{"id":"https://openalex.org/W4399757647","doi":"https://doi.org/10.1145/3625549.3658685","title":"DataStates-LLM: Lazy Asynchronous Checkpointing for Large Language Models","display_name":"DataStates-LLM: Lazy Asynchronous Checkpointing for Large Language Models","publication_year":2024,"publication_date":"2024-06-03","ids":{"openalex":"https://openalex.org/W4399757647","doi":"https://doi.org/10.1145/3625549.3658685"},"language":"en","primary_location":{"id":"doi:10.1145/3625549.3658685","is_oa":true,"landing_page_url":"https://doi.org/10.1145/3625549.3658685","pdf_url":null,"source":null,"license":"cc-by","license_id":"https://openalex.org/licenses/cc-by","version":"publishedVersion","is_accepted":true,"is_published":true,"raw_source_name":"Proceedings of the 33rd International Symposium on High-Performance Parallel and Distributed Computing","raw_type":"proceedings-article"},"type":"conference-paper","indexed_in":["arxiv","crossref"],"open_access":{"is_oa":true,"oa_status":"gold","oa_url":"https://doi.org/10.1145/3625549.3658685","any_repository_has_fulltext":true},"authorships":[{"author_position":"first","author":{"id":"https://openalex.org/A5049218913","display_name":"Avinash Maurya","orcid":"https://orcid.org/0000-0002-8200-0148"},"institutions":[{"id":"https://openalex.org/I155173764","display_name":"Rochester Institute of Technology","ror":"https://ror.org/00v4yb702","country_code":"US","type":"education","lineage":["https://openalex.org/I155173764"]}],"countries":["US"],"is_corresponding":false,"raw_author_name":"Avinash Maurya","raw_affiliation_strings":["Department of Computer Science, Rochester Institute of Technology, Rochester, New York, United States of America"],"raw_orcid":"https://orcid.org/0000-0002-8200-0148","affiliations":[{"raw_affiliation_string":"Department of Computer Science, Rochester Institute of Technology, Rochester, New York, United States of America","institution_ids":["https://openalex.org/I155173764"]}]},{"author_position":"middle","author":{"id":"https://openalex.org/A5056940311","display_name":"Robert Underwood","orcid":"https://orcid.org/0000-0002-1464-729X"},"institutions":[{"id":"https://openalex.org/I1282105669","display_name":"Argonne National Laboratory","ror":"https://ror.org/05gvnxz63","country_code":"US","type":"facility","lineage":["https://openalex.org/I1282105669","https://openalex.org/I1330989302","https://openalex.org/I39565521","https://openalex.org/I40347166"]}],"countries":["US"],"is_corresponding":false,"raw_author_name":"Robert Underwood","raw_affiliation_strings":["Argonne National Laboratory, Lemont, Illinois, United States of America","ANL - Argonne National Laboratory [Lemont] (9700 S Cass Ave B109, Lemont, IL, 60439 - United States)"],"raw_orcid":"https://orcid.org/0000-0002-1464-729X","affiliations":[{"raw_affiliation_string":"Argonne National Laboratory, Lemont, Illinois, United States of America","institution_ids":["https://openalex.org/I1282105669"]},{"raw_affiliation_string":"ANL - Argonne National Laboratory [Lemont] (9700 S Cass Ave B109, Lemont, IL, 60439 - United States)","institution_ids":["https://openalex.org/I1282105669"]}]},{"author_position":"middle","author":{"id":"https://openalex.org/A5054011637","display_name":"M. Mustafa Rafique","orcid":"https://orcid.org/0000-0002-5034-2880"},"institutions":[{"id":"https://openalex.org/I155173764","display_name":"Rochester Institute of Technology","ror":"https://ror.org/00v4yb702","country_code":"US","type":"education","lineage":["https://openalex.org/I155173764"]}],"countries":["US"],"is_corresponding":false,"raw_author_name":"M. Mustafa Rafique","raw_affiliation_strings":["Department of Computer Science, Rochester Institute of Technology, Rochester, New York, United States of America"],"raw_orcid":"https://orcid.org/0000-0002-5034-2880","affiliations":[{"raw_affiliation_string":"Department of Computer Science, Rochester Institute of Technology, Rochester, New York, United States of America","institution_ids":["https://openalex.org/I155173764"]}]},{"author_position":"middle","author":{"id":"https://openalex.org/A5046613458","display_name":"Franck