{"id":"https://openalex.org/W2963964896","doi":"https://doi.org/10.1109/icdm.2016.0028","title":"Model Accuracy and Runtime Tradeoff in Distributed Deep Learning: A Systematic Study","display_name":"Model Accuracy and Runtime Tradeoff in Distributed Deep Learning: A Systematic Study","alternate_titles":[],"publication_year":2016,"publication_date":"2016-12-01","ids":{"openalex":"https://openalex.org/W2963964896","doi":"https://doi.org/10.1109/icdm.2016.0028","mag":"2963964896"},"language":"en","primary_location":{"id":"doi:10.1109/icdm.2016.0028","doi":"https://doi.org/10.1109/icdm.2016.0028","is_oa":false,"landing_page_url":"https://doi.org/10.1109/icdm.2016.0028","pdf_url":null,"source":{"id":"https://openalex.org/S4210214883","display_name":"Proceedings","issn_l":"1550-4786","issn":["1550-4786","2374-8486"],"is_oa":false,"is_in_doaj":false,"is_core":false,"listed_in":["norway-1"],"host_organization":"https://openalex.org/P4310319808","host_organization_name":"Institute of Electrical and Electronics Engineers","host_organization_lineage":["https://openalex.org/P4310319808"],"host_organization_lineage_names":["Institute of Electrical and Electronics Engineers"],"type":"conference"},"license":null,"license_id":null,"version":"publishedVersion","is_accepted":true,"is_published":true,"raw_source_name":"2016 IEEE 16th International Conference on Data Mining (ICDM)","raw_type":"proceedings-article"},"parent_works":[],"lineage":["https://openalex.org/W2963964896"],"child_works_count":0,"type":"conference-paper","is_traditional":true,"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/A5110305071","display_name":"Suyog Gupta","orcid":null,"observed_orcids":[]},"institutions":[{"id":"https://openalex.org/I4210114115","display_name":"IBM Research - Thomas J. Watson Research Center","ror":"https://ror.org/0265w5591","country_code":"US","type":"facility","lineage":["https://openalex.org/I1341412227","https://openalex.org/I4210114115"]}],"countries":["US"],"is_corresponding":false,"contributor_roles":[],"raw_author_name":"Suyog Gupta","raw_affiliation_strings":["IBM T. J. Watson Research Center, Yorktown Heights, NY"],"raw_orcid":null,"affiliations":[{"raw_affiliation_string":"IBM T. J. Watson Research Center, Yorktown Heights, NY","institution_ids":["https://openalex.org/I4210114115"]}]},{"author_position":"middle","author":{"id":"https://openalex.org/A5100651774","display_name":"Wei Zhang","orcid":"https://orcid.org/0000-0001-5447-7793","observed_orcids":["https://orcid.org/0000-0001-5447-7793"]},"institutions":[{"id":"https://openalex.org/I4210114115","display_name":"IBM Research - Thomas J. Watson Research Center","ror":"https://ror.org/0265w5591","country_code":"US","type":"facility","lineage":["https://openalex.org/I1341412227","https://openalex.org/I4210114115"]}],"countries":["US"],"is_corresponding":false,"contributor_roles":[],"raw_author_name":"Wei Zhang","raw_affiliation_strings":["IBM T. J. Watson Research Center, Yorktown Heights, NY"],"raw_orcid":null,"affiliations":[{"raw_affiliation_string":"IBM T. J. Watson Research Center, Yorktown Heights, NY","institution_ids":["https://openalex.org/I4210114115"]}]},{"author_position":"last","author":{"id":"https://openalex.org/A5100455788","display_name":"Fei Wang","orcid":"https://orcid.org/0000-0002-1024-5867","observed_orcids":["https://orcid.org/0000-0002-1024-5867"]},"institutions":[{"id":"https://openalex.org/I4387153466","display_name":"Weill Cornell Medicine","ror":"https://ror.org/02r109517","country_code":"US","type":"education","lineage":["https://openalex.org/I205783295","https://openalex.org/I4387153466"]}],"countries":["US"],"is_corresponding":false,"contributor_roles":[],"raw_author_name":"Fei