{"id":"https://openalex.org/W4398796293","doi":"https://doi.org/10.1145/3694715.3695960","title":"ReCycle: Resilient Training of Large DNNs using Pipeline Adaptation","display_name":"ReCycle: Resilient Training of Large DNNs using Pipeline Adaptation","publication_year":2024,"publication_date":"2024-11-04","ids":{"openalex":"https://openalex.org/W4398796293","doi":"https://doi.org/10.1145/3694715.3695960"},"language":"en","primary_location":{"id":"doi:10.1145/3694715.3695960","is_oa":true,"landing_page_url":"https://doi.org/10.1145/3694715.3695960","pdf_url":"https://dl.acm.org/doi/pdf/10.1145/3694715.3695960","source":null,"license":null,"license_id":null,"version":"publishedVersion","is_accepted":true,"is_published":true,"raw_source_name":"Proceedings of the ACM SIGOPS 30th Symposium on Operating Systems Principles","raw_type":"proceedings-article"},"type":"conference-paper","indexed_in":["arxiv","crossref"],"open_access":{"is_oa":true,"oa_status":"gold","oa_url":"https://dl.acm.org/doi/pdf/10.1145/3694715.3695960","any_repository_has_fulltext":true},"authorships":[{"author_position":"first","author":{"id":"https://openalex.org/A5104313201","display_name":"Swapnil Gandhi","orcid":null},"institutions":[{"id":"https://openalex.org/I97018004","display_name":"Stanford University","ror":"https://ror.org/00f54p054","country_code":"US","type":"education","lineage":["https://openalex.org/I97018004"]}],"countries":["US"],"is_corresponding":false,"raw_author_name":"Swapnil Gandhi","raw_affiliation_strings":["Stanford University, Stanford, CA, US"],"raw_orcid":"https://orcid.org/0000-0003-3689-9591","affiliations":[{"raw_affiliation_string":"Stanford University, Stanford, CA, US","institution_ids":["https://openalex.org/I97018004"]}]},{"author_position":"middle","author":{"id":"https://openalex.org/A5065406046","display_name":"Mark Zhao","orcid":"https://orcid.org/0000-0002-0706-5208"},"institutions":[{"id":"https://openalex.org/I97018004","display_name":"Stanford University","ror":"https://ror.org/00f54p054","country_code":"US","type":"education","lineage":["https://openalex.org/I97018004"]}],"countries":["US"],"is_corresponding":false,"raw_author_name":"Mark Zhao","raw_affiliation_strings":["Stanford University, Stanford, CA, US"],"raw_orcid":"https://orcid.org/0000-0002-0706-5208","affiliations":[{"raw_affiliation_string":"Stanford University, Stanford, CA, US","institution_ids":["https://openalex.org/I97018004"]}]},{"author_position":"middle","author":{"id":"https://openalex.org/A5058444456","display_name":"Athinagoras Skiadopoulos","orcid":"https://orcid.org/0000-0003-2933-0380"},"institutions":[{"id":"https://openalex.org/I97018004","display_name":"Stanford University","ror":"https://ror.org/00f54p054","country_code":"US","type":"education","lineage":["https://openalex.org/I97018004"]}],"countries":["US"],"is_corresponding":false,"raw_author_name":"Athinagoras Skiadopoulos","raw_affiliation_strings":["Stanford University, Stanford, CA, USA"],"raw_orcid":"https://orcid.org/0000-0003-2933-0380","affiliations":[{"raw_affiliation_string":"Stanford University, Stanford, CA, USA","institution_ids":["https://openalex.org/I97018004"]}]},{"author_position":"last","author":{"id":"https://openalex.org/A5042148531","display_name":"Christos Kozyrakis","orcid":"https://orcid.org/0000-0002-3154-7530"},"institutions":[{"id":"https://openalex.org/I97018004","display_name":"Stanford University","ror":"https://ror.org/00f54p054","country_code":"US","type":"education","lineage":["https://openalex.org/I97018004"]}],"countries":["US"],"is_corresponding":false,"raw_author_name":"Christos Kozyrakis","raw_affiliation_strings":["Stanford