{"id":"https://openalex.org/W2981114289","doi":"https://doi.org/10.1147/jrd.2019.2947013","title":"BlueConnect: Decomposing all-reduce for deep learning on heterogeneous network hierarchy","display_name":"BlueConnect: Decomposing all-reduce for deep learning on heterogeneous network hierarchy","publication_year":2019,"publication_date":"2019-10-14","ids":{"openalex":"https://openalex.org/W2981114289","doi":"https://doi.org/10.1147/jrd.2019.2947013","mag":"2981114289"},"language":"en","primary_location":{"id":"doi:10.1147/jrd.2019.2947013","is_oa":false,"landing_page_url":"https://doi.org/10.1147/jrd.2019.2947013","pdf_url":null,"source":{"id":"https://openalex.org/S4210219925","display_name":"IBM Journal of Research and Development","issn_l":"0018-8646","issn":["0018-8646","2151-8556"],"is_oa":false,"is_in_doaj":false,"is_core":true,"host_organization":"https://openalex.org/P4310320652","host_organization_name":"IBM","host_organization_lineage":["https://openalex.org/P4310320652"],"host_organization_lineage_names":["IBM"],"type":"journal"},"license":null,"license_id":null,"version":"publishedVersion","is_accepted":true,"is_published":true,"raw_source_name":"IBM Journal of Research and Development","raw_type":"journal-article"},"type":"article","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/A5102977691","display_name":"Min Chul Cho","orcid":"https://orcid.org/0000-0001-7288-1706"},"institutions":[],"countries":[],"is_corresponding":false,"raw_author_name":"M. Cho","raw_affiliation_strings":[],"raw_orcid":null,"affiliations":[]},{"author_position":"middle","author":{"id":"https://openalex.org/A5000286196","display_name":"Ulrich Finkler","orcid":null},"institutions":[],"countries":[],"is_corresponding":false,"raw_author_name":"U. Finkler","raw_affiliation_strings":[],"raw_orcid":null,"affiliations":[]},{"author_position":"middle","author":{"id":"https://openalex.org/A5042998510","display_name":"Maur\u00edcio Serrano","orcid":"https://orcid.org/0000-0003-0250-5881"},"institutions":[],"countries":[],"is_corresponding":false,"raw_author_name":"M. Serrano","raw_affiliation_strings":[],"raw_orcid":null,"affiliations":[]},{"author_position":"middle","author":{"id":"https://openalex.org/A5063522852","display_name":"David S. Kung","orcid":null},"institutions":[],"countries":[],"is_corresponding":false,"raw_author_name":"D. Kung","raw_affiliation_strings":[],"raw_orcid":null,"affiliations":[]},{"author_position":"last","author":{"id":"https://openalex.org/A5030648912","display_name":"Hillery C. Hunter","orcid":null},"institutions":[],"countries":[],"is_corresponding":false,"raw_author_name":"H. Hunter","raw_affiliation_strings":[],"raw_orcid":null,"affiliations":[]}],"institutions":[],"countries_distinct_count":0,"institutions_distinct_count":0,"corresponding_author_ids":[],"corresponding_institution_ids":[],"apc_list":null,"apc_paid":null,"fwci":3.7547,"has_fulltext":false,"cited_by_count":88,"citation_normalized_percentile":{"value":0.94281269,"is_in_top_1_percent":false,"is_in_top_10_percent":true},"cited_by_percentile_year":{"min":99,"max":100},"biblio":{"volume":"63","issue":"6","first_page":"1:1","last_page":"1:11"},"is_retracted":false,"is_paratext":false,"is_xpac":false,"primary_topic":{"id":"https://openalex.org/T10036","display_name":"Advanced Neural Network Applications","score":0.9997000098228455,"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"}},"topics":[{"id":"https://openalex.org/T10036","display_name":"Advanced Neural Network Applications","score":0.9997000098228455,"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/T10502","display_name":"Advanced Memory and Neural Computing","score":0.9988999962806702,"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"}},{"id":"https://openalex.org/T12808","display_name":"Ferroelectric and Negative Capacitance