{"id":"https://openalex.org/W4392089995","doi":"https://doi.org/10.1145/3644815.3644967","title":"Green AI: a Preliminary Empirical Study on Energy Consumption in DL Models Across Different Runtime Infrastructures","display_name":"Green AI: a Preliminary Empirical Study on Energy Consumption in DL Models Across Different Runtime Infrastructures","publication_year":2024,"publication_date":"2024-04-14","ids":{"openalex":"https://openalex.org/W4392089995","doi":"https://doi.org/10.1145/3644815.3644967"},"language":"en","primary_location":{"id":"doi:10.1145/3644815.3644967","is_oa":true,"landing_page_url":"https://doi.org/10.1145/3644815.3644967","pdf_url":"https://dl.acm.org/doi/pdf/10.1145/3644815.3644967","source":null,"license":null,"license_id":null,"version":"publishedVersion","is_accepted":true,"is_published":true,"raw_source_name":"Proceedings of the IEEE/ACM 3rd International Conference on AI Engineering - Software Engineering for AI","raw_type":"proceedings-article"},"type":"preprint","indexed_in":["arxiv","crossref"],"open_access":{"is_oa":true,"oa_status":"gold","oa_url":"https://dl.acm.org/doi/pdf/10.1145/3644815.3644967","any_repository_has_fulltext":true},"authorships":[{"author_position":"first","author":{"id":"https://openalex.org/A5099048584","display_name":"Negar Alizadeh","orcid":"https://orcid.org/0009-0005-8155-165X"},"institutions":[{"id":"https://openalex.org/I193662353","display_name":"Utrecht University","ror":"https://ror.org/04pp8hn57","country_code":"NL","type":"education","lineage":["https://openalex.org/I193662353"]}],"countries":["NL"],"is_corresponding":true,"raw_author_name":"Negar Alizadeh","raw_affiliation_strings":["Utrecht University, Utrecht, Netherlands"],"raw_orcid":"https://orcid.org/0009-0005-8155-165X","affiliations":[{"raw_affiliation_string":"Utrecht University, Utrecht, Netherlands","institution_ids":["https://openalex.org/I193662353"]}]},{"author_position":"last","author":{"id":"https://openalex.org/A5062400717","display_name":"Fernando Castor","orcid":"https://orcid.org/0000-0002-6389-3630"},"institutions":[{"id":"https://openalex.org/I25112270","display_name":"Universidade Federal de Pernambuco","ror":"https://ror.org/047908t24","country_code":"BR","type":"education","lineage":["https://openalex.org/I25112270"]},{"id":"https://openalex.org/I94624287","display_name":"University of Twente","ror":"https://ror.org/006hf6230","country_code":"NL","type":"education","lineage":["https://openalex.org/I94624287"]}],"countries":["BR","NL"],"is_corresponding":false,"raw_author_name":"Fernando Castor","raw_affiliation_strings":["Federal University of Pernambuco, Recife, Brazil","University of Twente, Enschede, Netherlands"],"raw_orcid":"https://orcid.org/0000-0002-6389-3630","affiliations":[{"raw_affiliation_string":"Federal University of Pernambuco, Recife, Brazil","institution_ids":["https://openalex.org/I25112270"]},{"raw_affiliation_string":"University of Twente, Enschede, Netherlands","institution_ids":["https://openalex.org/I94624287"]}]}],"institutions":[],"countries_distinct_count":2,"institutions_distinct_count":2,"corresponding_author_ids":["https://openalex.org/A5099048584"],"corresponding_institution_ids":["https://openalex.org/I193662353"],"apc_list":null,"apc_paid":null,"fwci":5.8874,"has_fulltext":true,"cited_by_count":26,"citation_normalized_percentile":{"value":0.97107635,"is_in_top_1_percent":false,"is_in_top_10_percent":true},"cited_by_percentile_year":{"min":98,"max":100},"biblio":{"volume":null,"issue":null,"first_page":"134","last_page":"139"},"is_retracted":false,"is_paratext":false,"is_xpac":false,"primary_topic":{"id":"https://openalex.org/T10036","display_name":"Advanced Neural Network Applications","score":0.9994999766349792,"subfield":{"id":"https://openalex.org/subfields/1707","display_name":"Computer