{"id":"https://openalex.org/W3193283516","doi":"https://doi.org/10.1109/tie.2021.3095790","title":"Cooperative Distributed GPU Power Capping for Deep Learning Clusters","display_name":"Cooperative Distributed GPU Power Capping for Deep Learning Clusters","publication_year":2021,"publication_date":"2021-08-24","ids":{"openalex":"https://openalex.org/W3193283516","doi":"https://doi.org/10.1109/tie.2021.3095790","mag":"3193283516"},"language":"en","primary_location":{"id":"doi:10.1109/tie.2021.3095790","is_oa":true,"landing_page_url":"https://doi.org/10.1109/tie.2021.3095790","pdf_url":"https://ieeexplore.ieee.org/ielx7/41/9709233/09484692.pdf","source":{"id":"https://openalex.org/S58031724","display_name":"IEEE Transactions on Industrial Electronics","issn_l":"0278-0046","issn":["0278-0046","1557-9948"],"is_oa":false,"is_in_doaj":false,"is_core":true,"host_organization":"https://openalex.org/P4310319808","host_organization_name":"Institute of Electrical and Electronics Engineers","host_organization_lineage":["https://openalex.org/P4310319808"],"host_organization_lineage_names":["Institute of Electrical and Electronics Engineers"],"type":"journal"},"license":null,"license_id":null,"version":"publishedVersion","is_accepted":true,"is_published":true,"raw_source_name":"IEEE Transactions on Industrial Electronics","raw_type":"journal-article"},"type":"article","indexed_in":["crossref"],"open_access":{"is_oa":true,"oa_status":"bronze","oa_url":"https://ieeexplore.ieee.org/ielx7/41/9709233/09484692.pdf","any_repository_has_fulltext":false},"authorships":[{"author_position":"first","author":{"id":"https://openalex.org/A5091359312","display_name":"Dong-Ki Kang","orcid":"https://orcid.org/0000-0003-1314-8189"},"institutions":[{"id":"https://openalex.org/I80611190","display_name":"Jeonbuk National University","ror":"https://ror.org/05q92br09","country_code":"KR","type":"education","lineage":["https://openalex.org/I80611190"]}],"countries":["KR"],"is_corresponding":false,"raw_author_name":"Dong-Ki Kang","raw_affiliation_strings":["Department of Information Technology, Jeonbuk National University, Jeonju, South Korea"],"raw_orcid":"https://orcid.org/0000-0003-1314-8189","affiliations":[{"raw_affiliation_string":"Department of Information Technology, Jeonbuk National University, Jeonju, South Korea","institution_ids":["https://openalex.org/I80611190"]}]},{"author_position":"middle","author":{"id":"https://openalex.org/A5054923848","display_name":"Yun-Gi Ha","orcid":"https://orcid.org/0000-0001-6437-6988"},"institutions":[{"id":"https://openalex.org/I157485424","display_name":"Korea Advanced Institute of Science and Technology","ror":"https://ror.org/05apxxy63","country_code":"KR","type":"education","lineage":["https://openalex.org/I157485424"]}],"countries":["KR"],"is_corresponding":false,"raw_author_name":"Yun-Gi Ha","raw_affiliation_strings":["School of Electrical Engineering, Korea Advanced Institute of Science and Technology, Daejeon, South Korea"],"raw_orcid":"https://orcid.org/0000-0001-6437-6988","affiliations":[{"raw_affiliation_string":"School of Electrical Engineering, Korea Advanced Institute of Science and Technology, Daejeon, South Korea","institution_ids":["https://openalex.org/I157485424"]}]},{"author_position":"middle","author":{"id":"https://openalex.org/A5045371931","display_name":"Limei Peng","orcid":"https://orcid.org/0000-0001-9984-9861"},"institutions":[{"id":"https://openalex.org/I31419693","display_name":"Kyungpook National University","ror":"https://ror.org/040c17130","country_code":"KR","type":"education","lineage":["https://openalex.org/I31419693"]}],"countries":["KR"],"is_corresponding":false,"raw_author_name":"Limei