{"id":"https://openalex.org/W7158267722","doi":"https://doi.org/10.1145/3805621.3807617","title":"Cost-Efficient Training and Checkpointing for Large Models on Preemptible Cloud VMs","display_name":"Cost-Efficient Training and Checkpointing for Large Models on Preemptible Cloud VMs","publication_year":2026,"publication_date":"2026-04-27","ids":{"openalex":"https://openalex.org/W7158267722","doi":"https://doi.org/10.1145/3805621.3807617"},"language":null,"primary_location":{"id":"doi:10.1145/3805621.3807617","is_oa":true,"landing_page_url":"https://doi.org/10.1145/3805621.3807617","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 Sixth European Workshop on Machine Learning and Systems","raw_type":"proceedings-article"},"type":"article","indexed_in":["crossref"],"open_access":{"is_oa":true,"oa_status":"gold","oa_url":"https://doi.org/10.1145/3805621.3807617","any_repository_has_fulltext":null},"authorships":[{"author_position":"first","author":{"id":"https://openalex.org/A5063257885","display_name":"Omkar Desai","orcid":"https://orcid.org/0000-0002-5678-7434"},"institutions":[{"id":"https://openalex.org/I70983195","display_name":"Syracuse University","ror":"https://ror.org/025r5qe02","country_code":"US","type":"education","lineage":["https://openalex.org/I70983195"]}],"countries":["US"],"is_corresponding":false,"raw_author_name":"Omkar Desai","raw_affiliation_strings":["Syracuse University, Syracuse, New York, USA"],"raw_orcid":"https://orcid.org/0000-0002-5678-7434","affiliations":[{"raw_affiliation_string":"Syracuse University, Syracuse, New York, USA","institution_ids":["https://openalex.org/I70983195"]}]},{"author_position":"middle","author":{"id":"https://openalex.org/A5085780376","display_name":"Shuyi Pei","orcid":"https://orcid.org/0000-0002-8649-4253"},"institutions":[{"id":"https://openalex.org/I4210101778","display_name":"Samsung (United States)","ror":"https://ror.org/01bfbvm65","country_code":"US","type":"company","lineage":["https://openalex.org/I2250650973","https://openalex.org/I4210101778"]}],"countries":["US"],"is_corresponding":false,"raw_author_name":"Shuyi Pei","raw_affiliation_strings":["Samsung Semiconductor, Inc., San Jose, CA, USA"],"raw_orcid":"https://orcid.org/0000-0002-8649-4253","affiliations":[{"raw_affiliation_string":"Samsung Semiconductor, Inc., San Jose, CA, USA","institution_ids":["https://openalex.org/I4210101778"]}]},{"author_position":"middle","author":{"id":"https://openalex.org/A5011474644","display_name":"Janki Bhimani","orcid":"https://orcid.org/0000-0001-7284-1464"},"institutions":[{"id":"https://openalex.org/I19700959","display_name":"Florida International University","ror":"https://ror.org/02gz6gg07","country_code":"US","type":"education","lineage":["https://openalex.org/I19700959"]}],"countries":["US"],"is_corresponding":false,"raw_author_name":"Janki Bhimani","raw_affiliation_strings":["Florida International University, Miami, FL, USA"],"raw_orcid":"https://orcid.org/0000-0002-4421-9923","affiliations":[{"raw_affiliation_string":"Florida International University, Miami, FL, USA","institution_ids":["https://openalex.org/I19700959"]}]},{"author_position":"last","author":{"id":"https://openalex.org/A5000081394","display_name":"Bryan S. Kim","orcid":"https://orcid.org/0000-0002-3689-3985"},"institutions":[{"id":"https://openalex.org/I70983195","display_name":"Syracuse University","ror":"https://ror.org/025r5qe02","country_code":"US","type":"education","lineage":["https://openalex.org/I70983195"]}],"countries":["US"],"is_corresponding":false,"raw_author_name":"Bryan S. Kim","raw_affiliation_strings":["Syracuse University, Syracuse, NY, USA"],"raw_orcid":"https://orcid.org/0000-0002-3689-3985","affiliations":[{"raw_affiliation_string":"Syracuse