Devices","score":0.9973000288009644,"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":0.8197960257530212},{"id":"https://openalex.org/keywords/exploit","display_name":"Exploit","score":0.7793641686439514},{"id":"https://openalex.org/keywords/deep-learning","display_name":"Deep learning","score":0.6759104132652283},{"id":"https://openalex.org/keywords/artificial-intelligence","display_name":"Artificial intelligence","score":0.5762236714363098},{"id":"https://openalex.org/keywords/distributed-computing","display_name":"Distributed computing","score":0.5406650304794312},{"id":"https://openalex.org/keywords/overhead","display_name":"Overhead (engineering)","score":0.5399461388587952},{"id":"https://openalex.org/keywords/artificial-neural-network","display_name":"Artificial neural network","score":0.5126561522483826},{"id":"https://openalex.org/keywords/margin","display_name":"Margin (machine learning)","score":0.5125970244407654},{"id":"https://openalex.org/keywords/residual-neural-network","display_name":"Residual neural network","score":0.4640582799911499},{"id":"https://openalex.org/keywords/convolutional-neural-network","display_name":"Convolutional neural network","score":0.45844224095344543},{"id":"https://openalex.org/keywords/low-latency","display_name":"Low latency (capital markets)","score":0.4477795660495758},{"id":"https://openalex.org/keywords/latency","display_name":"Latency (audio)","score":0.42935997247695923},{"id":"https://openalex.org/keywords/bandwidth","display_name":"Bandwidth (computing)","score":0.4233691096305847},{"id":"https://openalex.org/keywords/parallelizable-manifold","display_name":"Parallelizable manifold","score":0.41749876737594604},{"id":"https://openalex.org/keywords/computer-engineering","display_name":"Computer engineering","score":0.4008384644985199},{"id":"https://openalex.org/keywords/computer-architecture","display_name":"Computer architecture","score":0.3992428779602051},{"id":"https://openalex.org/keywords/machine-learning","display_name":"Machine learning","score":0.3388601541519165},{"id":"https://openalex.org/keywords/computer-network","display_name":"Computer network","score":0.19703948497772217},{"id":"https://openalex.org/keywords/telecommunications","display_name":"Telecommunications","score":0.12686562538146973},{"id":"https://openalex.org/keywords/algorithm","display_name":"Algorithm","score":0.0911523699760437},{"id":"https://openalex.org/keywords/operating-system","display_name":"Operating system","score":0.08612790703773499}],"concepts":[{"id":"https://openalex.org/C41008148","wikidata":"https://www.wikidata.org/wiki/Q21198","display_name":"Computer science","level":0,"score":0.8197960257530212},{"id":"https://openalex.org/C165696696","wikidata":"https://www.wikidata.org/wiki/Q11287","display_name":"Exploit","level":2,"score":0.7793641686439514},{"id":"https://openalex.org/C108583219","wikidata":"https://www.wikidata.org/wiki/Q197536","display_name":"Deep learning","level":2,"score":0.6759104132652283},{"id":"https://openalex.org/C154945302","wikidata":"https://www.wikidata.org/wiki/Q11660","display_name":"Artificial intelligence","level":1,"score":0.5762236714363098},{"id":"https://openalex.org/C120314980","wikidata":"https://www.wikidata.org/wiki/Q180634","display_name":"Distributed computing","level":1,"score":0.5406650304794312},{"id":"https://openalex.org/C2779960059","wikidata":"https://www.wikidata.org/wiki/Q7113681","display_name":"Overhead (engineering)","level":2,"score":0.5399461388587952},{"id":"https://openalex.org/C50644808","wikidata":"https://www.wikidata.org/wiki/Q192776","display_name":"Artificial neural network","level":2,"score":0.5126561522483826},{"id":"https://openalex.org/C774472","wikidata":"https://www.wikidata.org/wiki/Q6760393","display_name":"Margin (machine learning)","level":2,"score":0.5125970244407654},{"id":"https://openalex.org/C2944601119","wikidata":"https://www.wikidata.org/wiki/Q43744058","display_name":"Residual neural