{"id":"https://openalex.org/W7165358792","doi":"https://doi.org/10.1145/3765611.3815434","title":"LLM-GreenServe: Optimized Power Consumption with Power-aware Prefill-Decoding Scheduling","display_name":"LLM-GreenServe: Optimized Power Consumption with Power-aware Prefill-Decoding Scheduling","publication_year":2026,"publication_date":"2026-06-20","ids":{"openalex":"https://openalex.org/W7165358792","doi":"https://doi.org/10.1145/3765611.3815434"},"language":null,"primary_location":{"id":"doi:10.1145/3765611.3815434","is_oa":true,"landing_page_url":"https://doi.org/10.1145/3765611.3815434","pdf_url":null,"source":null,"license":"cc-by-nc-nd","license_id":"https://openalex.org/licenses/cc-by-nc-nd","version":"publishedVersion","is_accepted":true,"is_published":true,"raw_source_name":"Proceedings of the 2026 ACM Sustainability Week","raw_type":"proceedings-article"},"type":"conference-paper","indexed_in":["crossref"],"open_access":{"is_oa":true,"oa_status":"gold","oa_url":"https://doi.org/10.1145/3765611.3815434","any_repository_has_fulltext":null},"authorships":[{"author_position":"first","author":{"id":"https://openalex.org/A5139021393","display_name":"Zhihua Cui","orcid":"https://orcid.org/0009-0005-6092-6053"},"institutions":[{"id":"https://openalex.org/I99065089","display_name":"Tsinghua University","ror":"https://ror.org/03cve4549","country_code":"CN","type":"education","lineage":["https://openalex.org/I99065089"]}],"countries":["CN"],"is_corresponding":false,"raw_author_name":"Zhihua Cui","raw_affiliation_strings":["Tsinghua University, Beijing, Bejing, China"],"raw_orcid":"https://orcid.org/0009-0005-6092-6053","affiliations":[{"raw_affiliation_string":"Tsinghua University, Beijing, Bejing, China","institution_ids":["https://openalex.org/I99065089"]}]},{"author_position":"middle","author":{"id":"https://openalex.org/A5065859286","display_name":"Hui Li","orcid":"https://orcid.org/0000-0001-9198-3951"},"institutions":[{"id":"https://openalex.org/I99065089","display_name":"Tsinghua University","ror":"https://ror.org/03cve4549","country_code":"CN","type":"education","lineage":["https://openalex.org/I99065089"]}],"countries":["CN"],"is_corresponding":false,"raw_author_name":"Hanglong Li","raw_affiliation_strings":["Tsinghua University, Beijing, Bejing, China"],"raw_orcid":"https://orcid.org/0009-0005-6828-8690","affiliations":[{"raw_affiliation_string":"Tsinghua University, Beijing, Bejing, China","institution_ids":["https://openalex.org/I99065089"]}]},{"author_position":"middle","author":{"id":"https://openalex.org/A5100601934","display_name":"Zhen Chen","orcid":"https://orcid.org/0000-0001-5503-2630"},"institutions":[{"id":"https://openalex.org/I99065089","display_name":"Tsinghua University","ror":"https://ror.org/03cve4549","country_code":"CN","type":"education","lineage":["https://openalex.org/I99065089"]}],"countries":["CN"],"is_corresponding":false,"raw_author_name":"Zhen Chen","raw_affiliation_strings":["Tsinghua University, Beijing, Bejing, China"],"raw_orcid":"https://orcid.org/0000-0001-5503-2630","affiliations":[{"raw_affiliation_string":"Tsinghua University, Beijing, Bejing, China","institution_ids":["https://openalex.org/I99065089"]}]},{"author_position":"middle","author":{"id":"https://openalex.org/A5138992730","display_name":"Shuoyang Sun","orcid":"https://orcid.org/0009-0000-6263-6723"},"institutions":[{"id":"https://openalex.org/I99065089","display_name":"Tsinghua University","ror":"https://ror.org/03cve4549","country_code":"CN","type":"education","lineage":["https://openalex.org/I99065089"]}],"countries":["CN"],"is_corresponding":false,"raw_author_name":"Shuoyang Sun","raw_affiliation_strings":["Tsinghua