University, Syracuse, NY, USA","institution_ids":["https://openalex.org/I70983195"]}]}],"institutions":[],"countries_distinct_count":1,"institutions_distinct_count":4,"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":0.68789792,"is_in_top_1_percent":false,"is_in_top_10_percent":false},"cited_by_percentile_year":null,"biblio":{"volume":null,"issue":null,"first_page":"31","last_page":"40"},"is_retracted":false,"is_paratext":false,"is_xpac":false,"primary_topic":{"id":"https://openalex.org/T10101","display_name":"Cloud Computing and Resource Management","score":0.33550000190734863,"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":0.33550000190734863,"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/T12127","display_name":"Software System Performance and Reliability","score":0.17260000109672546,"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/T14347","display_name":"Big Data and Digital Economy","score":0.05090000107884407,"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"}}],"keywords":[{"id":"https://openalex.org/keywords/preemption","display_name":"Preemption","score":0.7490000128746033},{"id":"https://openalex.org/keywords/throughput","display_name":"Throughput","score":0.738099992275238},{"id":"https://openalex.org/keywords/training","display_name":"Training (meteorology)","score":0.6559000015258789},{"id":"https://openalex.org/keywords/cloud-computing","display_name":"Cloud computing","score":0.5831999778747559},{"id":"https://openalex.org/keywords/interval","display_name":"Interval (graph theory)","score":0.5738000273704529},{"id":"https://openalex.org/keywords/hot-spot","display_name":"Hot spot (computer programming)","score":0.41119998693466187}],"concepts":[{"id":"https://openalex.org/C41008148","wikidata":"https://www.wikidata.org/wiki/Q21198","display_name":"Computer science","level":0,"score":0.8098000288009644},{"id":"https://openalex.org/C206952183","wikidata":"https://www.wikidata.org/wiki/Q1193100","display_name":"Preemption","level":2,"score":0.7490000128746033},{"id":"https://openalex.org/C157764524","wikidata":"https://www.wikidata.org/wiki/Q1383412","display_name":"Throughput","level":3,"score":0.738099992275238},{"id":"https://openalex.org/C2777211547","wikidata":"https://www.wikidata.org/wiki/Q17141490","display_name":"Training (meteorology)","level":2,"score":0.6559000015258789},{"id":"https://openalex.org/C79974875","wikidata":"https://www.wikidata.org/wiki/Q483639","display_name":"Cloud computing","level":2,"score":0.5831999778747559},{"id":"https://openalex.org/C2778067643","wikidata":"https://www.wikidata.org/wiki/Q166507","display_name":"Interval (graph theory)","level":2,"score":0.5738000273704529},{"id":"https://openalex.org/C120314980","wikidata":"https://www.wikidata.org/wiki/Q180634","display_name":"Distributed computing","level":1,"score":0.47290000319480896},{"id":"https://openalex.org/C79403827","wikidata":"https://www.wikidata.org/wiki/Q3988","display_name":"Real-time computing","level":1,"score":0.4645000100135803},{"id":"https://openalex.org/C199672914","wikidata":"https://www.wikidata.org/wiki/Q4241353","display_name":"Hot spot (computer programming)","level":2,"score":0.41119998693466187},{"id":"https://openalex.org/C8505890","wikidata":"https://www.wikidata.org/wiki/Q605095","display_name":"Budget constraint","level":2,"score":0.29910001158714294},{"id":"https://openalex.org/C206345919","wikidata":"https://www.wikidata.org/wiki/Q20380951","display_name":"Resource (disambiguation)","level":2,"score":0.27959999442100525},{"id":"https://openalex.org/C99844830","wikidata":"https://www.wikidata.org/wiki/Q102441924","display_name":"Scaling","level":2,"score":0.27480000257492065},{"id":"https://openalex.org/C111919701","wikidata":"https://www.wikidata.org/wiki/Q9135","display_name":"Operating