{"id":"https://openalex.org/W7114929646","doi":"https://doi.org/10.1109/tc.2025.3642827","title":"GreenDLS: An Energy-Efficient and SLO-Aware Deep Learning Serving System","display_name":"GreenDLS: An Energy-Efficient and SLO-Aware Deep Learning Serving System","publication_year":2025,"publication_date":"2025-12-12","ids":{"openalex":"https://openalex.org/W7114929646","doi":"https://doi.org/10.1109/tc.2025.3642827"},"language":null,"primary_location":{"id":"doi:10.1109/tc.2025.3642827","is_oa":false,"landing_page_url":"https://doi.org/10.1109/tc.2025.3642827","pdf_url":null,"source":{"id":"https://openalex.org/S157670870","display_name":"IEEE Transactions on Computers","issn_l":"0018-9340","issn":["0018-9340","0016-9340","1557-9956","2326-3814"],"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 Computers","raw_type":"journal-article"},"type":"article","indexed_in":["crossref"],"open_access":{"is_oa":false,"oa_status":"closed","oa_url":null,"any_repository_has_fulltext":false},"authorships":[{"author_position":"first","author":{"id":null,"display_name":"Meng Hao","orcid":"https://orcid.org/0000-0003-0043-4370"},"institutions":[{"id":"https://openalex.org/I204983213","display_name":"Harbin Institute of Technology","ror":"https://ror.org/01yqg2h08","country_code":"CN","type":"education","lineage":["https://openalex.org/I204983213"]}],"countries":["CN"],"is_corresponding":false,"raw_author_name":"Meng Hao","raw_affiliation_strings":["Harbin Institute of Technology, Harbin, China"],"raw_orcid":"https://orcid.org/0000-0003-0043-4370","affiliations":[{"raw_affiliation_string":"Harbin Institute of Technology, Harbin, China","institution_ids":["https://openalex.org/I204983213"]}]},{"author_position":"middle","author":{"id":null,"display_name":"Xueyang Tian","orcid":"https://orcid.org/0009-0001-0415-327X"},"institutions":[{"id":"https://openalex.org/I204983213","display_name":"Harbin Institute of Technology","ror":"https://ror.org/01yqg2h08","country_code":"CN","type":"education","lineage":["https://openalex.org/I204983213"]}],"countries":["CN"],"is_corresponding":false,"raw_author_name":"Xueyang Tian","raw_affiliation_strings":["Harbin Institute of Technology, Harbin, China"],"raw_orcid":"https://orcid.org/0009-0001-0415-327X","affiliations":[{"raw_affiliation_string":"Harbin Institute of Technology, Harbin, China","institution_ids":["https://openalex.org/I204983213"]}]},{"author_position":"middle","author":{"id":null,"display_name":"Siyu Yang","orcid":"https://orcid.org/0009-0000-0071-6218"},"institutions":[{"id":"https://openalex.org/I204983213","display_name":"Harbin Institute of Technology","ror":"https://ror.org/01yqg2h08","country_code":"CN","type":"education","lineage":["https://openalex.org/I204983213"]}],"countries":["CN"],"is_corresponding":false,"raw_author_name":"Siyu Yang","raw_affiliation_strings":["Harbin Institute of Technology, Harbin, China"],"raw_orcid":"https://orcid.org/0009-0000-0071-6218","affiliations":[{"raw_affiliation_string":"Harbin Institute of Technology, Harbin, China","institution_ids":["https://openalex.org/I204983213"]}]},{"author_position":"middle","author":{"id":null,"display_name":"Hongwei Yang","orcid":"https://orcid.org/0000-0002-8386-0131"},"institutions":[{"id":"https://openalex.org/I204983213","display_name":"Harbin Institute of Technology","ror":"https://ror.org/01yqg2h08","country_code":"CN","type":"education","lineage":["https://openalex.org/I204983213"]}],"countries":["CN"],"is_corresponding":false,"raw_author_name":"Hongwei Yang","raw_affiliation_strings":["Harbin Institute of Technology, Harbin, China"],"raw_orcid":"https://orcid.org/0000-0002-8386-0131","affiliations":[{"raw_affiliation_string":"Harbin Institute of Technology, Harbin, China","institution_ids":["https://openalex.org/I204983213"]}]},{"author_position":"middle","author":{"id":null,"display_name":"Hui He","orcid":"https://orcid.org/0000-0002-6494-775X"},"institutions":[{"id":"https://openalex.org/I204983213","display_name":"Harbin Institute of