University, Beijing, Bejing, China"],"raw_orcid":"https://orcid.org/0009-0000-6263-6723","affiliations":[{"raw_affiliation_string":"Tsinghua University, Beijing, Bejing, China","institution_ids":["https://openalex.org/I99065089"]}]},{"author_position":"middle","author":{"id":"https://openalex.org/A5138991821","display_name":"Pinze Ren","orcid":"https://orcid.org/0009-0006-3090-7596"},"institutions":[{"id":"https://openalex.org/I99065089","display_name":"Tsinghua University","ror":"https://ror.org/03cve4549","country_code":"CN","type":"education","lineage":["https://openalex.org/I99065089"]}],"countries":["CN"],"is_corresponding":false,"raw_author_name":"Pinze Ren","raw_affiliation_strings":["Tsinghua University, Beijing, Bejing, China"],"raw_orcid":"https://orcid.org/0009-0006-3090-7596","affiliations":[{"raw_affiliation_string":"Tsinghua University, Beijing, Bejing, China","institution_ids":["https://openalex.org/I99065089"]}]},{"author_position":"middle","author":{"id":"https://openalex.org/A5138964606","display_name":"Weiran Lin","orcid":"https://orcid.org/0009-0000-8408-8940"},"institutions":[{"id":"https://openalex.org/I99065089","display_name":"Tsinghua University","ror":"https://ror.org/03cve4549","country_code":"CN","type":"education","lineage":["https://openalex.org/I99065089"]}],"countries":["CN"],"is_corresponding":false,"raw_author_name":"Weiran Lin","raw_affiliation_strings":["Tsinghua University, Beijing, Bejing, China"],"raw_orcid":"https://orcid.org/0009-0000-8408-8940","affiliations":[{"raw_affiliation_string":"Tsinghua University, Beijing, Bejing, China","institution_ids":["https://openalex.org/I99065089"]}]},{"author_position":"middle","author":{"id":"https://openalex.org/A5063073383","display_name":"Yinjun Wu","orcid":"https://orcid.org/0000-0002-9770-5765"},"institutions":[{"id":"https://openalex.org/I20231570","display_name":"Peking University","ror":"https://ror.org/02v51f717","country_code":"CN","type":"education","lineage":["https://openalex.org/I20231570"]}],"countries":["CN"],"is_corresponding":false,"raw_author_name":"Yinjun Wu","raw_affiliation_strings":["Peking University, Beijing, Bejing, China"],"raw_orcid":"https://orcid.org/0000-0002-9770-5765","affiliations":[{"raw_affiliation_string":"Peking University, Beijing, Bejing, China","institution_ids":["https://openalex.org/I20231570"]}]},{"author_position":"middle","author":{"id":"https://openalex.org/A5066296581","display_name":"C Li","orcid":null},"institutions":[{"id":"https://openalex.org/I99065089","display_name":"Tsinghua University","ror":"https://ror.org/03cve4549","country_code":"CN","type":"education","lineage":["https://openalex.org/I99065089"]}],"countries":["CN"],"is_corresponding":false,"raw_author_name":"Chao Li","raw_affiliation_strings":["Tsinghua University, Beijing, Bejing, China"],"raw_orcid":"https://orcid.org/0000-0003-0204-1749","affiliations":[{"raw_affiliation_string":"Tsinghua University, Beijing, Bejing, China","institution_ids":["https://openalex.org/I99065089"]}]},{"author_position":"last","author":{"id":"https://openalex.org/A5078133609","display_name":"Shuangshou Li","orcid":null},"institutions":[{"id":"https://openalex.org/I99065089","display_name":"Tsinghua University","ror":"https://ror.org/03cve4549","country_code":"CN","type":"education","lineage":["https://openalex.org/I99065089"]}],"countries":["CN"],"is_corresponding":false,"raw_author_name":"Shuangshou Li","raw_affiliation_strings":["Tsinghua University, Beijing, Bejing, China"],"raw_orcid":"https://orcid.org/0000-0003-1692-4657","affiliations":[{"raw_affiliation_string":"Tsinghua