system","level":1,"score":0.2635999917984009},{"id":"https://openalex.org/C29202148","wikidata":"https://www.wikidata.org/wiki/Q287260","display_name":"Resource allocation","level":2,"score":0.26269999146461487},{"id":"https://openalex.org/C51632099","wikidata":"https://www.wikidata.org/wiki/Q3985153","display_name":"Training set","level":2,"score":0.2558000087738037},{"id":"https://openalex.org/C31258907","wikidata":"https://www.wikidata.org/wiki/Q1301371","display_name":"Computer network","level":1,"score":0.2538999915122986}],"mesh":[],"locations_count":1,"locations":[{"id":"doi:10.1145/3805621.3807617","is_oa":true,"landing_page_url":"https://doi.org/10.1145/3805621.3807617","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 Sixth European Workshop on Machine Learning and Systems","raw_type":"proceedings-article"}],"best_oa_location":{"id":"doi:10.1145/3805621.3807617","is_oa":true,"landing_page_url":"https://doi.org/10.1145/3805621.3807617","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 Sixth European Workshop on Machine Learning and Systems","raw_type":"proceedings-article"},"sustainable_development_goals":[],"awards":[{"id":"https://openalex.org/G1489511460","display_name":null,"funder_award_id":"2443515","funder_id":"https://openalex.org/F4320306076","funder_display_name":"National Science Foundation"},{"id":"https://openalex.org/G641836305","display_name":null,"funder_award_id":"RS-2024-00428758","funder_id":"https://openalex.org/F4320306076","funder_display_name":"National Science Foundation"}],"funders":[{"id":"https://openalex.org/F4320306076","display_name":"National Science Foundation","ror":"https://ror.org/021nxhr62"}],"has_content":{"pdf":false,"grobid_xml":false},"content_urls":null,"referenced_works_count":4,"referenced_works":["https://openalex.org/W2033656974","https://openalex.org/W3081168214","https://openalex.org/W4387302750","https://openalex.org/W4404400583"],"related_works":[],"abstract_inverted_index":{"Training":[0],"large":[1,35],"models":[2,36],"on":[3],"discounted":[4],"spot":[5,38,48,97,109,152],"VMs":[6,110],"offers":[7],"significant":[8],"cost":[9,86,147],"savings":[10],"but":[11],"remains":[12],"challenging":[13],"due":[14],"to":[15,93,123,132,139],"their":[16],"low":[17,101],"availability":[18,99,114],"and":[19,31,69,75,83,153],"unilateral":[20],"preemptions.":[21],"To":[22],"address":[23],"these":[24],"challenges,":[25],"we":[26,41,55,78],"present":[27],"a":[28,47,61,144],"cost-effective":[29],"training":[30,81,85,112,119],"checkpointing":[32,58,74],"system":[33],"for":[34,46,111],"using":[37,50],"VMs.":[39],"First,":[40],"predict":[42],"the":[43,66,71,106],"preemption":[44,52,67],"rate":[45],"instance":[49],"historical":[51],"data.":[53],"Second,":[54],"dynamically":[56],"tune":[57],"interval":[59,128],"through":[60,87,148],"mathematical":[62],"model":[63],"that":[64],"uses":[65],"predictions":[68],"balances":[70],"overheads":[72],"of":[73,108,151],"recovery.":[76],"Finally,":[77],"guarantee":[79],"consistent":[80],"throughput":[82,120,142],"minimized":[84],"prediction-informed":[88],"hybrid":[89],"resource":[90],"utilization:":[91],"switching":[92],"on-demand":[94,154],"instances":[95],"when":[96,113,130],"VM":[98],"is":[100,115],"while":[102],"also":[103,136],"opportunistically":[104],"scaling":[105],"number":[107],"high.":[116],"We":[117,135],"improve":[118],"by":[121],"up":[122,138],"60.27%":[124],"with":[125],"dynamic":[126],"checkpoint":[127],"tuning":[129],"compared":[131],"fixed-interval":[133],"approaches.":[134],"achieve":[137],"2.04\u00d7":[140],"higher":[141],"at":[143],"51.41%":[145],"lower":[146],"effective":[149],"use":[150],"instances.":[155]},"counts_by_year":[],"updated_date":"2026-06-11T09:08:48.828518","created_date":"2026-04-30T00:00:00"}