{"id":"https://openalex.org/W4412703834","doi":"https://doi.org/10.1145/3696630.3728529","title":"dl\u00b2: Detecting Communication Deadlocks in Deep Learning Jobs","display_name":"dl\u00b2: Detecting Communication Deadlocks in Deep Learning Jobs","publication_year":2025,"publication_date":"2025-06-23","ids":{"openalex":"https://openalex.org/W4412703834","doi":"https://doi.org/10.1145/3696630.3728529"},"language":"en","primary_location":{"id":"doi:10.1145/3696630.3728529","is_oa":true,"landing_page_url":"https://doi.org/10.1145/3696630.3728529","pdf_url":"https://dl.acm.org/doi/pdf/10.1145/3696630.3728529","source":null,"license":"cc-by-nc","license_id":"https://openalex.org/licenses/cc-by-nc","version":"publishedVersion","is_accepted":true,"is_published":true,"raw_source_name":"Proceedings of the 33rd ACM International Conference on the Foundations of Software Engineering","raw_type":"proceedings-article"},"type":"conference-paper","indexed_in":["crossref"],"open_access":{"is_oa":true,"oa_status":"gold","oa_url":"https://dl.acm.org/doi/pdf/10.1145/3696630.3728529","any_repository_has_fulltext":null},"authorships":[{"author_position":"first","author":{"id":"https://openalex.org/A5031774552","display_name":"Yanjie Gao","orcid":"https://orcid.org/0000-0003-1899-8561"},"institutions":[{"id":"https://openalex.org/I4210113369","display_name":"Microsoft Research Asia (China)","ror":"https://ror.org/0300m5276","country_code":"CN","type":"company","lineage":["https://openalex.org/I1290206253","https://openalex.org/I4210113369"]}],"countries":["CN"],"is_corresponding":false,"raw_author_name":"Yanjie Gao","raw_affiliation_strings":["Microsoft Research, Beijing, China"],"raw_orcid":"https://orcid.org/0000-0003-1899-8561","affiliations":[{"raw_affiliation_string":"Microsoft Research, Beijing, China","institution_ids":["https://openalex.org/I4210113369"]}]},{"author_position":"middle","author":{"id":null,"display_name":"Jiyu Luo","orcid":"https://orcid.org/0009-0006-4315-8061"},"institutions":[{"id":"https://openalex.org/I126520041","display_name":"University of Science and Technology of China","ror":"https://ror.org/04c4dkn09","country_code":"CN","type":"education","lineage":["https://openalex.org/I126520041","https://openalex.org/I19820366"]}],"countries":["CN"],"is_corresponding":false,"raw_author_name":"Jiyu Luo","raw_affiliation_strings":["University of Science and Technology of China, Hefei, China"],"raw_orcid":"https://orcid.org/0009-0006-4315-8061","affiliations":[{"raw_affiliation_string":"University of Science and Technology of China, Hefei, China","institution_ids":["https://openalex.org/I126520041"]}]},{"author_position":"middle","author":{"id":"https://openalex.org/A5013547786","display_name":"Haoxiang Lin","orcid":"https://orcid.org/0000-0002-9148-5861"},"institutions":[{"id":"https://openalex.org/I4210113369","display_name":"Microsoft Research Asia (China)","ror":"https://ror.org/0300m5276","country_code":"CN","type":"company","lineage":["https://openalex.org/I1290206253","https://openalex.org/I4210113369"]}],"countries":["CN"],"is_corresponding":false,"raw_author_name":"Haoxiang Lin","raw_affiliation_strings":["Microsoft Research, Beijing, China"],"raw_orcid":"https://orcid.org/0000-0002-9148-5861","affiliations":[{"raw_affiliation_string":"Microsoft Research, Beijing, China","institution_ids":["https://openalex.org/I4210113369"]}]},{"author_position":"middle","author":{"id":"https://openalex.org/A5100412598","display_name":"Hongyu Zhang","orcid":"https://orcid.org/0000-0002-3063-9425"},"institutions":[{"id":"https://openalex.org/I158842170","display_name":"Chongqing University","ror":"https://ror.org/023rhb549","country_code":"CN","type":"education","lineage":["https://openalex.org/I158842170"]}],"countries":["CN"],"is_corresponding":false,"raw_author_name":"Hongyu Zhang","raw_affiliation_strings":["Chongqing University, Chongqing, China"],"raw_orcid":"https://orcid.org/0000-0002-3063-9425","affiliations":[{"raw_affiliation_string":"Chongqing University, Chongqing, China","institution_ids":["https://openalex.org/I158842170"]}]},{"author_position":"middle","author":{"id":"https://openalex.org/A5103080159","display_name":"Ming Wu","orcid":"https://orcid.org/0009-0001-6663-4861"},"institutions":[{"id":"https://openalex.org/I4210139061","display_name":"Zero Emissions Resource