University, Beijing, Bejing, China","institution_ids":["https://openalex.org/I99065089"]}]}],"institutions":[],"countries_distinct_count":1,"institutions_distinct_count":2,"corresponding_author_ids":[],"corresponding_institution_ids":[],"apc_list":null,"apc_paid":null,"fwci":0.0,"has_fulltext":false,"cited_by_count":0,"citation_normalized_percentile":{"value":0.78599152,"is_in_top_1_percent":false,"is_in_top_10_percent":false},"cited_by_percentile_year":null,"biblio":{"volume":null,"issue":null,"first_page":"309","last_page":"314"},"is_retracted":false,"is_paratext":false,"is_xpac":false,"primary_topic":{"id":"https://openalex.org/T10363","display_name":"Low-power high-performance VLSI design","score":0.23960000276565552,"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/T10363","display_name":"Low-power high-performance VLSI design","score":0.23960000276565552,"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/T10847","display_name":"Advanced Optical Network Technologies","score":0.09830000251531601,"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/T10829","display_name":"Interconnection Networks and Systems","score":0.042100001126527786,"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/scheduling","display_name":"Scheduling (production processes)","score":0.5443999767303467},{"id":"https://openalex.org/keywords/power-consumption","display_name":"Power consumption","score":0.5020999908447266},{"id":"https://openalex.org/keywords/job-shop-scheduling","display_name":"Job shop scheduling","score":0.3700999915599823},{"id":"https://openalex.org/keywords/energy-consumption","display_name":"Energy consumption","score":0.35499998927116394},{"id":"https://openalex.org/keywords/power","display_name":"Power (physics)","score":0.3483999967575073}],"concepts":[{"id":"https://openalex.org/C41008148","wikidata":"https://www.wikidata.org/wiki/Q21198","display_name":"Computer science","level":0,"score":0.6031000018119812},{"id":"https://openalex.org/C206729178","wikidata":"https://www.wikidata.org/wiki/Q2271896","display_name":"Scheduling (production processes)","level":2,"score":0.5443999767303467},{"id":"https://openalex.org/C2984118289","wikidata":"https://www.wikidata.org/wiki/Q29954","display_name":"Power consumption","level":3,"score":0.5020999908447266},{"id":"https://openalex.org/C126255220","wikidata":"https://www.wikidata.org/wiki/Q141495","display_name":"Mathematical optimization","level":1,"score":0.38749998807907104},{"id":"https://openalex.org/C55416958","wikidata":"https://www.wikidata.org/wiki/Q6206757","display_name":"Job shop scheduling","level":3,"score":0.3700999915599823},{"id":"https://openalex.org/C2780165032","wikidata":"https://www.wikidata.org/wiki/Q16869822","display_name":"Energy consumption","level":2,"score":0.35499998927116394},{"id":"https://openalex.org/C79403827","wikidata":"https://www.wikidata.org/wiki/Q3988","display_name":"Real-time computing","level":1,"score":0.3537999987602234},{"id":"https://openalex.org/C163258240","wikidata":"https://www.wikidata.org/wiki/Q25342","display_name":"Power (physics)","level":2,"score":0.3483999967575073},{"id":"https://openalex.org/C30772137","wikidata":"https://www.wikidata.org/wiki/Q5164762","display_name":"Consumption (sociology)","level":2,"score":0.2881999909877777},{"id":"https://openalex.org/C200601418","wikidata":"https://www.wikidata.org/wiki/Q2193887","display_name":"Reliability engineering","level":1,"score":0.26030001044273376}],"mesh":[],"locations_count":1,"locations":[{"id":"doi:10.1145/3765611.3815434","is_oa":true,"landing_page_url":"https://doi.org/10.1145/3765611.3815434","pdf_url":null,"source":null,"license":"cc-by-nc-nd","license_id":"https://openalex.org/licenses/cc-by-nc-nd","version":"publishedVersion","is_accepted":true,"is_published":true,"raw_source_name":"Proceedings