Peng","raw_affiliation_strings":["School of Computer Science and Engineering, Kyungpook National University, Daegu, South Korea"],"raw_orcid":"https://orcid.org/0000-0001-9984-9861","affiliations":[{"raw_affiliation_string":"School of Computer Science and Engineering, Kyungpook National University, Daegu, South Korea","institution_ids":["https://openalex.org/I31419693"]}]},{"author_position":"last","author":{"id":"https://openalex.org/A5045403869","display_name":"Chan\u2010Hyun Youn","orcid":"https://orcid.org/0000-0002-3970-7308"},"institutions":[{"id":"https://openalex.org/I157485424","display_name":"Korea Advanced Institute of Science and Technology","ror":"https://ror.org/05apxxy63","country_code":"KR","type":"education","lineage":["https://openalex.org/I157485424"]}],"countries":["KR"],"is_corresponding":false,"raw_author_name":"Chan-Hyun Youn","raw_affiliation_strings":["School of Electrical Engineering, Korea Advanced Institute of Science and Technology, Daejeon, South Korea"],"raw_orcid":"https://orcid.org/0000-0002-3970-7308","affiliations":[{"raw_affiliation_string":"School of Electrical Engineering, Korea Advanced Institute of Science and Technology, Daejeon, South Korea","institution_ids":["https://openalex.org/I157485424"]}]}],"institutions":[],"countries_distinct_count":1,"institutions_distinct_count":3,"corresponding_author_ids":[],"corresponding_institution_ids":[],"apc_list":null,"apc_paid":null,"fwci":0.9554,"has_fulltext":true,"cited_by_count":10,"citation_normalized_percentile":{"value":0.79924025,"is_in_top_1_percent":false,"is_in_top_10_percent":false},"cited_by_percentile_year":{"min":89,"max":97},"biblio":{"volume":"69","issue":"7","first_page":"7244","last_page":"7254"},"is_retracted":false,"is_paratext":false,"is_xpac":false,"primary_topic":{"id":"https://openalex.org/T11612","display_name":"Stochastic Gradient Optimization Techniques","score":0.9987000226974487,"subfield":{"id":"https://openalex.org/subfields/1702","display_name":"Artificial Intelligence"},"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/T11612","display_name":"Stochastic Gradient Optimization Techniques","score":0.9987000226974487,"subfield":{"id":"https://openalex.org/subfields/1702","display_name":"Artificial Intelligence"},"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/T13553","display_name":"Age of Information Optimization","score":0.9976000189781189,"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/T12288","display_name":"Optimization and Search Problems","score":0.9943000078201294,"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"}}],"keywords":[{"id":"https://openalex.org/keywords/computer-science","display_name":"Computer science","score":0.7627596855163574},{"id":"https://openalex.org/keywords/deep-learning","display_name":"Deep learning","score":0.4759911298751831},{"id":"https://openalex.org/keywords/artificial-neural-network","display_name":"Artificial neural network","score":0.4309011697769165},{"id":"https://openalex.org/keywords/parallel-computing","display_name":"Parallel computing","score":0.41429635882377625},{"id":"https://openalex.org/keywords/artificial-intelligence","display_name":"Artificial intelligence","score":0.35620978474617004},{"id":"https://openalex.org/keywords/algorithm","display_name":"Algorithm","score":0.3260191082954407}],"concepts":[{"id":"https://openalex.org/C41008148","wikidata":"https://www.wikidata.org/wiki/Q21198","display_name":"Computer science","level":0,"score":0.7627596855163574},{"id":"https://openalex.org/C108583219","wikidata":"https://www.wikidata.org/wiki/Q197536","display_name":"Deep