Cappello","orcid":"https://orcid.org/0000-0002-7890-3934"},"institutions":[{"id":"https://openalex.org/I1282105669","display_name":"Argonne National Laboratory","ror":"https://ror.org/05gvnxz63","country_code":"US","type":"facility","lineage":["https://openalex.org/I1282105669","https://openalex.org/I1330989302","https://openalex.org/I39565521","https://openalex.org/I40347166"]}],"countries":["US"],"is_corresponding":false,"raw_author_name":"Franck Cappello","raw_affiliation_strings":["Argonne National Laboratory, Lemont, Illinois, United States of America","ANL - Argonne National Laboratory [Lemont] (9700 S Cass Ave B109, Lemont, IL, 60439 - United States)"],"raw_orcid":"https://orcid.org/0000-0002-7890-3934","affiliations":[{"raw_affiliation_string":"Argonne National Laboratory, Lemont, Illinois, United States of America","institution_ids":["https://openalex.org/I1282105669"]},{"raw_affiliation_string":"ANL - Argonne National Laboratory [Lemont] (9700 S Cass Ave B109, Lemont, IL, 60439 - United States)","institution_ids":["https://openalex.org/I1282105669"]}]},{"author_position":"last","author":{"id":"https://openalex.org/A5085745891","display_name":"Bogdan Nicolae","orcid":"https://orcid.org/0000-0002-0661-7509"},"institutions":[{"id":"https://openalex.org/I1282105669","display_name":"Argonne National Laboratory","ror":"https://ror.org/05gvnxz63","country_code":"US","type":"facility","lineage":["https://openalex.org/I1282105669","https://openalex.org/I1330989302","https://openalex.org/I39565521","https://openalex.org/I40347166"]}],"countries":["US"],"is_corresponding":false,"raw_author_name":"Bogdan Nicolae","raw_affiliation_strings":["Argonne National Laboratory, Lemont, Illinois, USA","ANL - Argonne National Laboratory [Lemont] (9700 S Cass Ave B109, Lemont, IL, 60439 - United States)"],"raw_orcid":"https://orcid.org/0000-0002-0661-7509","affiliations":[{"raw_affiliation_string":"Argonne National Laboratory, Lemont, Illinois, USA","institution_ids":["https://openalex.org/I1282105669"]},{"raw_affiliation_string":"ANL - Argonne National Laboratory [Lemont] (9700 S Cass Ave B109, Lemont, IL, 60439 - United States)","institution_ids":["https://openalex.org/I1282105669"]}]}],"institutions":[],"countries_distinct_count":1,"institutions_distinct_count":2,"corresponding_author_ids":[],"corresponding_institution_ids":[],"apc_list":null,"apc_paid":null,"fwci":16.1577,"has_fulltext":true,"cited_by_count":29,"citation_normalized_percentile":{"value":0.99607167,"is_in_top_1_percent":true,"is_in_top_10_percent":true},"cited_by_percentile_year":{"min":96,"max":100},"biblio":{"volume":null,"issue":null,"first_page":"227","last_page":"239"},"is_retracted":false,"is_paratext":false,"is_xpac":false,"primary_topic":{"id":"https://openalex.org/T10772","display_name":"Distributed systems and fault tolerance","score":0.9969000220298767,"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":0.9969000220298767,"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":0.9957000017166138,"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"}},{"id":"https://openalex.org/T12808","display_name":"Ferroelectric and Negative Capacitance Devices","score":0.9952999949455261,"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"}}],"keywords":[{"id":"https://openalex.org/keywords/computer-science","display_name":"Computer science","score":0.870658278465271},{"id":"https://openalex.org/keywords/asynchronous-communication","display_name":"Asynchronous communication","score":0.8353250026702881},{"id":"https://openalex.org/keywords/scalability","display_name":"Scalability","score":0.7605575919151306},{"id":"https://openalex.org/keywords/granularity","display_name":"Granularity","score":0.6640466451644897},{"id":"https://openalex.org/keywords/parallel-computing","display_name":"Parallel computing","score":0.6063946485519409},{"id":"https://openalex.org/keywords/process","display_name":"Process (computing)","score":0.5675054788589478},{"id":"https://openalex.org/keywords/state","display_name":"State (computer science)","score":0.519860565662384},{"id":"https://openalex.org/keywords/distributed-computing","display_name":"Distributed computing","score":0.47809407114982605},{"id":"https://openalex.org/keywords/software","display_name":"Software","score":0.43111446499824524},{"id":"https://openalex.org/keywords/operating-system","display_name":"Operating system","score":0.29413124918937683},{"id":"https://openalex.org/keywords/programming-language","display_name":"Programming language","score":0.2712019383907318},{"id":"https://openalex.org/keywords/computer-network","display_name":"Computer network","score":0.13789686560630798}],"concepts":[{"id":"https://openalex.org/C41008148","wikidata":"https://www.wikidata.org/wiki/Q21198","display_name":"Computer science","level":0,"score":0.870658278465271},{"id":"https://openalex.org/C151319957","wikidata":"https://www.wikidata.org/wiki/Q752739","display_name":"Asynchronous communication","level":2,"score":0.8353250026702881},{"id":"https://openalex.org/C48044578","wikidata":"https://www.wikidata.org/wiki/Q727490","display_name":"Scalability","level":2,"score":0.7605575919151306},{"id":"https://openalex.org/C177774035","wikidata":"https://www.wikidata.org/wiki/Q1246948","display_name":"Granularity","level":2,"score":0.6640466451644897},{"id":"https://openalex.org/C173608175","wikidata":"https://www.wikidata.org/wiki/Q232661","display_name":"Parallel computing","level":1,"score":0.6063946485519409},{"id":"https://openalex.org/C98045186","wikidata":"https://www.wikidata.org/wiki/Q205663","display_name":"Process (computing)","level":2,"score":0.5675054788589478},{"id":"https://openalex.org/C48103436","wikidata":"https://www.wikidata.org/wiki/Q599031","display_name":"State (computer science)","level":2,"score":0.519860565662384},{"id":"https://openalex.org/C120314980","wikidata":"https://www.wikidata.org/wiki/Q180634","display_name":"Distributed computing","level":1,"score":0.47809407114982605},{"id":"https://openalex.org/C2777904410","wikidata":"https://www.wikidata.org/wiki/Q7397","display_name":"Software","level":2,"score":0.43111446499824524},{"id":"https://openalex.org/C111919701","wikidata":"https://www.wikidata.org/wiki/Q9135","display_name":"Operating system","level":1,"score":0.29413124918937683},{"id":"https://openalex.org/C199360897","wikidata":"https://www.wikidata.org/wiki/Q9143","display_name":"Programming language","level":1,"score":0.2712019383907318},{"id":"https://openalex.org/C31258907","wikidata":"https://www.wikidata.org/wiki/Q1301371","display_name":"Computer network","level":1,"score":0.13789686560630798}],"mesh":[],"locations_count":3,"locations":[{"id":"doi:10.1145/3625549.3658685","is_oa":true,"landing_page_url":"https://doi.org/10.1145/3625549.3658685","pdf_url":null,"source":null,"license":"cc-by","license_id":"https://openalex.org/licenses/cc-by","version":"publishedVersion","is_accepted":true,"is_published":true,"raw_source_name":"Proceedings of the 33rd International Symposium on High-Performance Parallel and Distributed Computing","raw_type":"proceedings-article"},{"id":"pmh:oai:arXiv.org:2406.10707","is_oa":true,"landing_page_url":"http://arxiv.org/abs/2406.10707","pdf_url":"https://arxiv.org/pdf/2406.10707","source":{"id":"https://openalex.org/S4306400194","display_name":"arXiv (Cornell University)","issn_l":"2331-8422","issn":["2331-8422"],"is_oa":true,"is_in_doaj":false,"is_core":false,"host_organization":"https://openalex.org/I205783295","host_organization_name":"Cornell University","host_organization_lineage":["https://openalex.org/I205783295"],"host_organization_lineage_names":[],"type":"repository"},"license":"cc-by","license_id":"https://openalex.org/licenses/cc-by","version":"submittedVersion","is_accepted":false,"is_published":false,"raw_source_name":null,"raw_type":"text"},{"id":"pmh:oai:HAL:hal-04614247v1","is_oa":true,"landing_page_url":"https://hal.science/hal-04614247","pdf_url":"https://hal.science/hal-04614247/document","source":{"id":"https://openalex.org/S4306402512","display_name":"HAL (Le Centre pour la Communication Scientifique Directe)","issn_l":null,"issn":null,"is_oa":false,"is_in_doaj":false,"is_core":false,"host_organization":"https://openalex.org/I1294671590","host_organization_name":"Centre National de la Recherche Scientifique","host_organization_lineage":["https://openalex.org/I1294671590"],"host_organization_lineage_names":[],"type":"repository"},"license":"cc-by","license_id":"https://openalex.org/licenses/cc-by","version":"submittedVersion","is_accepted":false,"is_published":false,"raw_source_name":"HPDC'24: 33nd International Symposium on High-Performance Parallel and Distributed Computing, Jun 2024, Pisa (IT), Italy. &#x27E8;10.1145/3625549.3658685&#x27E9;","raw_type":"info:eu-repo/semantics/conferenceObject"}],"best_oa_location":{"id":"doi:10.1145/3625549.3658685","is_oa":true,"landing_page_url":"https://doi.org/10.1145/3625549.3658685","pdf_url":null,"source":null,"license":"cc-by","license_id":"https://openalex.org/licenses/cc-by","version":"publishedVersion","is_accepted":true,"is_published":true,"raw_source_name":"Proceedings of the 33rd International Symposium on High-Performance Parallel and Distributed Computing","raw_type":"proceedings-article"},"sustainable_development_goals":[],"awards":[{"id":"https://openalex.org/G102163711","display_name":null,"funder_award_id":"DE\u2010 AC02\u201006CH11357","funder_id":"https://openalex.org/F4320306084","funder_display_name":"U.S. Department of Energy"},{"id":"https://openalex.org/G107546177","display_name":null,"funder_award_id":"DE- AC02-06-CH11357","funder_id":"https://openalex.org/F4320338284","funder_display_name":"Argonne National Laboratory"},{"id":"https://openalex.org/G2131453672","display_name":"Collaborative Research: CNS Core: Medium:HardLambda: A new FaaS Abstraction for Cross-Stack Resource Management in Disaggregated Datacenters","funder_award_id":"2106635","funder_id":"https://openalex.org/F4320306076","funder_display_name":"National Science Foundation"},{"id":"https://openalex.org/G3075337988","display_name":null,"funder_award_id":"06CH11357","funder_id":"https://openalex.org/F4320338284","funder_display_name":"Argonne National Laboratory"},{"id":"https://openalex.org/G3085353284","display_name":null,"funder_award_id":"2106634\\/2106635","funder_id":"https://openalex.org/F4320306076","funder_display_name":"National Science Foundation"},{"id":"https://openalex.org/G4552540348","display_name":null,"funder_award_id":"0F-60169","funder_id":"https://openalex.org/F4320306084","funder_display_name":"U.S. Department of Energy"},{"id":"https://openalex.org/G5367729914","display_name":null,"funder_award_id":"0F-60169","funder_id":"https://openalex.org/F4320338284","funder_display_name":"Argonne National Laboratory"},{"id":"https://openalex.org/G6848031779","display_name":null,"funder_award_id":"06CH11357","funder_id":"https://openalex.org/F4320306084","funder_display_name":"U.S. Department of Energy"},{"id":"https://openalex.org/G6918803902","display_name":null,"funder_award_id":"06CH11357","funder_id":"https://openalex.org/F4320332359","funder_display_name":"Office of Science"},{"id":"https://openalex.org/G8515454090","display_name":"Collaborative Research: CNS Core: Medium:HardLambda: A new FaaS Abstraction for Cross-Stack Resource Management in Disaggregated Datacenters","funder_award_id":"2106634","funder_id":"https://openalex.org/F4320306076","funder_display_name":"National Science Foundation"},{"id":"https://openalex.org/G878601127","display_name":null,"funder_award_id":"DE-AC02-06CH11357","funder_id":"https://openalex.org/F4320337506","funder_display_name":"Advanced Scientific Computing Research"},{"id":"https://openalex.org/G993149817","display_name":null,"funder_award_id":"DE-AC02-06CH11357 ","funder_id":"https://openalex.org/F4320332359","funder_display_name":"Office of Science"}],"funders":[{"id":"https://openalex.org/F4320306076","display_name":"National Science Foundation","ror":"https://ror.org/021nxhr62"},{"id":"https://openalex.org/F4320306084","display_name":"U.S. Department of