{"id":"https://openalex.org/W4401176820","doi":"https://doi.org/10.1145/3651890.3672239","title":"Crux: GPU-Efficient Communication Scheduling for Deep Learning Training","display_name":"Crux: GPU-Efficient Communication Scheduling for Deep Learning Training","publication_year":2024,"publication_date":"2024-07-31","ids":{"openalex":"https://openalex.org/W4401176820","doi":"https://doi.org/10.1145/3651890.3672239"},"language":"en","primary_location":{"id":"doi:10.1145/3651890.3672239","is_oa":false,"landing_page_url":"https://doi.org/10.1145/3651890.3672239","pdf_url":null,"source":null,"license":null,"license_id":null,"version":"publishedVersion","is_accepted":true,"is_published":true,"raw_source_name":"Proceedings of the ACM SIGCOMM 2024 Conference","raw_type":"proceedings-article"},"type":"conference-paper","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":"https://openalex.org/A5114200761","display_name":"Jiamin Cao","orcid":null},"institutions":[{"id":"https://openalex.org/I45928872","display_name":"Alibaba Group (China)","ror":"https://ror.org/00k642b80","country_code":"CN","type":"company","lineage":["https://openalex.org/I45928872"]}],"countries":["CN"],"is_corresponding":false,"raw_author_name":"Jiamin Cao","raw_affiliation_strings":["Alibaba Cloud, Beijing, China"],"raw_orcid":"https://orcid.org/0009-0009-5468-7366","affiliations":[{"raw_affiliation_string":"Alibaba Cloud, Beijing, China","institution_ids":["https://openalex.org/I45928872"]}]},{"author_position":"middle","author":{"id":"https://openalex.org/A5000348407","display_name":"Yu Guan","orcid":"https://orcid.org/0000-0002-0726-3933"},"institutions":[{"id":"https://openalex.org/I45928872","display_name":"Alibaba Group (China)","ror":"https://ror.org/00k642b80","country_code":"CN","type":"company","lineage":["https://openalex.org/I45928872"]}],"countries":["CN"],"is_corresponding":false,"raw_author_name":"Yu Guan","raw_affiliation_strings":["Alibaba Cloud, Beijing, China"],"raw_orcid":"https://orcid.org/0000-0002-0726-3933","affiliations":[{"raw_affiliation_string":"Alibaba Cloud, Beijing, China","institution_ids":["https://openalex.org/I45928872"]}]},{"author_position":"middle","author":{"id":"https://openalex.org/A5008523628","display_name":"Kun Qian","orcid":"https://orcid.org/0000-0001-9882-9279"},"institutions":[{"id":"https://openalex.org/I45928872","display_name":"Alibaba Group (China)","ror":"https://ror.org/00k642b80","country_code":"CN","type":"company","lineage":["https://openalex.org/I45928872"]}],"countries":["CN"],"is_corresponding":false,"raw_author_name":"Kun Qian","raw_affiliation_strings":["Alibaba Cloud, Hangzhou, China"],"raw_orcid":"https://orcid.org/0000-0001-9882-9279","affiliations":[{"raw_affiliation_string":"Alibaba Cloud, Hangzhou, China","institution_ids":["https://openalex.org/I45928872"]}]},{"author_position":"middle","author":{"id":"https://openalex.org/A5050612924","display_name":"Jiaqi Gao","orcid":"https://orcid.org/0000-0003-3548-2030"},"institutions":[{"id":"https://openalex.org/I45928872","display_name":"Alibaba Group (China)","ror":"https://ror.org/00k642b80","country_code":"CN","type":"company","lineage":["https://openalex.org/I45928872"]}],"countries":["CN"],"is_corresponding":false,"raw_author_name":"Jiaqi Gao","raw_affiliation_strings":["Alibaba Cloud, Beijing, China"],"raw_orcid":"https://orcid.org/0000-0003-3548-2030","affiliations":[{"raw_affiliation_string":"Alibaba Cloud, Beijing, China","institution_ids":["https://openalex.org/I45928872"]}]},{"author_position":"middle","author":{"id":"https://openalex.org/A5086945155","display_name":"Wencong