learning","level":2,"score":0.4759911298751831},{"id":"https://openalex.org/C50644808","wikidata":"https://www.wikidata.org/wiki/Q192776","display_name":"Artificial neural network","level":2,"score":0.4309011697769165},{"id":"https://openalex.org/C173608175","wikidata":"https://www.wikidata.org/wiki/Q232661","display_name":"Parallel computing","level":1,"score":0.41429635882377625},{"id":"https://openalex.org/C154945302","wikidata":"https://www.wikidata.org/wiki/Q11660","display_name":"Artificial intelligence","level":1,"score":0.35620978474617004},{"id":"https://openalex.org/C11413529","wikidata":"https://www.wikidata.org/wiki/Q8366","display_name":"Algorithm","level":1,"score":0.3260191082954407}],"mesh":[],"locations_count":1,"locations":[{"id":"doi:10.1109/tie.2021.3095790","is_oa":true,"landing_page_url":"https://doi.org/10.1109/tie.2021.3095790","pdf_url":"https://ieeexplore.ieee.org/ielx7/41/9709233/09484692.pdf","source":{"id":"https://openalex.org/S58031724","display_name":"IEEE Transactions on Industrial Electronics","issn_l":"0278-0046","issn":["0278-0046","1557-9948"],"is_oa":false,"is_in_doaj":false,"is_core":true,"host_organization":"https://openalex.org/P4310319808","host_organization_name":"Institute of Electrical and Electronics Engineers","host_organization_lineage":["https://openalex.org/P4310319808"],"host_organization_lineage_names":["Institute of Electrical and Electronics Engineers"],"type":"journal"},"license":null,"license_id":null,"version":"publishedVersion","is_accepted":true,"is_published":true,"raw_source_name":"IEEE Transactions on Industrial Electronics","raw_type":"journal-article"}],"best_oa_location":{"id":"doi:10.1109/tie.2021.3095790","is_oa":true,"landing_page_url":"https://doi.org/10.1109/tie.2021.3095790","pdf_url":"https://ieeexplore.ieee.org/ielx7/41/9709233/09484692.pdf","source":{"id":"https://openalex.org/S58031724","display_name":"IEEE Transactions on Industrial Electronics","issn_l":"0278-0046","issn":["0278-0046","1557-9948"],"is_oa":false,"is_in_doaj":false,"is_core":true,"host_organization":"https://openalex.org/P4310319808","host_organization_name":"Institute of Electrical and Electronics Engineers","host_organization_lineage":["https://openalex.org/P4310319808"],"host_organization_lineage_names":["Institute of Electrical and Electronics Engineers"],"type":"journal"},"license":null,"license_id":null,"version":"publishedVersion","is_accepted":true,"is_published":true,"raw_source_name":"IEEE Transactions on Industrial Electronics","raw_type":"journal-article"},"sustainable_development_goals":[{"score":0.6299999952316284,"display_name":"Decent work and economic growth","id":"https://metadata.un.org/sdg/8"}],"awards":[{"id":"https://openalex.org/G1979152044","display_name":null,"funder_award_id":"20ZS1100","funder_id":"https://openalex.org/F4320322093","funder_display_name":"Electronics and Telecommunications Research Institute"}],"funders":[{"id":"https://openalex.org/F4320322093","display_name":"Electronics and Telecommunications Research Institute","ror":"https://ror.org/03ysstz10"},{"id":"https://openalex.org/F4320323103","display_name":"Agency for Defense Development","ror":"https://ror.org/05fhe0r85"}],"has_content":{"grobid_xml":true,"pdf":true},"content_urls":{"pdf":"https://content.openalex.org/works/W3193283516.pdf","grobid_xml":"https://content.openalex.org/works/W3193283516.grobid-xml"},"referenced_works_count":26,"referenced_works":["https://openalex.org/W1686810756","https://openalex.org/W2008114469","https://openalex.org/W2038337218","https://openalex.org/W2061328016","https://openalex.org/W2062454966","https://openalex.org/W2074084090","https://openalex.org/W2094336028","https://openalex.org/W2114013570","https://openalex.org/W2132101661","https://