University, Stanford, CA, US"],"raw_orcid":"https://orcid.org/0000-0002-3154-7530","affiliations":[{"raw_affiliation_string":"Stanford University, Stanford, CA, US","institution_ids":["https://openalex.org/I97018004"]}]}],"institutions":[],"countries_distinct_count":1,"institutions_distinct_count":1,"corresponding_author_ids":[],"corresponding_institution_ids":["https://openalex.org/I97018004"],"apc_list":null,"apc_paid":null,"fwci":18.3894,"has_fulltext":true,"cited_by_count":18,"citation_normalized_percentile":{"value":0.99647826,"is_in_top_1_percent":true,"is_in_top_10_percent":true},"cited_by_percentile_year":{"min":99,"max":100},"biblio":{"volume":null,"issue":null,"first_page":"211","last_page":"228"},"is_retracted":false,"is_paratext":false,"is_xpac":false,"primary_topic":{"id":"https://openalex.org/T11932","display_name":"Wireless Body Area Networks","score":0.9433000087738037,"subfield":{"id":"https://openalex.org/subfields/2204","display_name":"Biomedical Engineering"},"field":{"id":"https://openalex.org/fields/22","display_name":"Engineering"},"domain":{"id":"https://openalex.org/domains/3","display_name":"Physical Sciences"}},"topics":[{"id":"https://openalex.org/T11932","display_name":"Wireless Body Area Networks","score":0.9433000087738037,"subfield":{"id":"https://openalex.org/subfields/2204","display_name":"Biomedical Engineering"},"field":{"id":"https://openalex.org/fields/22","display_name":"Engineering"},"domain":{"id":"https://openalex.org/domains/3","display_name":"Physical Sciences"}},{"id":"https://openalex.org/T10444","display_name":"Context-Aware Activity Recognition Systems","score":0.9330999851226807,"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/T10273","display_name":"IoT and Edge/Fog Computing","score":0.9164000153541565,"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/training","display_name":"Training (meteorology)","score":0.8272995948791504},{"id":"https://openalex.org/keywords/computer-science","display_name":"Computer science","score":0.46346914768218994},{"id":"https://openalex.org/keywords/pipeline-transport","display_name":"Pipeline transport","score":0.4378371238708496},{"id":"https://openalex.org/keywords/aeronautics","display_name":"Aeronautics","score":0.4174150228500366},{"id":"https://openalex.org/keywords/environmental-science","display_name":"Environmental science","score":0.1961253583431244},{"id":"https://openalex.org/keywords/engineering","display_name":"Engineering","score":0.19453927874565125},{"id":"https://openalex.org/keywords/meteorology","display_name":"Meteorology","score":0.13729262351989746},{"id":"https://openalex.org/keywords/geography","display_name":"Geography","score":0.1106647253036499}],"concepts":[{"id":"https://openalex.org/C2777211547","wikidata":"https://www.wikidata.org/wiki/Q17141490","display_name":"Training (meteorology)","level":2,"score":0.8272995948791504},{"id":"https://openalex.org/C41008148","wikidata":"https://www.wikidata.org/wiki/Q21198","display_name":"Computer science","level":0,"score":0.46346914768218994},{"id":"https://openalex.org/C175309249","wikidata":"https://www.wikidata.org/wiki/Q725864","display_name":"Pipeline transport","level":2,"score":0.4378371238708496},{"id":"https://openalex.org/C178802073","wikidata":"https://www.wikidata.org/wiki/Q8421","display_name":"Aeronautics","level":1,"score":0.4174150228500366},{"id":"https://openalex.org/C39432304","wikidata":"https://www.wikidata.org/wiki/Q188847","display_name":"Environmental