Xiao","orcid":"https://orcid.org/0000-0002-3043-522X"},"institutions":[{"id":"https://openalex.org/I45928872","display_name":"Alibaba Group (China)","ror":"https://ror.org/00k642b80","country_code":"CN","type":"company","lineage":["https://openalex.org/I45928872"]}],"countries":["CN"],"is_corresponding":false,"raw_author_name":"Wencong Xiao","raw_affiliation_strings":["Alibaba Cloud, Hangzhou, China"],"raw_orcid":"https://orcid.org/0000-0002-3043-522X","affiliations":[{"raw_affiliation_string":"Alibaba Cloud, Hangzhou, China","institution_ids":["https://openalex.org/I45928872"]}]},{"author_position":"middle","author":{"id":"https://openalex.org/A5100743206","display_name":"Jianbo Dong","orcid":"https://orcid.org/0000-0003-0939-8943"},"institutions":[{"id":"https://openalex.org/I45928872","display_name":"Alibaba Group (China)","ror":"https://ror.org/00k642b80","country_code":"CN","type":"company","lineage":["https://openalex.org/I45928872"]}],"countries":["CN"],"is_corresponding":false,"raw_author_name":"Jianbo Dong","raw_affiliation_strings":["Alibaba Cloud, Beijing, China"],"raw_orcid":"https://orcid.org/0000-0003-0939-8943","affiliations":[{"raw_affiliation_string":"Alibaba Cloud, Beijing, China","institution_ids":["https://openalex.org/I45928872"]}]},{"author_position":"middle","author":{"id":"https://openalex.org/A5006224078","display_name":"Binzhang Fu","orcid":"https://orcid.org/0009-0008-1213-0554"},"institutions":[{"id":"https://openalex.org/I45928872","display_name":"Alibaba Group (China)","ror":"https://ror.org/00k642b80","country_code":"CN","type":"company","lineage":["https://openalex.org/I45928872"]}],"countries":["CN"],"is_corresponding":false,"raw_author_name":"Binzhang Fu","raw_affiliation_strings":["Alibaba Cloud, Beijing, China"],"raw_orcid":"https://orcid.org/0009-0008-1213-0554","affiliations":[{"raw_affiliation_string":"Alibaba Cloud, Beijing, China","institution_ids":["https://openalex.org/I45928872"]}]},{"author_position":"middle","author":{"id":"https://openalex.org/A5043589586","display_name":"Dennis Cai","orcid":"https://orcid.org/0009-0001-7272-8143"},"institutions":[{"id":"https://openalex.org/I45928872","display_name":"Alibaba Group (China)","ror":"https://ror.org/00k642b80","country_code":"CN","type":"company","lineage":["https://openalex.org/I45928872"]}],"countries":["CN"],"is_corresponding":false,"raw_author_name":"Dennis Cai","raw_affiliation_strings":["Alibaba Cloud, Beijing, China"],"raw_orcid":"https://orcid.org/0009-0001-7272-8143","affiliations":[{"raw_affiliation_string":"Alibaba Cloud, Beijing, China","institution_ids":["https://openalex.org/I45928872"]}]},{"author_position":"last","author":{"id":"https://openalex.org/A5103040133","display_name":"Ennan Zhai","orcid":"https://orcid.org/0000-0003-4352-7497"},"institutions":[{"id":"https://openalex.org/I45928872","display_name":"Alibaba Group (China)","ror":"https://ror.org/00k642b80","country_code":"CN","type":"company","lineage":["https://openalex.org/I45928872"]}],"countries":["CN"],"is_corresponding":false,"raw_author_name":"Ennan Zhai","raw_affiliation_strings":["Alibaba Cloud, Beijing, China"],"raw_orcid":"https://orcid.org/0000-0003-4352-7497","affiliations":[{"raw_affiliation_string":"Alibaba