openalex.org/W2162986857","https://openalex.org/W2183341477","https://openalex.org/W2194775991","https://openalex.org/W2512094163","https://openalex.org/W2513383847","https://openalex.org/W2522820499","https://openalex.org/W2618530766","https://openalex.org/W2760837370","https://openalex.org/W2794513473","https://openalex.org/W2945148926","https://openalex.org/W2988430901","https://openalex.org/W3022298203","https://openalex.org/W3037110041","https://openalex.org/W3043273434","https://openalex.org/W4250589301","https://openalex.org/W4254052724","https://openalex.org/W6758283263"],"related_works":["https://openalex.org/W2731899572","https://openalex.org/W3215138031","https://openalex.org/W2117014006","https://openalex.org/W3009238340","https://openalex.org/W2939353110","https://openalex.org/W4233815414","https://openalex.org/W4321369474","https://openalex.org/W4360585206","https://openalex.org/W2372170743","https://openalex.org/W4327774331"],"abstract_inverted_index":{"The":[0],"recent":[1,204],"GPU-based":[2,47,64,151],"clusters":[3,37,48],"that":[4,167],"handle":[5],"deep":[6,20],"learning":[7],"(DL)":[8],"tasks":[9,76,198],"have":[10,177],"the":[11,30,46,69,79,87,92,98,106,127,149,168,172,191],"features":[12],"of":[13,19,73,117,182,195],"GPU":[14,40,58,118,144],"device":[15],"heterogeneity,":[16],"a":[17,55,132,178],"variety":[18],"neural":[21],"network":[22],"(DNN)":[23],"models,":[24],"and":[25,114,155,189],"high":[26],"computational":[27],"complexity.":[28],"Thus,":[29],"traditional":[31],"power":[32,59,81,115,173],"capping":[33,60,174],"methods":[34],"for":[35,63,109,141,161],"CPU-based":[36],"or":[38],"small-scale":[39],"devices":[41,119],"cannot":[42],"be":[43],"applied":[44],"to":[45,67,176],"handling":[49],"DL":[50,75,110,197],"tasks.":[51],"This":[52,103],"article":[53],"develops":[54],"cooperative":[56],"distributed":[57],"(CD-GPC)":[61],"system":[62,170],"clusters,":[65],"aiming":[66],"minimize":[68],"training":[70,112],"completion":[71],"time":[72,113],"invoked":[74,196],"without":[77,120],"exceeding":[78],"limited":[80],"budget.":[82],"Specifically,":[83],"we":[84,125],"first":[85],"design":[86],"frequency":[88],"scaling":[89],"approach":[90,104],"using":[91],"online":[93],"model":[94,137],"estimation":[95],"based":[96],"on":[97],"recursive":[99],"least":[100],"square":[101],"method.":[102],"achieves":[105],"accurate":[107],"tuning":[108],"task":[111],"usage":[116],"needing":[121],"offline":[122],"profiling.":[123],"Then,":[124],"formulate":[126],"proposed":[128,169],"FS":[129],"problem":[130,140],"as":[131],"Lagrangian":[133],"dual":[134],"decomposition-based":[135],"economic":[136],"predictive":[138],"control":[139],"large-scale":[142],"heterogeneous":[143],"clusters.":[145],"We":[146],"conduct":[147],"both":[148],"NVIDIA":[150],"lab-scale":[152],"real":[153,156],"experiments":[154,160],"job":[157],"trace-based":[158],"simulation":[159],"performance":[162],"evaluation.":[163],"Experimental":[164],"results":[165],"validate":[166],"improves":[171],"accuracy":[175],"mean":[179],"absolute":[180],"error":[181],"<inline-formula":[183],"xmlns:mml=\"http://www.w3.org/1998/Math/MathML\"":[184],"xmlns:xlink=\"http://www.w3.org/1999/xlink\"><tex-math":[185],"notation=\"LaTeX\">$&lt;":[186],"\\!1\\%$</tex-math></inline-formula>":[187],",":[188],"reduces":[190],"deadline":[192],"violation":[193],"ratio":[194],"by":[199],"21.5%":[200],"compared":[201],"with":[202],"other":[203],"counterparts.":[205]},"counts_by_year":[{"year":2025,"cited_by_count":3},{"year":2024,"cited_by_count":3},{"year":2023,"cited_by_count":1},{"year":2022,"cited_by_count":3}],"updated_date":"2025-11-06T03:46:38.306776","created_date":"2025-10-10T00:00:00"}