{"id":"https://openalex.org/W7123919998","doi":"https://doi.org/10.1145/3772052.3772242","title":"Defragmentation Scheduling with Deep Reinforcement Learning in Shared GPU Clusters","display_name":"Defragmentation Scheduling with Deep Reinforcement Learning in Shared GPU Clusters","publication_year":2025,"publication_date":"2025-11-19","ids":{"openalex":"https://openalex.org/W7123919998","doi":"https://doi.org/10.1145/3772052.3772242"},"language":null,"primary_location":{"id":"doi:10.1145/3772052.3772242","is_oa":false,"landing_page_url":"https://doi.org/10.1145/3772052.3772242","pdf_url":null,"source":null,"license":null,"license_id":null,"version":"publishedVersion","is_accepted":true,"is_published":true,"raw_source_name":"Proceedings of the 2025 ACM Symposium on Cloud Computing","raw_type":"proceedings-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":"Qingfu Wu","orcid":"https://orcid.org/0009-0008-9767-8366"},"institutions":[{"id":"https://openalex.org/I157773358","display_name":"Sun Yat-sen University","ror":"https://ror.org/0064kty71","country_code":"CN","type":"education","lineage":["https://openalex.org/I157773358"]}],"countries":["CN"],"is_corresponding":true,"raw_author_name":"Qingfu Wu","raw_affiliation_strings":["Sun Yat-sen University, Guangzhou, Guangdong, China"],"affiliations":[{"raw_affiliation_string":"Sun Yat-sen University, Guangzhou, Guangdong, China","institution_ids":["https://openalex.org/I157773358"]}]},{"author_position":"middle","author":{"id":"https://openalex.org/A5122923202","display_name":"Pengfei Chen","orcid":null},"institutions":[{"id":"https://openalex.org/I157773358","display_name":"Sun Yat-sen University","ror":"https://ror.org/0064kty71","country_code":"CN","type":"education","lineage":["https://openalex.org/I157773358"]}],"countries":["CN"],"is_corresponding":false,"raw_author_name":"Pengfei Chen","raw_affiliation_strings":["Sun Yat-sen University, Guangzhou, Guangdong, China"],"affiliations":[{"raw_affiliation_string":"Sun Yat-sen University, Guangzhou, Guangdong, China","institution_ids":["https://openalex.org/I157773358"]}]},{"author_position":"last","author":{"id":"https://openalex.org/A5122910822","display_name":"Yilun Wang","orcid":null},"institutions":[{"id":"https://openalex.org/I157773358","display_name":"Sun Yat-sen University","ror":"https://ror.org/0064kty71","country_code":"CN","type":"education","lineage":["https://openalex.org/I157773358"]}],"countries":["CN"],"is_corresponding":false,"raw_author_name":"Yilun Wang","raw_affiliation_strings":["Sun Yat-sen University, Guangzhou, Guangdong, China"],"affiliations":[{"raw_affiliation_string":"Sun Yat-sen University, Guangzhou, Guangdong, China","institution_ids":["https://openalex.org/I157773358"]}]}],"institutions":[],"countries_distinct_count":1,"institutions_distinct_count":3,"corresponding_author_ids":[],"corresponding_institution_ids":["https://openalex.org/I157773358"],"apc_list":null,"apc_paid":null,"fwci":0.0,"has_fulltext":false,"cited_by_count":0,"citation_normalized_percentile":{"value":0.68839579,"is_in_top_1_percent":false,"is_in_top_10_percent":false},"cited_by_percentile_year":null,"biblio":{"volume":null,"issue":null,"first_page":"402","last_page":"415"},"is_retracted":false,"is_paratext":false,"is_xpac":false,"primary_topic":{"id":"https://openalex.org/T10715","display_name":"Distributed and Parallel Computing Systems","score":0.5218999981880188,"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"}},"topics":[{"id":"https://openalex.org/T10715","display_name":"Distributed and Parallel Computing Systems","score":0.5218999981880188,"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/T10101","display_name":"Cloud Computing and Resource Management","score":0.30079999566078186,"subfield":{"id":"https://openalex.org/subfields/1710","display_name":"Information Systems"},"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":0.06159999966621399,"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/reinforcement-learning","display_name":"Reinforcement learning","score":0.7390000224113464},{"id":"https://openalex.org/keywords/scheduling","display_name":"Scheduling (production