of the 2026 ACM Sustainability Week","raw_type":"proceedings-article"}],"best_oa_location":{"id":"doi:10.1145/3765611.3815434","is_oa":true,"landing_page_url":"https://doi.org/10.1145/3765611.3815434","pdf_url":null,"source":null,"license":"cc-by-nc-nd","license_id":"https://openalex.org/licenses/cc-by-nc-nd","version":"publishedVersion","is_accepted":true,"is_published":true,"raw_source_name":"Proceedings of the 2026 ACM Sustainability Week","raw_type":"proceedings-article"},"sustainable_development_goals":[{"id":"https://metadata.un.org/sdg/7","score":0.7630156874656677,"display_name":"Affordable and clean energy"}],"awards":[{"id":"https://openalex.org/G3410076442","display_name":null,"funder_award_id":"No.58302201426","funder_id":"https://openalex.org/F4320322392","funder_display_name":"Tsinghua University"}],"funders":[{"id":"https://openalex.org/F4320322392","display_name":"Tsinghua University","ror":"https://ror.org/03cve4549"}],"has_content":{"pdf":false,"grobid_xml":false},"content_urls":null,"referenced_works_count":7,"referenced_works":["https://openalex.org/W2978041899","https://openalex.org/W4387321091","https://openalex.org/W4388041083","https://openalex.org/W4406460939","https://openalex.org/W4412889635","https://openalex.org/W7133227460","https://openalex.org/W7133233472"],"related_works":[],"abstract_inverted_index":{"Large":[0],"language":[1],"model":[2,117],"(LLM)":[3],"serving":[4,20,53],"imposes":[5],"substantial":[6],"GPU":[7,82,106,140],"energy":[8,33,107],"costs,":[9],"making":[10],"energy-efficient":[11],"inference":[12],"an":[13],"increasingly":[14],"important":[15],"systems":[16],"problem.":[17],"Existing":[18],"LLM":[19,150],"engines,":[21],"such":[22],"as":[23],"vLLM,":[24],"primarily":[25],"optimize":[26],"throughput":[27],"and":[28,32,76,94,119,121,137],"latency,":[29],"while":[30],"power":[31,83,133],"are":[34],"not":[35],"first-class":[36],"scheduling":[37,50,134],"objectives.":[38],"This":[39,142],"paper":[40],"presents":[41],"LLM-GreenServe,":[42],"a":[43,64,144],"vLLM-based":[44],"prototype":[45],"that":[46,131],"integrates":[47],"power-aware":[48],"prefill\u2013decode":[49],"into":[51],"the":[52,70],"loop.":[54],"Its":[55],"core":[56],"policy,":[57],"Power-Aware":[58],"Dynamic":[59],"Prefill\u2013Decode":[60],"Scheduling":[61],"(PAPDS),":[62],"uses":[63],"PID":[65],"feedback":[66],"controller":[67,123],"to":[68,110],"adjust":[69],"token":[71],"budget":[72],"allocation":[73],"between":[74],"prefill":[75],"decode":[77],"tasks":[78],"based":[79],"on":[80,88],"device-level":[81],"measurements.":[84],"We":[85],"evaluate":[86],"PAPDS":[87],"Qwen3":[89],"models":[90],"with":[91,116],"8B,":[92],"14B,":[93],"32B":[95],"parameters":[96],"across":[97],"multiple":[98],"workload":[99],"types.":[100],"The":[101],"best-performing":[102],"configuration":[103],"reduces":[104,138],"active":[105,139],"by":[108],"up":[109],"17.24%.":[111],"While":[112],"effectiveness":[113],"scales":[114],"non-linearly":[115],"size":[118],"workload,":[120],"current":[122],"implementations":[124],"introduce":[125],"total-energy":[126],"overhead,":[127],"our":[128],"results":[129],"demonstrate":[130],"closed-loop":[132],"effectively":[135],"isolates":[136],"energy.":[141],"establishes":[143],"foundational,":[145],"software-only":[146],"paradigm":[147],"for":[148],"sustainable":[149],"deployment,":[151],"guiding":[152],"future":[153],"work":[154],"toward":[155],"ultra-lightweight":[156],"adaptive":[157],"control.":[158]},"counts_by_year":[],"updated_date":"2026-07-29T14:22:42.915294","created_date":"2026-06-20T00:00:00"}