Cloud, Beijing, China","institution_ids":["https://openalex.org/I45928872"]}]}],"institutions":[],"countries_distinct_count":1,"institutions_distinct_count":1,"corresponding_author_ids":[],"corresponding_institution_ids":["https://openalex.org/I45928872"],"apc_list":null,"apc_paid":null,"fwci":null,"has_fulltext":false,"cited_by_count":33,"citation_normalized_percentile":null,"cited_by_percentile_year":null,"biblio":{"volume":null,"issue":null,"first_page":"1","last_page":"15"},"is_retracted":false,"is_paratext":false,"is_xpac":false,"primary_topic":{"id":"https://openalex.org/T10101","display_name":"Cloud Computing and Resource Management","score":0.9990000128746033,"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"}},"topics":[{"id":"https://openalex.org/T10101","display_name":"Cloud Computing and Resource Management","score":0.9990000128746033,"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/T12292","display_name":"Graph Theory and Algorithms","score":0.9987999796867371,"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/T10054","display_name":"Parallel Computing and Optimization Techniques","score":0.9987000226974487,"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":0.8313450217247009},{"id":"https://openalex.org/keywords/scheduling","display_name":"Scheduling (production processes)","score":0.5201323628425598},{"id":"https://openalex.org/keywords/deep-learning","display_name":"Deep learning","score":0.49967360496520996},{"id":"https://openalex.org/keywords/training","display_name":"Training (meteorology)","score":0.4529725909233093},{"id":"https://openalex.org/keywords/artificial-intelligence","display_name":"Artificial intelligence","score":0.4458690583705902},{"id":"https://openalex.org/keywords/parallel-computing","display_name":"Parallel computing","score":0.4188828468322754}],"concepts":[{"id":"https://openalex.org/C41008148","wikidata":"https://www.wikidata.org/wiki/Q21198","display_name":"Computer science","level":0,"score":0.8313450217247009},{"id":"https://openalex.org/C206729178","wikidata":"https://www.wikidata.org/wiki/Q2271896","display_name":"Scheduling (production processes)","level":2,"score":0.5201323628425598},{"id":"https://openalex.org/C108583219","wikidata":"https://www.wikidata.org/wiki/Q197536","display_name":"Deep learning","level":2,"score":0.49967360496520996},{"id":"https://openalex.org/C2777211547","wikidata":"https://www.wikidata.org/wiki/Q17141490","display_name":"Training (meteorology)","level":2,"score":0.4529725909233093},{"id":"https://openalex.org/C154945302","wikidata":"https://www.wikidata.org/wiki/Q11660","display_name":"Artificial intelligence","level":1,"score":0.4458690583705902},{"id":"https://openalex.org/C173608175","wikidata":"https://www.wikidata.org/wiki/Q232661","display_name":"Parallel computing","level":1,"score":0.4188828468322754},{"id":"https://openalex.org/C162324750","wikidata":"https://www.wikidata.org/wiki/Q8134","display_name":"Economics","level":0,"score":0.0},{"id":"https://openalex.org/C153294291","wikidata":"https://www.wikidata.org/wiki/Q25261","display_name":"Meteorology","level":1,"score":0.0},{"id":"https://openalex.org/C121332964","wikidata":"https://www.wikidata.org/wiki/Q413","display_name":"Physics","level":0,"score":0.0},{"id":"https://openalex.org/C21547014","wikidata":"https://www.wikidata.org/wiki/Q1423657","display_name":"Operations management","level":1,"score":0.0}],"mesh":[],"locations_count":1,"locations":[{"id":"doi:10.1145/3651890.3672239","is_oa":false,"landing_page_url":"https://doi.org/10.1145/3651890.3672239","pdf_url":null,"source":null,"license":null,"license_id":null,"version":"publishedVersion","is_accepted":true,"is_published":true,"raw_source_name":"Proceedings of the ACM SIGCOMM 2024 