processes)","score":0.6571999788284302},{"id":"https://openalex.org/keywords/testbed","display_name":"Testbed","score":0.5803999900817871},{"id":"https://openalex.org/keywords/asynchronous-communication","display_name":"Asynchronous communication","score":0.4526999890804291},{"id":"https://openalex.org/keywords/job-shop-scheduling","display_name":"Job shop scheduling","score":0.4203999936580658},{"id":"https://openalex.org/keywords/quality-of-service","display_name":"Quality of service","score":0.3799999952316284},{"id":"https://openalex.org/keywords/gpu-cluster","display_name":"GPU cluster","score":0.3643999993801117},{"id":"https://openalex.org/keywords/fragmentation","display_name":"Fragmentation (computing)","score":0.35510000586509705}],"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/C97541855","wikidata":"https://www.wikidata.org/wiki/Q830687","display_name":"Reinforcement learning","level":2,"score":0.7390000224113464},{"id":"https://openalex.org/C206729178","wikidata":"https://www.wikidata.org/wiki/Q2271896","display_name":"Scheduling (production processes)","level":2,"score":0.6571999788284302},{"id":"https://openalex.org/C31395832","wikidata":"https://www.wikidata.org/wiki/Q1318674","display_name":"Testbed","level":2,"score":0.5803999900817871},{"id":"https://openalex.org/C120314980","wikidata":"https://www.wikidata.org/wiki/Q180634","display_name":"Distributed computing","level":1,"score":0.5092999935150146},{"id":"https://openalex.org/C151319957","wikidata":"https://www.wikidata.org/wiki/Q752739","display_name":"Asynchronous communication","level":2,"score":0.4526999890804291},{"id":"https://openalex.org/C55416958","wikidata":"https://www.wikidata.org/wiki/Q6206757","display_name":"Job shop scheduling","level":3,"score":0.4203999936580658},{"id":"https://openalex.org/C5119721","wikidata":"https://www.wikidata.org/wiki/Q220501","display_name":"Quality of service","level":2,"score":0.3799999952316284},{"id":"https://openalex.org/C2781335571","wikidata":"https://www.wikidata.org/wiki/Q2633544","display_name":"GPU cluster","level":3,"score":0.3643999993801117},{"id":"https://openalex.org/C191015642","wikidata":"https://www.wikidata.org/wiki/Q1132459","display_name":"Fragmentation (computing)","level":2,"score":0.35510000586509705},{"id":"https://openalex.org/C173608175","wikidata":"https://www.wikidata.org/wiki/Q232661","display_name":"Parallel computing","level":1,"score":0.3544999957084656},{"id":"https://openalex.org/C2984822820","wikidata":"https://www.wikidata.org/wiki/Q1123036","display_name":"Processor scheduling","level":3,"score":0.328900009393692},{"id":"https://openalex.org/C173801870","wikidata":"https://www.wikidata.org/wiki/Q201413","display_name":"Heuristic","level":2,"score":0.3027999997138977},{"id":"https://openalex.org/C83283714","wikidata":"https://www.wikidata.org/wiki/Q121117","display_name":"Supercomputer","level":2,"score":0.290800005197525},{"id":"https://openalex.org/C2778119891","wikidata":"https://www.wikidata.org/wiki/Q477690","display_name":"CUDA","level":2,"score":0.28119999170303345},{"id":"https://openalex.org/C82876162","wikidata":"https://www.wikidata.org/wiki/Q17096504","display_name":"Latency (audio)","level":2,"score":0.2752000093460083},{"id":"https://openalex.org/C50630238","wikidata":"https://www.wikidata.org/wiki/Q971505","display_name":"General-purpose computing on graphics processing units","level":3,"score":0.27160000801086426},{"id":"https://openalex.org/C177212765","wikidata":"https://www.wikidata.org/wiki/Q627335","display_name":"Workflow","level":2,"score":0.27129998803138733},{"id":"https://openalex.org/C78766204","wikidata":"https://www.wikidata.org/wiki/Q555032","display_name":"Multi-core processor","level":2,"score":0.26660001277923584},{"id":"https://openalex.org/C2989134064","wikidata":"https://www.wikidata.org/wiki/Q288510","display_name":"Execution time","level":2,"score":0.26089999079704285},{"id":"https://openalex.org/C79403827","wikidata":"https://www.wikidata.org/wiki/Q3988","display_name":"Real-time computing","level":1,"score":0.2597000002861023},{"id":"https://openalex.org/C154945302","wikidata":"https://www.wikidata.org/wiki/Q11660","display_name":"Artificial