Technology","ror":"https://ror.org/01yqg2h08","country_code":"CN","type":"education","lineage":["https://openalex.org/I204983213"]}],"countries":["CN"],"is_corresponding":false,"raw_author_name":"Hui He","raw_affiliation_strings":["Harbin Institute of Technology, Harbin, China"],"raw_orcid":"https://orcid.org/0000-0002-6494-775X","affiliations":[{"raw_affiliation_string":"Harbin Institute of Technology, Harbin, China","institution_ids":["https://openalex.org/I204983213"]}]},{"author_position":"middle","author":{"id":null,"display_name":"Xuan Li","orcid":"https://orcid.org/0009-0005-6658-3290"},"institutions":[{"id":"https://openalex.org/I204983213","display_name":"Harbin Institute of Technology","ror":"https://ror.org/01yqg2h08","country_code":"CN","type":"education","lineage":["https://openalex.org/I204983213"]}],"countries":["CN"],"is_corresponding":false,"raw_author_name":"Xuan Li","raw_affiliation_strings":["Harbin Institute of Technology, Harbin, China"],"raw_orcid":"https://orcid.org/0009-0005-6658-3290","affiliations":[{"raw_affiliation_string":"Harbin Institute of Technology, Harbin, China","institution_ids":["https://openalex.org/I204983213"]}]},{"author_position":"middle","author":{"id":null,"display_name":"Yiming Wang","orcid":"https://orcid.org/0000-0003-3298-5134"},"institutions":[{"id":"https://openalex.org/I204983213","display_name":"Harbin Institute of Technology","ror":"https://ror.org/01yqg2h08","country_code":"CN","type":"education","lineage":["https://openalex.org/I204983213"]}],"countries":["CN"],"is_corresponding":false,"raw_author_name":"Yiming Wang","raw_affiliation_strings":["Harbin Institute of Technology, Harbin, China"],"raw_orcid":"https://orcid.org/0000-0003-3298-5134","affiliations":[{"raw_affiliation_string":"Harbin Institute of Technology, Harbin, China","institution_ids":["https://openalex.org/I204983213"]}]},{"author_position":"middle","author":{"id":null,"display_name":"Zheng Liu","orcid":"https://orcid.org/0009-0006-4934-480X"},"institutions":[{"id":"https://openalex.org/I204983213","display_name":"Harbin Institute of Technology","ror":"https://ror.org/01yqg2h08","country_code":"CN","type":"education","lineage":["https://openalex.org/I204983213"]}],"countries":["CN"],"is_corresponding":false,"raw_author_name":"Zheng Liu","raw_affiliation_strings":["Harbin Institute of Technology, Harbin, China"],"raw_orcid":"https://orcid.org/0009-0006-4934-480X","affiliations":[{"raw_affiliation_string":"Harbin Institute of Technology, Harbin, China","institution_ids":["https://openalex.org/I204983213"]}]},{"author_position":"middle","author":{"id":null,"display_name":"Desheng Wang","orcid":"https://orcid.org/0000-0002-7502-7094"},"institutions":[{"id":"https://openalex.org/I204983213","display_name":"Harbin Institute of Technology","ror":"https://ror.org/01yqg2h08","country_code":"CN","type":"education","lineage":["https://openalex.org/I204983213"]}],"countries":["CN"],"is_corresponding":false,"raw_author_name":"Desheng Wang","raw_affiliation_strings":["Guangdong Provincial Key Laboratory of Novel Security Intelligence Technologies, Harbin Institute of Technology (Shenzhen), Shenzhen, China"],"raw_orcid":"https://orcid.org/0000-0002-7502-7094","affiliations":[{"raw_affiliation_string":"Guangdong Provincial Key Laboratory of Novel Security Intelligence Technologies, Harbin Institute of Technology (Shenzhen), Shenzhen, China","institution_ids":["https://openalex.org/I204983213"]}]},{"author_position":"last","author":{"id":null,"display_name":"Weizhe Zhang","orcid":"https://orcid.org/0000-0003-4783-876X"},"institutions":[{"id":"https://openalex.org/I204983213","display_name":"Harbin Institute of Technology","ror":"https://ror.org/01yqg2h08","country_code":"CN","type":"education","lineage":["https://openalex.org/I204983213"]}],"countries":["CN"],"is_corresponding":false,"raw_author_name":"Weizhe Zhang","raw_affiliation_strings":["Harbin Institute of Technology, Harbin, China"],"raw_orcid":"https://orcid.org/0000-0003-4783-876X","affiliations":[{"raw_affiliation_string":"Harbin