science","level":0,"score":0.1961253583431244},{"id":"https://openalex.org/C127413603","wikidata":"https://www.wikidata.org/wiki/Q11023","display_name":"Engineering","level":0,"score":0.19453927874565125},{"id":"https://openalex.org/C153294291","wikidata":"https://www.wikidata.org/wiki/Q25261","display_name":"Meteorology","level":1,"score":0.13729262351989746},{"id":"https://openalex.org/C205649164","wikidata":"https://www.wikidata.org/wiki/Q1071","display_name":"Geography","level":0,"score":0.1106647253036499},{"id":"https://openalex.org/C87717796","wikidata":"https://www.wikidata.org/wiki/Q146326","display_name":"Environmental engineering","level":1,"score":0.0}],"mesh":[],"locations_count":2,"locations":[{"id":"doi:10.1145/3694715.3695960","is_oa":true,"landing_page_url":"https://doi.org/10.1145/3694715.3695960","pdf_url":"https://dl.acm.org/doi/pdf/10.1145/3694715.3695960","source":null,"license":null,"license_id":null,"version":"publishedVersion","is_accepted":true,"is_published":true,"raw_source_name":"Proceedings of the ACM SIGOPS 30th Symposium on Operating Systems Principles","raw_type":"proceedings-article"},{"id":"pmh:oai:arXiv.org:2405.14009","is_oa":true,"landing_page_url":"http://arxiv.org/abs/2405.14009","pdf_url":"https://arxiv.org/pdf/2405.14009","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":"other-oa","license_id":"https://openalex.org/licenses/other-oa","version":"submittedVersion","is_accepted":false,"is_published":false,"raw_source_name":null,"raw_type":"text"}],"best_oa_location":{"id":"doi:10.1145/3694715.3695960","is_oa":true,"landing_page_url":"https://doi.org/10.1145/3694715.3695960","pdf_url":"https://dl.acm.org/doi/pdf/10.1145/3694715.3695960","source":null,"license":null,"license_id":null,"version":"publishedVersion","is_accepted":true,"is_published":true,"raw_source_name":"Proceedings of the ACM SIGOPS 30th Symposium on Operating Systems Principles","raw_type":"proceedings-article"},"sustainable_development_goals":[],"awards":[{"id":"https://openalex.org/G1452086972","display_name":null,"funder_award_id":"JUMP 2.0","funder_id":"https://openalex.org/F4320306087","funder_display_name":"Semiconductor Research Corporation"}],"funders":[{"id":"https://openalex.org/F4320306087","display_name":"Semiconductor Research Corporation","ror":"https://ror.org/047z4n946"},{"id":"https://openalex.org/F4320332180","display_name":"Defense Advanced Research Projects Agency","ror":"https://ror.org/02caytj08"}],"has_content":{"pdf":true,"grobid_xml":false},"content_urls":{"pdf":"https://content.openalex.org/works/W4398796293.pdf"},"referenced_works_count":45,"referenced_works":["https://openalex.org/W1598866093","https://openalex.org/W2147504831","https://openalex.org/W2160815625","https://openalex.org/W2618530766","https://openalex.org/W2622263826","https://openalex.org/W2755682530","https://openalex.org/W2767346922","https://openalex.org/W2798515322","https://openalex.org/W2883830791","https://openalex.org/W2889565680","https://openalex.org/W2908510526","https://openalex.org/W2940920241","https://openalex.org/W2962684017","https://openalex.org/W2963351145","https://openalex.org/W2969388332","https://openalex.org/W2981937105","https://openalex.org/W3001279689","https://openalex.org/W3012897490","https://openalex.org/W3019389960","https://openalex.org/W3040573126","https://openalex.org/W3081168214","https://openalex.org/W3086105743","https://openalex.org/W3129831491","https://openalex.org/W3195577433","https://openalex.org/W3204998121","https://openalex.org/W3205803342","https://openalex.org/W3206832494","https://openalex.org/W4220741164","https://openalex.org/W4221167110","https://openalex.org/W4225591000","https://openalex.org/W4226335648","https://openalex.org/W4229005866","https://openalex.org/W4284974526","https://openalex.org/W4287391717","https://openalex.org/W4292779060