network","level":3,"score":0.4640582799911499},{"id":"https://openalex.org/C81363708","wikidata":"https://www.wikidata.org/wiki/Q17084460","display_name":"Convolutional neural network","level":2,"score":0.45844224095344543},{"id":"https://openalex.org/C46637626","wikidata":"https://www.wikidata.org/wiki/Q6693015","display_name":"Low latency (capital markets)","level":2,"score":0.4477795660495758},{"id":"https://openalex.org/C82876162","wikidata":"https://www.wikidata.org/wiki/Q17096504","display_name":"Latency (audio)","level":2,"score":0.42935997247695923},{"id":"https://openalex.org/C2776257435","wikidata":"https://www.wikidata.org/wiki/Q1576430","display_name":"Bandwidth (computing)","level":2,"score":0.4233691096305847},{"id":"https://openalex.org/C148047603","wikidata":"https://www.wikidata.org/wiki/Q1014612","display_name":"Parallelizable manifold","level":2,"score":0.41749876737594604},{"id":"https://openalex.org/C113775141","wikidata":"https://www.wikidata.org/wiki/Q428691","display_name":"Computer engineering","level":1,"score":0.4008384644985199},{"id":"https://openalex.org/C118524514","wikidata":"https://www.wikidata.org/wiki/Q173212","display_name":"Computer architecture","level":1,"score":0.3992428779602051},{"id":"https://openalex.org/C119857082","wikidata":"https://www.wikidata.org/wiki/Q2539","display_name":"Machine learning","level":1,"score":0.3388601541519165},{"id":"https://openalex.org/C31258907","wikidata":"https://www.wikidata.org/wiki/Q1301371","display_name":"Computer network","level":1,"score":0.19703948497772217},{"id":"https://openalex.org/C76155785","wikidata":"https://www.wikidata.org/wiki/Q418","display_name":"Telecommunications","level":1,"score":0.12686562538146973},{"id":"https://openalex.org/C11413529","wikidata":"https://www.wikidata.org/wiki/Q8366","display_name":"Algorithm","level":1,"score":0.0911523699760437},{"id":"https://openalex.org/C111919701","wikidata":"https://www.wikidata.org/wiki/Q9135","display_name":"Operating system","level":1,"score":0.08612790703773499},{"id":"https://openalex.org/C38652104","wikidata":"https://www.wikidata.org/wiki/Q3510521","display_name":"Computer security","level":1,"score":0.0}],"mesh":[],"locations_count":1,"locations":[{"id":"doi:10.1147/jrd.2019.2947013","is_oa":false,"landing_page_url":"https://doi.org/10.1147/jrd.2019.2947013","pdf_url":null,"source":{"id":"https://openalex.org/S4210219925","display_name":"IBM Journal of Research and Development","issn_l":"0018-8646","issn":["0018-8646","2151-8556"],"is_oa":false,"is_in_doaj":false,"is_core":true,"host_organization":"https://openalex.org/P4310320652","host_organization_name":"IBM","host_organization_lineage":["https://openalex.org/P4310320652"],"host_organization_lineage_names":["IBM"],"type":"journal"},"license":null,"license_id":null,"version":"publishedVersion","is_accepted":true,"is_published":true,"raw_source_name":"IBM Journal of Research and Development","raw_type":"journal-article"}],"best_oa_location":null,"sustainable_development_goals":[{"display_name":"Industry, innovation and infrastructure","id":"https://metadata.un.org/sdg/9","score":0.5199999809265137}],"awards":[],"funders":[],"has_content":{"grobid_xml":false,"pdf":false},"content_urls":null,"referenced_works_count":52,"referenced_works":["https://openalex.org/W1598866093","https://openalex.org/W1836465849","https://openalex.org/W1879670505","https://openalex.org/W1995653323","https://openalex.org/W2109573501","https://openalex.org/W2130531694","https://openalex.org/W2131613942","https://openalex.org/W2138243089","https://openalex.org/W2155893237","https://openalex.org/W2163605009","https://openalex.org/W2164945803","https://openalex.org/W2186615578","https://openalex.org/W2193413348","https://openalex.org/W2194775991","https://openalex.org/W2336650964","https://openalex.org/W2513383847","https://openalex.org/W2606722458","https://openalex.org/W