Institute of Technology, Harbin, China","institution_ids":["https://openalex.org/I204983213"]}]}],"institutions":[],"countries_distinct_count":1,"institutions_distinct_count":1,"corresponding_author_ids":[],"corresponding_institution_ids":["https://openalex.org/I204983213"],"apc_list":null,"apc_paid":null,"fwci":0.0,"has_fulltext":false,"cited_by_count":0,"citation_normalized_percentile":{"value":0.46694755,"is_in_top_1_percent":false,"is_in_top_10_percent":false},"cited_by_percentile_year":null,"biblio":{"volume":"75","issue":"8","first_page":"2687","last_page":"2701"},"is_retracted":false,"is_paratext":false,"is_xpac":false,"primary_topic":{"id":"https://openalex.org/T12238","display_name":"Green IT and Sustainability","score":0.21770000457763672,"subfield":{"id":"https://openalex.org/subfields/2208","display_name":"Electrical and Electronic Engineering"},"field":{"id":"https://openalex.org/fields/22","display_name":"Engineering"},"domain":{"id":"https://openalex.org/domains/3","display_name":"Physical Sciences"}},"topics":[{"id":"https://openalex.org/T12238","display_name":"Green IT and Sustainability","score":0.21770000457763672,"subfield":{"id":"https://openalex.org/subfields/2208","display_name":"Electrical and Electronic Engineering"},"field":{"id":"https://openalex.org/fields/22","display_name":"Engineering"},"domain":{"id":"https://openalex.org/domains/3","display_name":"Physical Sciences"}},{"id":"https://openalex.org/T10273","display_name":"IoT and Edge/Fog Computing","score":0.1889999955892563,"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":0.12849999964237213,"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/energy-consumption","display_name":"Energy consumption","score":0.7961000204086304},{"id":"https://openalex.org/keywords/frequency-scaling","display_name":"Frequency scaling","score":0.7228999733924866},{"id":"https://openalex.org/keywords/inference","display_name":"Inference","score":0.6420999765396118},{"id":"https://openalex.org/keywords/deep-learning","display_name":"Deep learning","score":0.6035000085830688},{"id":"https://openalex.org/keywords/efficient-energy-use","display_name":"Efficient energy use","score":0.5929999947547913},{"id":"https://openalex.org/keywords/reinforcement-learning","display_name":"Reinforcement learning","score":0.5044000148773193},{"id":"https://openalex.org/keywords/energy","display_name":"Energy (signal processing)","score":0.46459999680519104},{"id":"https://openalex.org/keywords/latency","display_name":"Latency (audio)","score":0.45210000872612}],"concepts":[{"id":"https://openalex.org/C41008148","wikidata":"https://www.wikidata.org/wiki/Q21198","display_name":"Computer science","level":0,"score":0.8458999991416931},{"id":"https://openalex.org/C2780165032","wikidata":"https://www.wikidata.org/wiki/Q16869822","display_name":"Energy consumption","level":2,"score":0.7961000204086304},{"id":"https://openalex.org/C157742956","wikidata":"https://www.wikidata.org/wiki/Q3237776","display_name":"Frequency scaling","level":3,"score":0.7228999733924866},{"id":"https://openalex.org/C2776214188","wikidata":"https://www.wikidata.org/wiki/Q408386","display_name":"Inference","level":2,"score":0.6420999765396118},{"id":"https://openalex.org/C108583219","wikidata":"https://www.wikidata.org/wiki/Q197536","display_name":"Deep learning","level":2,"score":0.6035000085830688},{"id":"https://openalex.org/C2742236","wikidata":"https://www.wikidata.org/wiki/Q924713","display_name":"Efficient energy use","level":2,"score":0.5929999947547913},{"id":"https://openalex.org/C97541855","wikidata":"https://www.wikidata.org/wiki/Q830687","display_name":"Reinforcement learning","level":2,"score":0.5044000148773193},{"id":"https://openalex.org/C79403827","wikidata":"https://www.wikidata.org/wiki/Q3988","display_name":"Real-time computing","level":1,"score":0.48030000925064087},{"id":"https://openalex.org/C186370098","wikidata":"https://www.wikidata.org/wiki/Q442787","display_name":"Energy (signal