Conference","raw_type":"proceedings-article"}],"best_oa_location":null,"sustainable_development_goals":[],"awards":[],"funders":[],"has_content":{"pdf":false,"grobid_xml":false},"content_urls":null,"referenced_works_count":29,"referenced_works":["https://openalex.org/W95608104","https://openalex.org/W2009319645","https://openalex.org/W2013932952","https://openalex.org/W2057332538","https://openalex.org/W2059300917","https://openalex.org/W2091520147","https://openalex.org/W2096237581","https://openalex.org/W2108229655","https://openalex.org/W2110514750","https://openalex.org/W2133569115","https://openalex.org/W2149939304","https://openalex.org/W2157614013","https://openalex.org/W2168427242","https://openalex.org/W2323909431","https://openalex.org/W2505830093","https://openalex.org/W2824896242","https://openalex.org/W2963596039","https://openalex.org/W2968108410","https://openalex.org/W2975712713","https://openalex.org/W3016395792","https://openalex.org/W3042338533","https://openalex.org/W3046750740","https://openalex.org/W3097875570","https://openalex.org/W3175449831","https://openalex.org/W4281657584","https://openalex.org/W4290991121","https://openalex.org/W4308995136","https://openalex.org/W4308995365","https://openalex.org/W4385966327"],"related_works":["https://openalex.org/W2731899572","https://openalex.org/W3215138031","https://openalex.org/W3009238340","https://openalex.org/W4321369474","https://openalex.org/W4360585206","https://openalex.org/W4285208911","https://openalex.org/W3082895349","https://openalex.org/W4213079790","https://openalex.org/W2248239756","https://openalex.org/W4323565446"],"abstract_inverted_index":{"Deep":[0],"learning":[1],"training":[2,51],"(DLT),":[3],"e.g.,":[4],"large":[5],"language":[6],"model":[7],"(LLM)":[8],"training,":[9],"has":[10],"become":[11],"one":[12],"of":[13,49],"the":[14,39,46,50,71,110],"most":[15],"important":[16],"services":[17],"in":[18,45],"multitenant":[19],"cloud":[20],"computing.":[21],"By":[22],"deeply":[23],"studying":[24],"in-production":[25],"DLT":[26,35,75,111],"jobs,":[27],"we":[28,56,87,104],"observed":[29],"that":[30,62,108,127,143],"communication":[31,60,72,101,120],"contention":[32,73],"among":[33,74],"different":[34],"jobs":[36],"seriously":[37],"influences":[38],"overall":[40],"GPU":[41,66,78,99,115,133,146],"computation":[42,67,79,116,134,147],"utilization,":[43],"resulting":[44],"low":[47],"efficiency":[48],"cluster.":[52],"In":[53],"this":[54,96],"paper,":[55],"present":[57],"Crux,":[58],"a":[59,91],"scheduler":[61],"aims":[63],"to":[64,94,131,151],"maximize":[65],"utilization":[68,80,148],"by":[69,98,149],"mitigating":[70],"jobs.":[76],"Maximizing":[77],"for":[81],"DLT,":[82],"nevertheless,":[83],"is":[84],"NP-Complete;":[85],"thus,":[86],"formulate":[88],"and":[89,159],"prove":[90],"novel":[92],"theorem":[93],"approach":[95,107],"goal":[97],"intensity-aware":[100],"scheduling.":[102],"Then,":[103],"propose":[105],"an":[106],"prioritizes":[109],"flows":[112],"with":[113,154],"high":[114],"intensity,":[117],"reducing":[118],"potential":[119],"contention.":[121],"Our":[122],"96-GPU":[123],"testbed":[124],"experiments":[125],"show":[126],"Crux":[128,144],"improves":[129],"8.3%":[130],"14.8%":[132],"utilization.":[135],"The":[136],"large-scale":[137],"production":[138],"trace-based":[139],"simulation":[140],"further":[141],"shows":[142],"increases":[145],"up":[150],"23%":[152],"compared":[153],"alternatives":[155],"including":[156],"Sincronia,":[157],"TACCL,":[158],"CASSINI.":[160]},"counts_by_year":[{"year":2026,"cited_by_count":12},{"year":2025,"cited_by_count":21}],"updated_date":"2026-07-17T09:13:05.818461","created_date":"2025-10-10T00:00:00"}