Energy","ror":"https://ror.org/01bj3aw27"},{"id":"https://openalex.org/F4320332359","display_name":"Office of Science","ror":"https://ror.org/00mmn6b08"},{"id":"https://openalex.org/F4320337506","display_name":"Advanced Scientific Computing Research","ror":"https://ror.org/0012c7r22"},{"id":"https://openalex.org/F4320338284","display_name":"Argonne National Laboratory","ror":"https://ror.org/05gvnxz63"}],"has_content":{"pdf":false,"grobid_xml":false},"content_urls":null,"referenced_works_count":6,"referenced_works":["https://openalex.org/W2116115793","https://openalex.org/W3041193038","https://openalex.org/W3086105743","https://openalex.org/W3129831491","https://openalex.org/W3130554079","https://openalex.org/W3205803342"],"related_works":["https://openalex.org/W2931688134","https://openalex.org/W2002863235","https://openalex.org/W4401278057","https://openalex.org/W1982914007","https://openalex.org/W2159583675","https://openalex.org/W1824242903","https://openalex.org/W1493858311","https://openalex.org/W2155470929","https://openalex.org/W2394465510","https://openalex.org/W2111125783"],"abstract_inverted_index":{"LLMs":[0,58,132],"have":[1],"seen":[2],"rapid":[3],"adoption":[4],"in":[5,53,114,194],"all":[6],"domains.":[7],"They":[8],"need":[9,59],"to":[10,60,71,98,120,141,190,212,228],"be":[11,61,68,135],"trained":[12],"on":[13,150],"high-end":[14],"high-performance":[15],"computing":[16],"(HPC)":[17],"infrastructures":[18],"and":[19,48,75,95,128,175,221,232],"ingest":[20],"massive":[21],"amounts":[22],"of":[23,36,39,83,144,165,184,210],"input":[24],"data.":[25],"Unsurprisingly,":[26],"at":[27,137,208],"such":[28],"a":[29,54,72,85,102,157],"large":[30,81],"scale,":[31],"unexpected":[32],"events":[33],"(e.g.,":[34,101],"failures":[35],"components,":[37],"instability":[38],"the":[40,51,80,92,122,142,151,166,169,173,195,201,240],"software,":[41],"undesirable":[42],"learning":[43],"patterns,":[44],"etc.),":[45],"are":[46],"frequent":[47],"typically":[49],"impact":[50,149],"training":[52,152,202,236],"negative":[55],"fashion.":[56],"Thus,":[57],"checkpointed":[62],"frequently":[63],"so":[64],"that":[65,89,133,162,168],"they":[66],"can":[67,134],"rolled":[69],"back":[70],"stable":[73],"state":[74,97,177],"subsequently":[76],"fine-tuned.":[77],"However,":[78],"given":[79],"sizes":[82],"LLMs,":[84],"straightforward":[86],"checkpointing":[87,130,222,231,242],"solution":[88],"directly":[90],"writes":[91],"model":[93,174,217],"parameters":[94],"optimizer":[96,176],"persistent":[99],"storage":[100],"parallel":[103],"file":[104],"system),":[105],"incurs":[106],"significant":[107,148],"I/O":[108,123],"overheads.":[109],"To":[110],"address":[111],"this":[112,115],"challenge,":[113],"paper":[116],"we":[117,155],"study":[118],"how":[119],"reduce":[121],"overheads":[124],"for":[125,131,181],"enabling":[126],"fast":[127],"scalable":[129],"applied":[136],"high":[138],"frequency":[139],"(up":[140],"granularity":[143],"individual":[145],"iterations)":[146],"without":[147],"process.":[153,203],"Specifically,":[154],"introduce":[156],"lazy":[158],"asynchronous":[159],"multi-level":[160],"approach":[161,207],"takes":[163],"advantage":[164],"fact":[167],"tensors":[170],"making":[171],"up":[172,211,227],"shards":[178],"remain":[179],"immutable":[180],"extended":[182],"periods":[183],"time,":[185],"which":[186],"makes":[187],"it":[188],"possible":[189],"copy":[191],"their":[192],"content":[193],"background":[196],"with":[197,239],"minimal":[198],"interference":[199],"during":[200],"We":[204],"evaluate":[205],"our":[206],"scales":[209],"180":[213],"GPUs":[214],"using":[215],"different":[216],"sizes,":[218],"parallelism":[219],"settings,":[220],"frequencies.":[223],"The":[224],"results":[225],"show":[226],"48\u00d7":[229],"faster":[230,234],"2.2\u00d7":[233],"end-to-end":[235],"runtime":[237],"compared":[238],"state-of-art":[241],"approaches.":[243]},"counts_by_year":[{"year":2026,"cited_by_count":5},{"year":2025,"cited_by_count":21},{"year":2024,"cited_by_count":3}],"updated_date":"2026-08-01T09:00:35.917206","created_date":"2024-06-18T00:00:00"}