Organisation","ror":"https://ror.org/03j5ebv52","country_code":"NO","type":"nonprofit","lineage":["https://openalex.org/I4210139061"]}],"countries":["NO"],"is_corresponding":false,"raw_author_name":"Ming Wu","raw_affiliation_strings":["Zero Gravity Labs, San Francisco, USA"],"raw_orcid":"https://orcid.org/0009-0001-6663-4861","affiliations":[{"raw_affiliation_string":"Zero Gravity Labs, San Francisco, USA","institution_ids":["https://openalex.org/I4210139061"]}]},{"author_position":"last","author":{"id":"https://openalex.org/A5100438310","display_name":"Mao Yang","orcid":"https://orcid.org/0009-0009-6455-3898"},"institutions":[{"id":"https://openalex.org/I4210113369","display_name":"Microsoft Research Asia (China)","ror":"https://ror.org/0300m5276","country_code":"CN","type":"company","lineage":["https://openalex.org/I1290206253","https://openalex.org/I4210113369"]}],"countries":["CN"],"is_corresponding":false,"raw_author_name":"Mao Yang","raw_affiliation_strings":["Microsoft Research, Beijing, China"],"raw_orcid":"https://orcid.org/0009-0009-6455-3898","affiliations":[{"raw_affiliation_string":"Microsoft Research, Beijing, China","institution_ids":["https://openalex.org/I4210113369"]}]}],"institutions":[],"countries_distinct_count":2,"institutions_distinct_count":4,"corresponding_author_ids":[],"corresponding_institution_ids":[],"apc_list":null,"apc_paid":null,"fwci":null,"has_fulltext":true,"cited_by_count":1,"citation_normalized_percentile":null,"cited_by_percentile_year":null,"biblio":{"volume":null,"issue":null,"first_page":"27","last_page":"38"},"is_retracted":false,"is_paratext":false,"is_xpac":false,"primary_topic":{"id":"https://openalex.org/T10772","display_name":"Distributed systems and fault tolerance","score":0.9868999719619751,"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/T10772","display_name":"Distributed systems and fault tolerance","score":0.9868999719619751,"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/T12607","display_name":"Personal Information Management and User Behavior","score":0.9843000173568726,"subfield":{"id":"https://openalex.org/subfields/1802","display_name":"Information Systems and Management"},"field":{"id":"https://openalex.org/fields/18","display_name":"Decision Sciences"},"domain":{"id":"https://openalex.org/domains/2","display_name":"Social Sciences"}},{"id":"https://openalex.org/T10400","display_name":"Network Security and Intrusion Detection","score":0.9790999889373779,"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":0.8150007724761963},{"id":"https://openalex.org/keywords/distributed-computing","display_name":"Distributed computing","score":0.3800511360168457},{"id":"https://openalex.org/keywords/parallel-computing","display_name":"Parallel computing","score":0.35801568627357483},{"id":"https://openalex.org/keywords/artificial-intelligence","display_name":"Artificial intelligence","score":0.35108309984207153},{"id":"https://openalex.org/keywords/computer-network","display_name":"Computer network","score":0.33092164993286133}],"concepts":[{"id":"https://openalex.org/C41008148","wikidata":"https://www.wikidata.org/wiki/Q21198","display_name":"Computer science","level":0,"score":0.8150007724761963},{"id":"https://openalex.org/C120314980","wikidata":"https://www.wikidata.org/wiki/Q180634","display_name":"Distributed computing","level":1,"score":0.3800511360168457},{"id":"https://openalex.org/C173608175","wikidata":"https://www.wikidata.org/wiki/Q232661","display_name":"Parallel computing","level":1,"score":0.35801568627357483},{"id":"https://openalex.org/C154945302","wikidata":"https://www.wikidata.org/wiki/Q11660","display_name":"Artificial intelligence","level":1,"score":0.35108309984207153},{"id":"https://openalex.org/C31258907","wikidata":"https://www.wikidata.org/wiki/Q1301371","display_name":"Computer network","level":1,"score":0.33092164993286133}],"mesh":[],"locations_count":1,"locations":[{"id":"doi:10.1145/3696630.3728529","is_oa":true,"landing_page_url":"https://doi.org/10.1145/3696630.3728529","pdf_url":"https://dl.acm.org/doi/pdf/10.1145/3696630.3728529","source":null,"license":"cc-by-nc","license_id":"https://openalex.org/licenses/cc-by-nc","version":"publishedVersion","is_accepted":true,"is_published":true,"raw_source_name":"Proceedings of the 33rd ACM International Conference on the Foundations of Software