2617960902","https://openalex.org/W2622263826","https://openalex.org/W2749988060","https://openalex.org/W2884711234","https://openalex.org/W2953384591","https://openalex.org/W2962747323","https://openalex.org/W2962911728","https://openalex.org/W2962950660","https://openalex.org/W2963804082","https://openalex.org/W2963959597","https://openalex.org/W3212066104","https://openalex.org/W4233717502","https://openalex.org/W4297785678","https://openalex.org/W4298312531","https://openalex.org/W6635810480","https://openalex.org/W6638667902","https://openalex.org/W6638803421","https://openalex.org/W6639392911","https://openalex.org/W6684191040","https://openalex.org/W6684953698","https://openalex.org/W6686509673","https://openalex.org/W6687566353","https://openalex.org/W6692488971","https://openalex.org/W6703420464","https://openalex.org/W6713134421","https://openalex.org/W6726983090","https://openalex.org/W6732127223","https://openalex.org/W6738250615","https://openalex.org/W6739622702","https://openalex.org/W6743289643","https://openalex.org/W6743955621","https://openalex.org/W6747968877","https://openalex.org/W6753209298","https://openalex.org/W6754143808","https://openalex.org/W6764309476"],"related_works":["https://openalex.org/W4285069850","https://openalex.org/W2891970004","https://openalex.org/W1996916724","https://openalex.org/W4240963716","https://openalex.org/W1974634278","https://openalex.org/W1596637634","https://openalex.org/W2321918740","https://openalex.org/W2154273227","https://openalex.org/W4377082852","https://openalex.org/W2057947873"],"abstract_inverted_index":{"As":[0],"deep":[1,22,32,66],"neural":[2,23],"networks":[3],"get":[4,10],"more":[5],"complex":[6],"and":[7,89,98,100,120,153],"input":[8],"datasets":[9],"larger,":[11],"it":[12,41],"can":[13,112,142,157],"take":[14,121],"days":[15],"or":[16],"even":[17],"weeks":[18,51],"to":[19,25,45,52,92,102,115,133],"train":[20],"a":[21,35,78,83,103,116,150],"network":[24,106,118],"the":[26,43,47,94,124,129,144,154],"desired":[27],"accuracy.":[28],"Therefore,":[29,108],"enabling":[30],"distributed":[31,65],"learning":[33,67],"at":[34],"massive":[36],"scale":[37],"is":[38,69],"critical":[39],"since":[40],"offers":[42],"potential":[44],"reduce":[46,159],"training":[48,169],"time":[49],"from":[50],"hours.":[53],"In":[54],"this":[55],"article,":[56],"we":[57],"present":[58],"BlueConnect,":[59],"an":[60],"efficient":[61],"communication":[62,147],"library":[63,148],"for":[64,72,128,167],"that":[68],"highly":[70],"optimized":[71],"popular":[73],"GPU-based":[74],"platforms.":[75],"BlueConnect":[76,141],"decomposes":[77],"single":[79],"all-reduce":[80],"operation":[81,111],"into":[82],"large":[84],"number":[85],"of":[86,105,123],"parallelizable":[87],"reduce\u2013scatter":[88],"all-gather":[90],"operations":[91],"exploit":[93],"tradeoff":[95],"between":[96],"latency":[97],"bandwidth":[99],"adapt":[101],"variety":[104],"configurations.":[107],"each":[109],"individual":[110],"be":[113],"mapped":[114],"different":[117],"fabric":[119],"advantage":[122],"best":[125],"performing":[126],"implementation":[127],"corresponding":[130],"fabric.":[131],"According":[132],"our":[134],"experimental":[135],"results":[136],"on":[137,164],"two":[138],"system":[139],"configurations,":[140],"outperform":[143],"leading":[145],"industrial":[146],"by":[149,162],"wide":[151],"margin,":[152],"BlueConnect-integrated":[155],"Caffe2":[156],"significantly":[158],"synchronization":[160],"overhead":[161],"87%":[163],"192":[165],"GPUs":[166],"Resnet-50":[168],"over":[170],"prior":[171],"schemes.":[172]},"counts_by_year":[{"year":2026,"cited_by_count":5},{"year":2025,"cited_by_count":11},{"year":2024,"cited_by_count":18},{"year":2023,"cited_by_count":21},{"year":2022,"cited_by_count":12},{"year":2021,"cited_by_count":12},{"year":2020,"cited_by_count":9}],"updated_date":"2026-07-15T18:14:33.161393","created_date":"2025-10-10T00:00:00"}