Vision and Pattern Recognition"},"field":{"id":"https://openalex.org/fields/17","display_name":"Computer Science"},"domain":{"id":"https://openalex.org/domains/3","display_name":"Physical Sciences"}},"topics":[{"id":"https://openalex.org/T10036","display_name":"Advanced Neural Network Applications","score":0.9994999766349792,"subfield":{"id":"https://openalex.org/subfields/1707","display_name":"Computer Vision and Pattern Recognition"},"field":{"id":"https://openalex.org/fields/17","display_name":"Computer Science"},"domain":{"id":"https://openalex.org/domains/3","display_name":"Physical Sciences"}},{"id":"https://openalex.org/T10273","display_name":"IoT and Edge/Fog Computing","score":0.9889000058174133,"subfield":{"id":"https://openalex.org/subfields/1705","display_name":"Computer Networks and Communications"},"field":{"id":"https://openalex.org/fields/17","display_name":"Computer Science"},"domain":{"id":"https://openalex.org/domains/3","display_name":"Physical Sciences"}},{"id":"https://openalex.org/T10054","display_name":"Parallel Computing and Optimization Techniques","score":0.9832000136375427,"subfield":{"id":"https://openalex.org/subfields/1708","display_name":"Hardware and Architecture"},"field":{"id":"https://openalex.org/fields/17","display_name":"Computer Science"},"domain":{"id":"https://openalex.org/domains/3","display_name":"Physical Sciences"}}],"keywords":[{"id":"https://openalex.org/keywords/computer-science","display_name":"Computer science","score":0.86798095703125},{"id":"https://openalex.org/keywords/inference","display_name":"Inference","score":0.6987130641937256},{"id":"https://openalex.org/keywords/cuda","display_name":"CUDA","score":0.6778685450553894},{"id":"https://openalex.org/keywords/deep-learning","display_name":"Deep learning","score":0.5709565877914429},{"id":"https://openalex.org/keywords/energy-consumption","display_name":"Energy consumption","score":0.5698680877685547},{"id":"https://openalex.org/keywords/efficient-energy-use","display_name":"Efficient energy use","score":0.5635074377059937},{"id":"https://openalex.org/keywords/artificial-intelligence","display_name":"Artificial intelligence","score":0.498502254486084},{"id":"https://openalex.org/keywords/general-purpose-computing-on-graphics-processing-units","display_name":"General-purpose computing on graphics processing units","score":0.43396249413490295},{"id":"https://openalex.org/keywords/execution-time","display_name":"Execution time","score":0.4238240718841553},{"id":"https://openalex.org/keywords/machine-learning","display_name":"Machine learning","score":0.40808671712875366},{"id":"https://openalex.org/keywords/parallel-computing","display_name":"Parallel computing","score":0.3814036250114441},{"id":"https://openalex.org/keywords/computer-engineering","display_name":"Computer engineering","score":0.34456533193588257},{"id":"https://openalex.org/keywords/computer-architecture","display_name":"Computer architecture","score":0.33417022228240967},{"id":"https://openalex.org/keywords/operating-system","display_name":"Operating system","score":0.14996913075447083}],"concepts":[{"id":"https://openalex.org/C41008148","wikidata":"https://www.wikidata.org/wiki/Q21198","display_name":"Computer science","level":0,"score":0.86798095703125},{"id":"https://openalex.org/C2776214188","wikidata":"https://www.wikidata.org/wiki/Q408386","display_name":"Inference","level":2,"score":0.6987130641937256},{"id":"https://openalex.org/C2778119891","wikidata":"https://www.wikidata.org/wiki/Q477690","display_name":"CUDA","level":2,"score":0.6778685450553894},{"id":"https://openalex.org/C108583219","wikidata":"https://www.wikidata.org/wiki/Q197536","display_name":"Deep learning","level":2,"score":0.5709565877914429},{"id":"https://openalex.org/C2780165032","wikidata":"https://www.wikidata.org/wiki/Q16869822","display_name":"Energy