","https://openalex.org/W4301239768","https://openalex.org/W4308760226","https://openalex.org/W4362655494","https://openalex.org/W4366999304","https://openalex.org/W4384918448","https://openalex.org/W4386348101","https://openalex.org/W4386840193","https://openalex.org/W4387302750","https://openalex.org/W4390529129","https://openalex.org/W6891860974"],"related_works":["https://openalex.org/W4391375266","https://openalex.org/W2748952813","https://openalex.org/W230091440","https://openalex.org/W2390279801","https://openalex.org/W2233261550","https://openalex.org/W2358668433","https://openalex.org/W4396701345","https://openalex.org/W2810751659","https://openalex.org/W258997015","https://openalex.org/W2376932109"],"abstract_inverted_index":{"Training":[0],"large":[1],"Deep":[2],"Neural":[3],"Network":[4],"(DNN)":[5],"models":[6],"requires":[7],"thousands":[8],"of":[9,14,61],"GPUs":[10],"over":[11],"the":[12,59,70,85,144,152,163,171,182],"course":[13],"several":[15],"days":[16],"or":[17,194],"weeks.":[18],"At":[19],"this":[20,116],"scale,":[21],"failures":[22],"are":[23,141],"frequent":[24],"and":[25,90,165,208,228,234],"can":[26,118],"have":[27],"a":[28,51,204],"big":[29],"impact":[30],"on":[31,65],"training":[32,57,76,111,127,197,214,224],"throughput.":[33,128],"Utilizing":[34],"spare":[35,66],"GPU":[36],"servers":[37,80,99],"to":[38,105,125,232],"mitigate":[39],"performance":[40],"loss":[41],"becomes":[42],"increasingly":[43],"costly":[44],"as":[45,226],"model":[46,87],"sizes":[47],"grow.":[48],"ReCycle":[49,101,132,207],"is":[50],"system":[52],"designed":[53],"for":[54,109,159,162,167,170,206,222],"efficient":[55],"DNN":[56],"in":[58,74],"presence":[60],"failures,":[62,218],"without":[63],"relying":[64],"servers.":[67],"It":[68],"exploits":[69],"inherent":[71],"functional":[72],"redundancy":[73],"distributed":[75],"systems":[77],"-":[78,89],"where":[79],"across":[81,121,177],"data-parallel":[82,96,106],"groups":[83],"store":[84],"same":[86],"parameters":[88],"pipeline":[91,122,146,178,190],"schedule":[92],"bubbles":[93],"within":[94,143],"each":[95],"group.":[97],"When":[98],"fails,":[100],"dynamically":[102],"re-routes":[103],"microbatches":[104,140],"peers,":[107],"allowing":[108],"uninterrupted":[110],"despite":[112],"multiple":[113,217],"failures.":[114,201],"However,":[115],"re-routing":[117],"create":[119],"imbalances":[120],"stages,":[123],"leading":[124],"reduced":[126],"To":[129],"address":[130],"this,":[131],"introduces":[133],"two":[134,156],"key":[135],"optimizations":[136,187],"that":[137,192,210],"ensure":[138],"re-routed":[139],"processed":[142],"original":[145],"schedule's":[147],"bubbles.":[148],"First,":[149],"it":[150,174,211],"decouples":[151],"backward":[153],"pass":[154],"into":[155],"phases:":[157],"one":[158],"computing":[160],"gradients":[161,169],"input":[164],"another":[166],"calculating":[168],"parameters.":[172],"Second,":[173],"avoids":[175],"synchronization":[176],"stages":[179],"by":[180,230],"staggering":[181],"optimizer":[183],"step.":[184],"Together,":[185],"these":[186],"enable":[188],"adaptive":[189],"schedules":[191],"minimize":[193],"even":[195],"eliminate":[196],"throughput":[198,215],"degradation":[199],"during":[200],"We":[202],"describe":[203],"prototype":[205],"show":[209],"achieves":[212],"high":[213],"under":[216],"outperforming":[219],"recent":[220],"proposals":[221],"fault-tolerant":[223],"such":[225],"Oobleck":[227],"Bamboo":[229],"up":[231],"1.46\u00d7":[233],"1.64\u00d7,":[235],"respectively.":[236]},"counts_by_year":[{"year":2026,"cited_by_count":8},{"year":2025,"cited_by_count":10}],"updated_date":"2026-08-11T07:18:39.950985","created_date":"2025-10-10T00:00:00"}