intelligence","level":1,"score":0.257099986076355},{"id":"https://openalex.org/C108583219","wikidata":"https://www.wikidata.org/wiki/Q197536","display_name":"Deep learning","level":2,"score":0.25290000438690186},{"id":"https://openalex.org/C107568181","wikidata":"https://www.wikidata.org/wiki/Q5319000","display_name":"Dynamic priority scheduling","level":3,"score":0.2522999942302704}],"mesh":[],"locations_count":1,"locations":[{"id":"doi:10.1145/3772052.3772242","is_oa":false,"landing_page_url":"https://doi.org/10.1145/3772052.3772242","pdf_url":null,"source":null,"license":null,"license_id":null,"version":"publishedVersion","is_accepted":true,"is_published":true,"raw_source_name":"Proceedings of the 2025 ACM Symposium on Cloud Computing","raw_type":"proceedings-article"}],"best_oa_location":null,"sustainable_development_goals":[{"score":0.6826539039611816,"id":"https://metadata.un.org/sdg/8","display_name":"Decent work and economic growth"}],"awards":[{"id":"https://openalex.org/G2477268016","display_name":null,"funder_award_id":"2024YFB4505904","funder_id":"https://openalex.org/F4320335777","funder_display_name":"National Key Research and Development Program of China"},{"id":"https://openalex.org/G7403938909","display_name":null,"funder_award_id":"62272495","funder_id":"https://openalex.org/F4320334062","funder_display_name":"National Natural Science Foundation of China-Liaoning Joint Fund"}],"funders":[{"id":"https://openalex.org/F4320334062","display_name":"National Natural Science Foundation of China-Liaoning Joint Fund","ror":null},{"id":"https://openalex.org/F4320335777","display_name":"National Key Research and Development Program of China","ror":null}],"has_content":{"grobid_xml":false,"pdf":false},"content_urls":null,"referenced_works_count":11,"referenced_works":["https://openalex.org/W3121689374","https://openalex.org/W3162118826","https://openalex.org/W3197816522","https://openalex.org/W4205916020","https://openalex.org/W4206174365","https://openalex.org/W4224313989","https://openalex.org/W4318541537","https://openalex.org/W4372262787","https://openalex.org/W4387938003","https://openalex.org/W4403537006","https://openalex.org/W4408867487"],"related_works":[],"abstract_inverted_index":{"Modern":[0],"GPU":[1,30,37,59,91,101,125],"clusters":[2,118],"in":[3,9,32,97],"computing":[4],"centers":[5],"face":[6],"significant":[7],"challenges":[8],"resource":[10],"utilization":[11,102],"due":[12],"to":[13,28,57,73,89,131],"fragmentation":[14,31,126],"caused":[15],"by":[16,80,103,128],"GPU-sharing":[17],"mechanisms,":[18],"job":[19,23],"diversity":[20],"and":[21,55,77,87,115,140],"asynchronous":[22],"lifecycles.":[24],"Existing":[25],"methods":[26],"fail":[27],"address":[29],"dynamic":[33],"scheduling":[34],"scenarios":[35],"under":[36],"sharing.":[38],"To":[39],"tackle":[40],"this":[41,43],"issue,":[42],"paper":[44],"proposes":[45],"DRR,":[46],"a":[47,63],"defragmentation":[48],"scheduler":[49],"with":[50],"deep":[51],"reinforcement":[52],"learning":[53,69],"(DRL)":[54],"rescheduling":[56,95],"mitigate":[58],"fragmentation.":[60,92],"DRR":[61,98,121],"employs":[62],"DRL":[64],"agent":[65],"trained":[66],"via":[67],"imitation":[68],"from":[70],"heuristic":[71],"algorithms":[72],"overcome":[74],"cold-start":[75],"issues,":[76],"further":[78,99],"enhanced":[79],"multi-scale":[81],"policy":[82],"optimization":[83],"for":[84,143],"balanced":[85],"exploration":[86],"exploitation":[88],"reduce":[90],"Additionally,":[93],"the":[94,105,111,123],"strategy":[96],"optimizes":[100],"relocating":[104],"running":[106],"jobs.":[107],"Evaluations":[108],"conducted":[109],"on":[110],"physical":[112],"Kubernetes-based":[113],"testbed":[114],"large-scale":[116],"simulated":[117],"demonstrate":[119],"that":[120],"reduces":[122],"average":[124],"rate":[127],"50%":[129],"compared":[130],"state-of-the-art":[132],"methods,":[133],"while":[134],"maintaining":[135],"Quality":[136],"of":[137],"Service":[138],"(QoS)":[139],"ensuring":[141],"fairness":[142],"users.":[144]},"counts_by_year":[],"updated_date":"2026-04-09T08:11:56.329763","created_date":"2026-01-14T00:00:00"}