processing)","level":2,"score":0.46459999680519104},{"id":"https://openalex.org/C82876162","wikidata":"https://www.wikidata.org/wiki/Q17096504","display_name":"Latency (audio)","level":2,"score":0.45210000872612},{"id":"https://openalex.org/C75027835","wikidata":"https://www.wikidata.org/wiki/Q1064746","display_name":"Green computing","level":3,"score":0.41130000352859497},{"id":"https://openalex.org/C154945302","wikidata":"https://www.wikidata.org/wiki/Q11660","display_name":"Artificial intelligence","level":1,"score":0.40059998631477356},{"id":"https://openalex.org/C16320812","wikidata":"https://www.wikidata.org/wiki/Q1812200","display_name":"Idle","level":2,"score":0.36059999465942383},{"id":"https://openalex.org/C74912251","wikidata":"https://www.wikidata.org/wiki/Q6815727","display_name":"Memory footprint","level":2,"score":0.3334999978542328},{"id":"https://openalex.org/C149635348","wikidata":"https://www.wikidata.org/wiki/Q193040","display_name":"Embedded system","level":1,"score":0.3287000060081482},{"id":"https://openalex.org/C157764524","wikidata":"https://www.wikidata.org/wiki/Q1383412","display_name":"Throughput","level":3,"score":0.3188000023365021},{"id":"https://openalex.org/C520301825","wikidata":"https://www.wikidata.org/wiki/Q380170","display_name":"Energy conservation","level":2,"score":0.31310001015663147},{"id":"https://openalex.org/C4822641","wikidata":"https://www.wikidata.org/wiki/Q846651","display_name":"Multiprocessing","level":2,"score":0.295199990272522},{"id":"https://openalex.org/C173608175","wikidata":"https://www.wikidata.org/wiki/Q232661","display_name":"Parallel computing","level":1,"score":0.27709999680519104},{"id":"https://openalex.org/C132853815","wikidata":"https://www.wikidata.org/wiki/Q1341254","display_name":"Energy accounting","level":3,"score":0.26170000433921814},{"id":"https://openalex.org/C120314980","wikidata":"https://www.wikidata.org/wiki/Q180634","display_name":"Distributed computing","level":1,"score":0.25920000672340393},{"id":"https://openalex.org/C44154836","wikidata":"https://www.wikidata.org/wiki/Q45045","display_name":"Simulation","level":1,"score":0.2565000057220459},{"id":"https://openalex.org/C2780745134","wikidata":"https://www.wikidata.org/wiki/Q7940751","display_name":"Voltage reduction","level":3,"score":0.25609999895095825},{"id":"https://openalex.org/C30772137","wikidata":"https://www.wikidata.org/wiki/Q5164762","display_name":"Consumption (sociology)","level":2,"score":0.25369998812675476},{"id":"https://openalex.org/C93996380","wikidata":"https://www.wikidata.org/wiki/Q44127","display_name":"Server","level":2,"score":0.2526000142097473},{"id":"https://openalex.org/C206729178","wikidata":"https://www.wikidata.org/wiki/Q2271896","display_name":"Scheduling (production processes)","level":2,"score":0.2506999969482422}],"mesh":[],"locations_count":1,"locations":[{"id":"doi:10.1109/tc.2025.3642827","is_oa":false,"landing_page_url":"https://doi.org/10.1109/tc.2025.3642827","pdf_url":null,"source":{"id":"https://openalex.org/S157670870","display_name":"IEEE Transactions on Computers","issn_l":"0018-9340","issn":["0018-9340","0016-9340","1557-9956","2326-3814"],"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 Computers","raw_type":"journal-article"}],"best_oa_location":null,"sustainable_development_goals":[{"id":"https://metadata.un.org/sdg/7","display_name":"Affordable and clean energy","score":0.8961318731307983}],"awards":[{"id":"https://openalex.org/G1446795574","display_name":null,"funder_award_id":"LH2024F022","funder_id":"https://openalex.org/F4320323085","funder_display_name":"Natural Science Foundation of Heilongjiang Province"},{"id":"https://openalex.org/G420673388","display_name":null,"funder_award_id":"HIT.NSFJG202433","funder_id":"https://openalex.org/F4320335787","funder_display_name":"Fundamental Research Funds for the Central Universities"},{"id":"https://openalex.org/G6382396254","display_name":null,"funder_award_id":"62402141","funder_id":"https://openalex.org/F4320321001","funder_display_name":"National