Wang","raw_affiliation_strings":["Department of Healthcare Policy and Research, Weill Cornell Medical College, New York City, NY"],"raw_orcid":null,"affiliations":[{"raw_affiliation_string":"Department of Healthcare Policy and Research, Weill Cornell Medical College, New York City, NY","institution_ids":["https://openalex.org/I4387153466"]}]}],"institutions":[],"countries_distinct_count":1,"institutions_distinct_count":2,"corresponding_author_ids":[],"corresponding_institution_ids":[],"apc_list":null,"apc_paid":null,"fwci":6.8273,"has_fulltext":false,"cited_by_count":168,"citation_normalized_percentile":{"value":0.97986235,"is_in_top_1_percent":false,"is_in_top_10_percent":true},"cited_by_percentile_year":{"min":87,"max":100},"biblio":{"volume":"2016","issue":null,"first_page":"171","last_page":"180"},"is_retracted":false,"is_pulled":false,"is_paratext":false,"is_xpac":false,"primary_topic":{"id":"https://openalex.org/T11612","display_name":"Stochastic Gradient Optimization Techniques","score":0.9082000255584717,"subfield":{"id":"https://openalex.org/subfields/1702","display_name":"Artificial Intelligence"},"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/T11612","display_name":"Stochastic Gradient Optimization Techniques","score":0.9082000255584717,"subfield":{"id":"https://openalex.org/subfields/1702","display_name":"Artificial Intelligence"},"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":0.04270000010728836,"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/T10101","display_name":"Cloud Computing and Resource Management","score":0.016200000420212746,"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/distributed-deep-learning","display_name":"distributed deep learning","score":0.9919999837875366},{"id":"https://openalex.org/keywords/parameter-server","display_name":"parameter server","score":0.7390000224113464},{"id":"https://openalex.org/keywords/asynchronous-stochastic-gradient-descent","display_name":"asynchronous stochastic gradient descent","score":0.9369999766349792},{"id":"https://openalex.org/keywords/synchronization-protocols","display_name":"synchronization protocols","score":0.7559999823570251},{"id":"https://openalex.org/keywords/learning-rate-modulation","display_name":"learning rate modulation","score":0.9440000057220459},{"id":"https://openalex.org/keywords/mini-batch-size","display_name":"mini-batch size","score":0.9440000057220459},{"id":"https://openalex.org/keywords/large-scale-neural-networks","display_name":"large-scale neural networks","score":0.7630000114440918}],"concepts":[{"id":"https://openalex.org/C41008148","wikidata":"https://www.wikidata.org/wiki/Q21198","display_name":"Computer science","level":0,"score":0.8589277267456055},{"id":"https://openalex.org/C151319957","wikidata":"https://www.wikidata.org/wiki/Q752739","display_name":"Asynchronous communication","level":2,"score":0.819416880607605},{"id":"https://openalex.org/C206688291","wikidata":"https://www.wikidata.org/wiki/Q7617819","display_name":"Stochastic gradient descent","level":3,"score":0.7453237771987915},{"id":"https://openalex.org/C2778562939","wikidata":"https://www.wikidata.org/wiki/Q1298791","display_name":"Synchronization (alternating current)","level":3,"score":0.7003160715103149},{"id":"https://openalex.org/C2780385302","wikidata":"https://www.wikidata.org/wiki/Q367158","display_name":"Protocol (science)","level":3,"score":0.6192167401313782},{"id":"https://openalex.org/C108583219","wikidata":"https://www.wikidata.org/wiki/Q197536","display_name":"Deep