Engineering","raw_type":"proceedings-article"}],"best_oa_location":{"id":"doi:10.1145/3696630.3728529","is_oa":true,"landing_page_url":"https://doi.org/10.1145/3696630.3728529","pdf_url":"https://dl.acm.org/doi/pdf/10.1145/3696630.3728529","source":null,"license":"cc-by-nc","license_id":"https://openalex.org/licenses/cc-by-nc","version":"publishedVersion","is_accepted":true,"is_published":true,"raw_source_name":"Proceedings of the 33rd ACM International Conference on the Foundations of Software Engineering","raw_type":"proceedings-article"},"sustainable_development_goals":[{"score":0.5899999737739563,"id":"https://metadata.un.org/sdg/8","display_name":"Decent work and economic growth"}],"awards":[],"funders":[],"has_content":{"grobid_xml":true,"pdf":true},"content_urls":{"pdf":"https://content.openalex.org/works/W4412703834.pdf","grobid_xml":"https://content.openalex.org/works/W4412703834.grobid-xml"},"referenced_works_count":22,"referenced_works":["https://openalex.org/W95608104","https://openalex.org/W2294580796","https://openalex.org/W2604348044","https://openalex.org/W2734941459","https://openalex.org/W2911233971","https://openalex.org/W2911270308","https://openalex.org/W2912193749","https://openalex.org/W2913107170","https://openalex.org/W2914526098","https://openalex.org/W2979826702","https://openalex.org/W2990138404","https://openalex.org/W2998338832","https://openalex.org/W4213153339","https://openalex.org/W4213308398","https://openalex.org/W4214708455","https://openalex.org/W4224308101","https://openalex.org/W4244832872","https://openalex.org/W4254124737","https://openalex.org/W4301045096","https://openalex.org/W4396895148","https://openalex.org/W4400798811","https://openalex.org/W6810081322"],"related_works":["https://openalex.org/W4391375266","https://openalex.org/W2899084033","https://openalex.org/W2748952813","https://openalex.org/W2390279801","https://openalex.org/W4391913857","https://openalex.org/W2358668433","https://openalex.org/W4396701345","https://openalex.org/W2376932109","https://openalex.org/W2001405890","https://openalex.org/W4396696052"],"abstract_inverted_index":{"In":[0],"recent":[1],"years,":[2],"deep":[3,72,100,157],"learning":[4,73,101,158],"has":[5],"seen":[6],"widespread":[7],"adoption":[8],"across":[9],"various":[10],"domains,":[11],"giving":[12],"rise":[13],"to":[14,57,65,75,95,123],"large-scale":[15],"models":[16,104],"such":[17],"as":[18,111],"large":[19],"language":[20],"models.":[21],"Training":[22],"these":[23],"models,":[24],"particularly":[25],"in":[26,43,99],"distributed":[27],"environments,":[28],"presents":[29,87],"substantial":[30],"computational":[31],"and":[32,60,118,136,155,174],"communication":[33,40,97,132,169],"challenges.":[34],"A":[35],"critical":[36],"issue":[37],"is":[38],"the":[39,105],"deadlock\u2014a":[41],"state":[42],"which":[44,55,176],"processes":[45],"become":[46],"indefinitely":[47],"stalled":[48],"while":[49],"awaiting":[50],"network":[51],"messages":[52],"from":[53],"others,":[54],"leads":[56],"resource":[58],"wastage":[59],"reduced":[61],"productivity.":[62],"Current":[63],"approaches":[64],"deadlock":[66,125],"handling":[67],"are":[68],"either":[69],"unsuitable":[70],"for":[71,139],"due":[74],"its":[76,178],"unique":[77],"hybrid":[78],"programming":[79],"paradigm":[80],"or":[81],"limit":[82],"optimization":[83],"opportunities.":[84],"This":[85],"paper":[86],"dl2,":[88],"a":[89,109,120,149],"novel":[90],"dynamic":[91],"analysis":[92],"tool":[93],"designed":[94],"detect":[96],"deadlocks":[98],"jobs.":[102],"dl2":[103,127,145,165],"runtime":[106],"trace":[107],"of":[108,151],"job":[110],"an":[112],"execution":[113],"graph,":[114],"detects":[115,167],"unmatched":[116],"communications,":[117],"constructs":[119],"wait-for":[121],"graph":[122],"identify":[124],"cycles.":[126],"can":[128],"also":[129],"handle":[130],"nondeterministic":[131],"behaviors,":[133],"providing":[134],"replay":[135],"diagnostic":[137],"support":[138],"root":[140],"cause":[141],"analysis.":[142],"We":[143],"evaluate":[144],"using":[146],"PyTorch":[147],"with":[148],"combination":[150],"synthetic":[152],"test":[153],"cases":[154],"real-world":[156],"workloads.":[159],"The":[160],"experimental":[161],"results":[162],"show":[163],"that":[164],"successfully":[166],"all":[168],"deadlocks,":[170],"achieving":[171],"100%":[172],"precision":[173],"recall,":[175],"highlights":[177],"effectiveness.":[179]},"counts_by_year":[{"year":2026,"cited_by_count":1}],"updated_date":"2026-07-14T23:27:15.235271","created_date":"2025-10-10T00:00:00"}