consumption","level":2,"score":0.5698680877685547},{"id":"https://openalex.org/C2742236","wikidata":"https://www.wikidata.org/wiki/Q924713","display_name":"Efficient energy use","level":2,"score":0.5635074377059937},{"id":"https://openalex.org/C154945302","wikidata":"https://www.wikidata.org/wiki/Q11660","display_name":"Artificial intelligence","level":1,"score":0.498502254486084},{"id":"https://openalex.org/C50630238","wikidata":"https://www.wikidata.org/wiki/Q971505","display_name":"General-purpose computing on graphics processing units","level":3,"score":0.43396249413490295},{"id":"https://openalex.org/C2989134064","wikidata":"https://www.wikidata.org/wiki/Q288510","display_name":"Execution time","level":2,"score":0.4238240718841553},{"id":"https://openalex.org/C119857082","wikidata":"https://www.wikidata.org/wiki/Q2539","display_name":"Machine learning","level":1,"score":0.40808671712875366},{"id":"https://openalex.org/C173608175","wikidata":"https://www.wikidata.org/wiki/Q232661","display_name":"Parallel computing","level":1,"score":0.3814036250114441},{"id":"https://openalex.org/C113775141","wikidata":"https://www.wikidata.org/wiki/Q428691","display_name":"Computer engineering","level":1,"score":0.34456533193588257},{"id":"https://openalex.org/C118524514","wikidata":"https://www.wikidata.org/wiki/Q173212","display_name":"Computer architecture","level":1,"score":0.33417022228240967},{"id":"https://openalex.org/C111919701","wikidata":"https://www.wikidata.org/wiki/Q9135","display_name":"Operating system","level":1,"score":0.14996913075447083},{"id":"https://openalex.org/C21442007","wikidata":"https://www.wikidata.org/wiki/Q1027879","display_name":"Graphics","level":2,"score":0.0},{"id":"https://openalex.org/C127413603","wikidata":"https://www.wikidata.org/wiki/Q11023","display_name":"Engineering","level":0,"score":0.0},{"id":"https://openalex.org/C119599485","wikidata":"https://www.wikidata.org/wiki/Q43035","display_name":"Electrical engineering","level":1,"score":0.0},{"id":"https://openalex.org/C18903297","wikidata":"https://www.wikidata.org/wiki/Q7150","display_name":"Ecology","level":1,"score":0.0},{"id":"https://openalex.org/C86803240","wikidata":"https://www.wikidata.org/wiki/Q420","display_name":"Biology","level":0,"score":0.0}],"mesh":[],"locations_count":5,"locations":[{"id":"doi:10.1145/3644815.3644967","is_oa":true,"landing_page_url":"https://doi.org/10.1145/3644815.3644967","pdf_url":"https://dl.acm.org/doi/pdf/10.1145/3644815.3644967","source":null,"license":null,"license_id":null,"version":"publishedVersion","is_accepted":true,"is_published":true,"raw_source_name":"Proceedings of the IEEE/ACM 3rd International Conference on AI Engineering - Software Engineering for AI","raw_type":"proceedings-article"},{"id":"pmh:oai:ris.utwente.nl:openaire/c60d110f-be84-4b00-8f8b-4754f63e414c","is_oa":true,"landing_page_url":"https://research.utwente.nl/en/publications/c60d110f-be84-4b00-8f8b-4754f63e414c","pdf_url":null,"source":{"id":"https://openalex.org/S4406922991","display_name":"University of Twente Research Information","issn_l":null,"issn":null,"is_oa":false,"is_in_doaj":false,"is_core":false,"host_organization":null,"host_organization_name":null,"host_organization_lineage":[],"host_organization_lineage_names":[],"type":"repository"},"license":"other-oa","license_id":"https://openalex.org/licenses/other-oa","version":"publishedVersion","is_accepted":true,"is_published":true,"raw_source_name":"Alizadeh, N & Castor, F 2024, Green AI : A Preliminary Empirical Study on Energy Consumption in DL Models Across Different Runtime Infrastructures. in 2024 IEEE/ACM 3rd International Conference on AI Engineering \u2013 Software Engineering for AI (CAIN)., 10556302, IEEE, pp. 134-139, IEEE/ACM 3rd International Conference on AI Engineering \u2013 Software Engineering for AI, CAIN 2024, Lisbon, Portugal, 14/04/24. https://doi.org/10.1145/3644815.3644967","raw_type":"info:eu-repo/semantics/publishedVersion"},{"id":"pmh:oai:arXiv.org:2402.13640","is_oa":true,"landing_page_url":"http://arxiv.org/abs/2402.13640","pdf_url":"https://arxiv.org/pdf/2402.13640","source":{"id":"https://openalex.org/S4306400194","display_name":"arXiv (Cornell University)","issn_l":null,"issn":null,"is_oa":true,"is_in_doaj":false,"is_core":false,"host_organization":"https://openalex.org/I205783295","host_organization_name":"Cornell University","host_organization_lineage":["https://openalex.org/I205783295"],"host_organization_lineage_names":[],"type":"repository"},"license":null,"license_id":null,"version":"submittedVersion","is_accepted":false,"is_published":false,"raw_source_name":null,"raw_type":"text"},{"id":"pmh:oai:ris.utwente.nl:publications/c60d110f-be84-4b00-8f8b-4754f63e414c","is_oa":true,"landing_page_url":"https://research.utwente.nl/files/459347644/Green_AI_A_Preliminary_Empirical_Study_on_Energy_Consumption_in_DL_Models_Across_Different_Runtime_Infrastructures.pdf","pdf_url":"https://ris.utwente.nl/ws/files/459347644/Green_AI_A_Preliminary_Empirical_Study_on_Energy_Consumption_in_DL_Models_Across_Different_Runtime_Infrastructures.pdf","source":{"id":"https://openalex.org/S4406922991","display_name":"University of Twente Research Information","issn_l":null,"issn":null,"is_oa":false,"is_in_doaj":false,"is_core":false,"host_organization":null,"host_organization_name":null,"host_organization_lineage":[],"host_organization_lineage_names":[],"type":"repository"},"license":"other-oa","license_id":"https://openalex.org/licenses/other-oa","version":"submittedVersion","is_accepted":false,"is_published":false,"raw_source_name":"Alizadeh, N & Castor, F 2024, Green AI : A Preliminary Empirical Study on Energy Consumption in DL Models Across Different Runtime Infrastructures. in 2024 IEEE/ACM 3rd International Conference on AI Engineering \u2013 Software Engineering for AI (CAIN)., 10556302, IEEE, pp. 134-139, IEEE/ACM 3rd International Conference on AI Engineering \u2013 Software Engineering for AI, CAIN 2024, Lisbon, Portugal, 14/04/24. https://doi.org/10.1145/3644815.3644967","raw_type":"info:eu-repo/semantics/publishedVersion"},{"id":"pmh:oai:dspace.library.uu.nl:1874/482340","is_oa":true,"landing_page_url":"https://dspace.library.uu.nl/handle/1874/482340","pdf_url":null,"source":{"id":"https://openalex.org/S4306401649","display_name":"Utrecht University Repository (Utrecht University)","issn_l":null,"issn":null,"is_oa":false,"is_in_doaj":false,"is_core":false,"host_organization":"https://openalex.org/I193662353","host_organization_name":"Utrecht University","host_organization_lineage":["https://openalex.org/I193662353"],"host_organization_lineage_names":[],"type":"repository"},"license":"other-oa","license_id":"https://openalex.org/licenses/other-oa","version":"submittedVersion","is_accepted":false,"is_published":false,"raw_source_name":null,"raw_type":"Part of book"}],"best_oa_location":{"id":"doi:10.1145/3644815.3644967","is_oa":true,"landing_page_url":"https://doi.org/10.1145/3644815.3644967","pdf_url":"https://dl.acm.org/doi/pdf/10.1145/3644815.3644967","source":null,"license":null,"license_id":null,"version":"publishedVersion","is_accepted":true,"is_published":true,"raw_source_name":"Proceedings of the IEEE/ACM 3rd International Conference on AI Engineering - Software Engineering for AI","raw_type":"proceedings-article"},"sustainable_development_goals":[{"score":0.6499999761581421,"display_name":"Affordable and clean