learning","level":2,"score":0.5150172710418701},{"id":"https://openalex.org/C98045186","wikidata":"https://www.wikidata.org/wiki/Q205663","display_name":"Process (computing)","level":2,"score":0.5105829834938049},{"id":"https://openalex.org/C154945302","wikidata":"https://www.wikidata.org/wiki/Q11660","display_name":"Artificial intelligence","level":1,"score":0.49922776222229004},{"id":"https://openalex.org/C119857082","wikidata":"https://www.wikidata.org/wiki/Q2539","display_name":"Machine learning","level":1,"score":0.49112841486930847},{"id":"https://openalex.org/C120314980","wikidata":"https://www.wikidata.org/wiki/Q180634","display_name":"Distributed computing","level":1,"score":0.4775489270687103},{"id":"https://openalex.org/C50644808","wikidata":"https://www.wikidata.org/wiki/Q192776","display_name":"Artificial neural network","level":2,"score":0.4322965145111084},{"id":"https://openalex.org/C31258907","wikidata":"https://www.wikidata.org/wiki/Q1301371","display_name":"Computer network","level":1,"score":0.1361408829689026},{"id":"https://openalex.org/C111919701","wikidata":"https://www.wikidata.org/wiki/Q9135","display_name":"Operating system","level":1,"score":0.09673985838890076},{"id":"https://openalex.org/C127162648","wikidata":"https://www.wikidata.org/wiki/Q16858953","display_name":"Channel (broadcasting)","level":2,"score":0.0},{"id":"https://openalex.org/C142724271","wikidata":"https://www.wikidata.org/wiki/Q7208","display_name":"Pathology","level":1,"score":0.0},{"id":"https://openalex.org/C204787440","wikidata":"https://www.wikidata.org/wiki/Q188504","display_name":"Alternative medicine","level":2,"score":0.0},{"id":"https://openalex.org/C71924100","wikidata":"https://www.wikidata.org/wiki/Q11190","display_name":"Medicine","level":0,"score":0.0}],"mesh":[],"locations_count":2,"locations":[{"id":"doi:10.1109/icdm.2016.0028","doi":"https://doi.org/10.1109/icdm.2016.0028","is_oa":false,"landing_page_url":"https://doi.org/10.1109/icdm.2016.0028","pdf_url":null,"source":{"id":"https://openalex.org/S4210214883","display_name":"Proceedings","issn_l":"1550-4786","issn":["1550-4786","2374-8486"],"is_oa":false,"is_in_doaj":false,"is_core":false,"listed_in":["norway-1"],"host_organization":"https://openalex.org/P4310319808","host_organization_name":"Institute of Electrical and Electronics Engineers","host_organization_lineage":["https://openalex.org/P4310319808"],"host_organization_lineage_names":["Institute of Electrical and Electronics Engineers"],"type":"conference"},"license":null,"license_id":null,"version":"publishedVersion","is_accepted":true,"is_published":true,"raw_source_name":"2016 IEEE 16th International Conference on Data Mining (ICDM)","raw_type":"proceedings-article"},{"id":"mag:2752943650","doi":null,"is_oa":false,"landing_page_url":"http://jglobal.jst.go.jp/public/20090422/201702253125691630","pdf_url":null,"source":{"id":"https://openalex.org/S4306512817","display_name":"IEEE Conference Proceedings","issn_l":null,"issn":null,"is_oa":false,"is_in_doaj":false,"is_core":false,"listed_in":[],"host_organization":null,"host_organization_name":null,"host_organization_lineage":[],"host_organization_lineage_names":[],"type":"conference"},"license":null,"license_id":null,"version":null,"is_accepted":false,"is_published":false,"raw_source_name":"IEEE Conference