual Categorization, IEEE Conference on Computer Vision and Pattern Recognition, 2011, p. 2."},{"key":"10.1016\/j.csi.2026.104170_b52","doi-asserted-by":"crossref","first-page":"94","DOI":"10.3390\/jimaging7060094","article-title":"Determining chess game state from an image","volume":"7","author":"W\u00f6lflein","year":"2021","journal-title":"J. Imaging"},{"key":"10.1016\/j.csi.2026.104170_b53","series-title":"Chess pieces dataset","author":"Roboflow","year":"2021"},{"key":"10.1016\/j.csi.2026.104170_b54","series-title":"Estimating Deep Learning energy consumption based on model architecture and training environment - Replication package","author":"del Rey","year":"2024"},{"key":"10.1016\/j.csi.2026.104170_b55","series-title":"Green software engineering done right: a scientific guide to set up energy efficiency experiments","author":"Cruz","year":"2021"},{"key":"10.1016\/j.csi.2026.104170_b56","series-title":"Machine Learning and Principles and Practice of Knowledge Discovery in Databases","first-page":"39","article-title":"A unified framework for assessing energy efficiency of machine learning","author":"Fischer","year":"2023"},{"key":"10.1016\/j.csi.2026.104170_b57","series-title":"2017 IEEE International Conference on Data Mining","first-page":"117","article-title":"Matrix profile VIII: Domain agnostic online semantic segmentation at superhuman performance levels","author":"Gharghabi","year":"2017"},{"issue":"39","key":"10.1016\/j.csi.2026.104170_b58","doi-asserted-by":"crossref","first-page":"1504","DOI":"10.21105\/joss.01504","article-title":"STUMPY: A powerful and scalable python library for time series data mining","volume":"4","author":"Law","year":"2019","journal-title":"J. Open Source Softw."},{"key":"10.1016\/j.csi.2026.104170_b59","series-title":"Proceedings of the IEEE\/ACM 3rd International Conference on AI Engineering - Software Engineering for AI","first-page":"118","article-title":"Engineering carbon emission-aware machine learning pipelines","author":"Husom","year":"2024"},{"issue":"4","key":"10.1016\/j.csi.2026.104170_b60","first-page":"34","article-title":"Serverless computing (Dagstuhl Seminar 21201)","volume":"11","author":"Abad","year":"2021","journal-title":"Dagstuhl Rep."},{"key":"10.1016\/j.csi.2026.104170_b61","series-title":"Proceedings of the 44th International Conference on Software Engineering","first-page":"1082","article-title":"Green AI: do deep learning frameworks have different costs?","author":"Georgiou","year":"2022"},{"issue":"5","key":"10.1016\/j.csi.2026.104170_b62","doi-asserted-by":"crossref","first-page":"137:1","DOI":"10.1145\/3708525","article-title":"Efficient and green large language models for software engineering: Literature review, vision, and the road ahead","volume":"34","author":"Shi","year":"2025","journal-title":"ACM Trans. Softw. Eng. Methodol."},{"key":"10.1016\/j.csi.2026.104170_b63","series-title":"2023 ACM\/IEEE International Symposium on Empirical Software Engineering and Measurement","first-page":"1","article-title":"Exploring the carbon footprint of hugging face\u2019s ML models: A repository mining study","author":"Casta\u00f1o","year":"2023"},{"key":"10.1016\/j.csi.2026.104170_b64","series-title":"Towards accurate and reliable energy measurement of NLP models","author":"Cao","year":"2020"},{"issue":"248","key":"10.1016\/j.csi.2026.104170_b65","first-page":"1","article-title":"Towards the systematic reporting of the energy and carbon footprints of machine learning","volume":"21","author":"Henderson","year":"2020","journal-title":"J. Mach. Learn. Res."},{"issue":"4","key":"10.1016\/j.csi.2026.104170_b66","doi-asserted-by":"crossref","first-page":"301","DOI":"10.1007\/s42979-021-00702-9","article-title":"A survey on variational autoencoders from a green AI perspective","volume":"2","author":"Asperti","year":"2021","journal-title":"SN Comput. Sci."},{"key":"10.1016\/j.csi.2026.104170_b67","series-title":"2023 IEEE\/ACM 45th International Conference on Software Engineering","first-page":"2098","article-title":"Twins or false friends? A study on energy consumption and performance of configurable software","author":"Weber","year":"2023"},{"key":"10.1016\/j.csi.2026.104170_b68","series-title":"AI Energy Score Leaderboard - December 2025","author":"Luccioni","year":"2025"},{"key":"10.1016\/j.csi.2026.104170_b69","series-title":"DATE","first-page":"551","article-title":"ICNN: An iterative implementation of convolutional neural networks to enable energy and computational complexity aware dynamic approximation","author":"Neshatpour","year":"2018"},{"key":"10.1016\/j.csi.2026.104170_b70","series-title":"IJCB","first-page":"1","article-title":"MixFaceNets: Extremely efficient face recognition networks","author":"Boutros","year":"2021"},{"key":"10.1016\/j.csi.2026.104170_b71","series-title":"Proceedings of the IEEE\/ACM 3rd International Conference on AI Engineering - Software Engineering for AI","first-page":"256","article-title":"Software design decisions for greener machine learning-based systems","author":"del Rey","year":"2024"}],"container-title":["Computer Standards & Interfaces"],"original-title":[],"language":"en","link":[{"URL":"https:\/\/api.elsevier.com\/content\/article\/PII:S0920548926000449?httpAccept=text\/xml","content-type":"text\/xml","content-version":"vor","intended-application":"text-mining"},{"URL":"https:\/\/api.elsevier.com\/content\/article\/PII:S0920548926000449?httpAccept=text\/plain","content-type":"text\/plain","content-version":"vor","intended-application":"text-mining"}],"deposited":{"date-parts":[[2026,5,23]],"date-time":"2026-05-23T16:48:03Z","timestamp":1779554883000},"score":1,"resource":{"primary":{"URL":"https:\/\/linkinghub.elsevier.com\/retrieve\/pii\/S0920548926000449"}},"subtitle":[],"short-title":[],"issued":{"date-parts":[[2027,1]]},"references-count":71,"alternative-id":["S0920548926000449"],"URL":"https:\/\/doi.org\/10.1016\/j.csi.2026.104170","relation":{},"ISSN":["0920-5489"],"issn-type":[{"value":"0920-5489","type":"print"}],"subject":[],"published":{"date-parts":[[2027,1]]},"assertion":[{"value":"Elsevier","name":"publisher","label":"This article is maintained by"},{"value":"Estimating Deep Learning energy consumption based on model architecture and training environment","name":"articletitle","label":"Article Title"},{"value":"Computer Standards & Interfaces","name":"journaltitle","label":"Journal Title"},{"value":"https:\/\/doi.org\/10.1016\/j.csi.2026.104170","name":"articlelink","label":"CrossRef DOI link to publisher maintained version"},{"value":"article","name":"content_type","label":"Content Type"},{"value":"\u00a9 2026 The Authors. Published by Elsevier B.V.","name":"copyright","label":"Copyright"}],"article-number":"104170"}}