Natural Science Foundation of China"},{"id":"https://openalex.org/G8725492322","display_name":null,"funder_award_id":"62202123","funder_id":"https://openalex.org/F4320321001","funder_display_name":"National Natural Science Foundation of China"}],"funders":[{"id":"https://openalex.org/F4320321001","display_name":"National Natural Science Foundation of China","ror":"https://ror.org/01h0zpd94"},{"id":"https://openalex.org/F4320323085","display_name":"Natural Science Foundation of Heilongjiang Province","ror":null},{"id":"https://openalex.org/F4320335787","display_name":"Fundamental Research Funds for the Central Universities","ror":null}],"has_content":{"grobid_xml":false,"pdf":false},"content_urls":null,"referenced_works_count":0,"referenced_works":[],"related_works":[],"abstract_inverted_index":{"The":[0],"growing":[1],"demand":[2],"for":[3,136],"deploying":[4],"deep":[5,92],"learning":[6,94],"(DL)":[7],"models,":[8],"particularly":[9],"large":[10],"language":[11],"models":[12,99],"(LLMs),":[13],"has":[14],"made":[15],"it":[16],"imperative":[17],"to":[18,42,63,100,106,154,164,178,197,223,249,253,258,274],"optimize":[19,64,101],"GPU":[20,119,125,137,187],"energy":[21,29,77,102,112,138,151,160,193,219,244,271],"consumption":[22,103,139,245,272],"while":[23,104,289],"meeting":[24],"service-level":[25],"objectives":[26],"(SLOs).":[27],"Significant":[28],"use":[30],"and":[31,56,59,75,124,149,170,203,229,255,287],"carbon":[32],"dioxide":[33],"(CO2)":[34],"emissions":[35],"from":[36],"GPU-based":[37],"inference":[38,107,130,237],"tasks":[39,238],"contribute":[40],"substantially":[41],"the":[43,47,155,167,173,179,200,206,210,226,232],"environmental":[44],"footprint":[45],"of":[46,162,195,221,285],"DL":[48,87],"deployment.":[49],"Existing":[50],"approaches":[51],"primarily":[52,279],"rely":[53,278],"on":[54,129,166,172,199,205,225,231,280],"batching":[55,185,281,286],"dynamic":[57,80],"voltage":[58],"frequency":[60,73,127,189],"scaling":[61],"(DVFS)":[62],"service":[65],"performance":[66],"or":[67,282],"throughput,":[68],"but":[69],"often":[70],"overlook":[71],"memory":[72,126],"adjustments":[74],"holistic":[76],"optimization":[78],"under":[79],"workloads.":[81],"This":[82],"study":[83],"presents":[84],"GREENDLS,":[85],"a":[86,283],"serving":[88],"system":[89],"that":[90,266,277],"integrates":[91],"reinforcement":[93],"(DRL)":[95],"with":[96,186],"offline":[97],"prediction":[98],"adhering":[105],"latency":[108],"SLOs,":[109],"achieving":[110],"significant":[111],"savings.":[113],"GREENDLS":[114,158,191,217,241,267],"dynamically":[115],"adjusts":[116],"batch":[117,145],"size,":[118],"streaming":[120],"multiprocessor":[121],"(SM)":[122],"frequency,":[123],"based":[128],"request":[131],"rates.":[132],"It":[133],"also":[134],"accounts":[135],"during":[140],"idle":[141],"phases,":[142],"such":[143],"as":[144],"filling,":[146],"enabling":[147],"multi-parameter":[148],"fine-grained":[150],"optimization.":[152],"Compared":[153],"Clipper":[156],"system,":[157,181],"achieves":[159,192],"savings":[161,194,220],"up":[163,196,222],"45.57%":[165],"RTX":[168,201,227],"3080Ti":[169,202,228],"39.44%":[171],"Tesla":[174,207,233],"V100S.":[175,208,234],"When":[176],"compared":[177,248,252,257,273],"EAIS":[180],"which":[182],"only":[183],"combines":[184],"SM":[188],"adjustment,":[190],"36.44%":[198],"15.44%":[204],"Against":[209],"method":[211],"proposed":[212],"by":[213,246],"Yu":[214,259],"et":[215,260],"al.,":[216],"attains":[218],"46.34%":[224],"33.74%":[230],"In":[235],"LLM":[236],"using":[239],"Qwen,":[240],"reduces":[242],"average":[243],"40.79%":[247],"Clipper,":[250],"10.42%":[251],"EAIS,":[254],"42.42%":[256],"al.":[261],"These":[262],"results":[263],"clearly":[264],"demonstrate":[265],"more":[268],"effectively":[269],"optimizes":[270],"traditional":[275],"methods":[276],"combination":[284],"DVFS,":[288],"still":[290],"ensuring":[291],"SLO":[292],"compliance.":[293]},"counts_by_year":[],"updated_date":"2026-07-11T05:44:25.926202","created_date":"2025-12-12T00:00:00"}