Proceedings","raw_type":null}],"best_oa_location":null,"sustainable_development_goals":[],"study_designs":[],"awards":[],"funders":[],"has_content":{"pdf":false,"grobid_xml":false},"content_urls":null,"referenced_works_count":25,"referenced_works":["https://openalex.org/W6908809","https://openalex.org/W104184427","https://openalex.org/W778657980","https://openalex.org/W1442374986","https://openalex.org/W1533861849","https://openalex.org/W1677182931","https://openalex.org/W1806891645","https://openalex.org/W1814095264","https://openalex.org/W1833917188","https://openalex.org/W1836465849","https://openalex.org/W1845277745","https://openalex.org/W2000200144","https://openalex.org/W2041517243","https://openalex.org/W2091590079","https://openalex.org/W2117539524","https://openalex.org/W2132737349","https://openalex.org/W2146502635","https://openalex.org/W2155893237","https://openalex.org/W2162390675","https://openalex.org/W2163605009","https://openalex.org/W2168231600","https://openalex.org/W2265775419","https://openalex.org/W2919115771","https://openalex.org/W3118608800","https://openalex.org/W3137220996"],"related_works":["https://openalex.org/W2739720758","https://openalex.org/W2982664135","https://openalex.org/W4205519540","https://openalex.org/W2898323977","https://openalex.org/W4283694394","https://openalex.org/W7203724885","https://openalex.org/W4404180200","https://openalex.org/W2971375861","https://openalex.org/W2737583410","https://openalex.org/W3132016054"],"abstract_inverted_index":{"Deep":[0],"learning":[1,73],"with":[2],"a":[3,86,99,120],"large":[4],"number":[5,76],"of":[6,27,65,77,95,126],"parametersrequires":[7],"distributed":[8,45,124],"training,":[9],"where":[10],"model":[11,35,82,115],"accuracy":[12],"and":[13,75,81,156],"runtimeare":[14],"two":[15,32],"important":[16],"factors":[17],"to":[18,91,140,143],"be":[19,134],"considered.":[20],"However,":[21],"there":[22],"hasbeen":[23],"no":[24],"systematic":[25],"study":[26],"the":[28,34,93,106,128,141,145],"tradeoff":[29],"between":[30],"these":[31],"factorsduring":[33],"training":[36,49,125],"process.":[37],"This":[38],"paper":[39],"presents":[40],"Rudra,":[41],"aparameter":[42],"server":[43],"based":[44],"computing":[46],"framework":[47],"tunedfor":[48],"large-scale":[50],"deep":[51],"neural":[52],"networks.":[53],"Using":[54],"variants":[55],"ofthe":[56],"asynchronous":[57],"stochastic":[58],"gradient":[59,69],"descent":[60],"algorithm":[61],"we":[62],"studythe":[63],"impact":[64],"synchronization":[66,101],"protocol,":[67],"stale":[68,96],"updates,":[70],"minibatch":[71],"size,":[72],"rates,":[74],"learners":[78,137],"on":[79],"runtimeperformance":[80],"accuracy.":[83,116],"We":[84,147],"introduce":[85],"new":[87,100],"learningrate":[88],"modulation":[89],"strategy":[90],"counter":[92],"effect":[94],"gradientsand":[97],"propose":[98],"protocol":[102],"that":[103],"can":[104],"effectivelybound":[105],"staleness":[107],"in":[108],"gradients,":[109],"improve":[110],"runtime":[111],"performanceand":[112],"achieve":[113],"good":[114],"Our":[117],"empirical":[118],"investigationreveals":[119],"principled":[121],"approach":[122,150],"for":[123],"neuralnetworks:":[127],"mini-batch":[129],"size":[130],"per":[131],"learner":[132],"should":[133],"reducedas":[135],"more":[136],"are":[138],"added":[139],"system":[142],"preserve":[144],"modelaccuracy.":[146],"validate":[148],"this":[149],"using":[151],"commonly-used":[152],"imageclassification":[153],"benchmarks:":[154],"CIFAR10":[155],"ImageNet.":[157]},"cited_by_api_url":"https://api.openalex.org/works?filter=cites:W2963964896","counts_by_year":[{"year":2026,"cited_by_count":5},{"year":2025,"cited_by_count":10},{"year":2024,"cited_by_count":13},{"year":2023,"cited_by_count":16},{"year":2022,"cited_by_count":22},{"year":2021,"cited_by_count":21},{"year":2020,"cited_by_count":27},{"year":2019,"cited_by_count":26},{"year":2018,"cited_by_count":17},{"year":2017,"cited_by_count":10},{"year":2016,"cited_by_count":1}],"updated_date":"2026-11-21T10:27:49.728292","created_date":"2019-07-30T00:00:00"}