energy","id":"https://metadata.un.org/sdg/7"}],"awards":[],"funders":[],"has_content":{"grobid_xml":false,"pdf":true},"content_urls":{"pdf":"https://content.openalex.org/works/W4392089995.pdf"},"referenced_works_count":14,"referenced_works":["https://openalex.org/W2038964933","https://openalex.org/W2194775991","https://openalex.org/W2769228904","https://openalex.org/W2904169529","https://openalex.org/W2963809228","https://openalex.org/W2969580338","https://openalex.org/W2974470161","https://openalex.org/W3084812981","https://openalex.org/W3153920384","https://openalex.org/W4225793574","https://openalex.org/W4284664530","https://openalex.org/W4286287841","https://openalex.org/W4312825166","https://openalex.org/W4399433123"],"related_works":["https://openalex.org/W1963859303","https://openalex.org/W2364044215","https://openalex.org/W2389600408","https://openalex.org/W240129890","https://openalex.org/W3048701459","https://openalex.org/W2149078538","https://openalex.org/W2080146221","https://openalex.org/W2370314112","https://openalex.org/W1912958759","https://openalex.org/W2792081825"],"abstract_inverted_index":{"Deep":[0],"Learning":[1],"(DL)":[2],"frameworks":[3,78,120],"such":[4,55],"as":[5,56,68,121,123],"PyTorch":[6,166,193,205],"and":[7,24,63,79,109,152,167,183,194],"TensorFlow":[8,168,199],"include":[9],"runtime":[10,61,84,114],"infrastructures":[11,85,115],"responsible":[12],"for":[13,169],"executing":[14],"trained":[15],"models":[16,38,173,212],"on":[17,39,90],"target":[18],"hardware,":[19],"managing":[20],"memory,":[21],"data":[22],"transfers,":[23],"multi-accelerator":[25],"execution,":[26],"if":[27],"applicable.":[28],"Additionally,":[29],"it":[30],"is":[31,200],"a":[32,87],"common":[33],"practice":[34],"to":[35,49,159,179,213],"deploy":[36],"pre-trained":[37],"environments":[40],"distinct":[41],"from":[42],"their":[43,98],"native":[44],"development":[45],"settings.":[46],"This":[47],"led":[48],"the":[50,106,113,139,150,170,207,211,220],"introduction":[51],"of":[52,116,141,155,222,229],"interchange":[53],"formats":[54,70],"ONNX,":[57,124],"which":[58,66],"includes":[59],"its":[60],"infrastructure,":[62],"ONNX":[64,214],"Runtime,":[65],"work":[67],"standard":[69],"that":[71,149,234],"can":[72],"be":[73],"used":[74],"across":[75],"diverse":[76],"DL":[77,119,128,156],"languages.":[80],"Even":[81],"though":[82],"these":[83],"have":[86,131],"great":[88],"impact":[89,140],"inference":[91,110],"performance,":[92],"no":[93],"previous":[94],"paper":[95],"has":[96],"investigated":[97],"energy":[99,107,153],"efficiency.":[100],"In":[101],"this":[102],"study,":[103],"we":[104,136,232],"monitor":[105],"consumption":[108],"time":[111],"in":[112,133,219,225],"three":[117,126],"well-known":[118],"well":[122],"using":[125,142,174],"various":[127],"models.":[129],"To":[130],"nuance":[132],"our":[134,226],"investigation,":[135],"also":[137],"examine":[138],"different":[143],"execution":[144,230],"providers.":[145],"We":[146],"find":[147],"out":[148],"performance":[151,217],"efficiency":[154],"are":[157],"difficult":[158],"predict.":[160],"One":[161],"framework,":[162],"MXNet,":[163],"outperforms":[164,237],"both":[165],"computer":[171],"vision":[172],"batch":[175,189],"size":[176,190],"1,":[177],"due":[178],"efficient":[180],"GPU":[181],"usage":[182],"thus":[184],"low":[185],"CPU":[186],"usage.":[187],"However,":[188],"64":[191],"makes":[192],"MXNet":[195],"practically":[196],"indistinguishable,":[197],"while":[198],"outperformed":[201],"consistently.":[202],"For":[203],"BERT,":[204],"exhibits":[206],"best":[208],"performance.":[209],"Converting":[210],"yields":[215],"significant":[216],"improvements":[218],"majority":[221],"cases.":[223],"Finally,":[224],"preliminary":[227],"investigation":[228],"providers,":[231],"observe":[233],"TensorRT":[235],"always":[236],"CUDA.":[238]},"counts_by_year":[{"year":2026,"cited_by_count":2},{"year":2025,"cited_by_count":17},{"year":2024,"cited_by_count":7}],"updated_date":"2026-05-19T08:33:51.333923","created_date":"2024-02-23T00:00:00"}