{"id":"https://openalex.org/W4283217358","doi":"https://doi.org/10.1109/infocom48880.2022.9796820","title":"Mercury: A Simple Transport Layer Scheduler to Accelerate Distributed DNN Training","display_name":"Mercury: A Simple Transport Layer Scheduler to Accelerate Distributed DNN Training","publication_year":2022,"publication_date":"2022-05-02","ids":{"openalex":"https://openalex.org/W4283217358","doi":"https://doi.org/10.1109/infocom48880.2022.9796820"},"language":"en","primary_location":{"id":"doi:10.1109/infocom48880.2022.9796820","is_oa":false,"landing_page_url":"https://doi.org/10.1109/infocom48880.2022.9796820","pdf_url":null,"source":null,"license":null,"license_id":null,"version":"publishedVersion","is_accepted":true,"is_published":true,"raw_source_name":"IEEE INFOCOM 2022 - IEEE Conference on Computer Communications","raw_type":"proceedings-article"},"type":"conference-paper","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/A5036041576","display_name":"Qingyang Duan","orcid":"https://orcid.org/0000-0002-9268-1979"},"institutions":[{"id":"https://openalex.org/I24943067","display_name":"Fudan University","ror":"https://ror.org/013q1eq08","country_code":"CN","type":"education","lineage":["https://openalex.org/I24943067"]}],"countries":["CN"],"is_corresponding":false,"raw_author_name":"Qingyang Duan","raw_affiliation_strings":["School of Information Science and Technology, Fudan University,Shanghai,China","School of Information Science and Technology, Fudan University, Shanghai, China"],"raw_orcid":null,"affiliations":[{"raw_affiliation_string":"School of Information Science and Technology, Fudan University,Shanghai,China","institution_ids":["https://openalex.org/I24943067"]},{"raw_affiliation_string":"School of Information Science and Technology, Fudan University, Shanghai, China","institution_ids":["https://openalex.org/I24943067"]}]},{"author_position":"middle","author":{"id":"https://openalex.org/A5067078448","display_name":"Zeqin Wang","orcid":null},"institutions":[{"id":"https://openalex.org/I24943067","display_name":"Fudan University","ror":"https://ror.org/013q1eq08","country_code":"CN","type":"education","lineage":["https://openalex.org/I24943067"]}],"countries":["CN"],"is_corresponding":false,"raw_author_name":"Zeqin Wang","raw_affiliation_strings":["School of Information Science and Technology, Fudan University,Shanghai,China","School of Information Science and Technology, Fudan University, Shanghai, China"],"raw_orcid":null,"affiliations":[{"raw_affiliation_string":"School of Information Science and Technology, Fudan University,Shanghai,China","institution_ids":["https://openalex.org/I24943067"]},{"raw_affiliation_string":"School of Information Science and Technology, Fudan University, Shanghai, China","institution_ids":["https://openalex.org/I24943067"]}]},{"author_position":"middle","author":{"id":"https://openalex.org/A5063792608","display_name":"Yuedong Xu","orcid":"https://orcid.org/0000-0003-4168-3998"},"institutions":[{"id":"https://openalex.org/I24943067","display_name":"Fudan University","ror":"https://ror.org/013q1eq08","country_code":"CN","type":"education","lineage":["https://openalex.org/I24943067"]}],"countries":["CN"],"is_corresponding":false,"raw_author_name":"Yuedong Xu","raw_affiliation_strings":["School of Information Science and Technology, Fudan University,Shanghai,China","School of Information Science and Technology, Fudan University, Shanghai, China"],"raw_orcid":null,"affiliations":[{"raw_affiliation_string":"School of Information Science and Technology, Fudan University,Shanghai,China","institution_ids":["https://openalex.org/I24943067"]},{"raw_affiliation_string":"School of Information Science and Technology, Fudan University, Shanghai, China","institution_ids":["https://openalex.org/I24943067"]}]},{"author_position":"middle","author":{"id":"https://openalex.org/A5083783069","display_name":"Shaoteng Liu","orcid":"https://orcid.org/0000-0001-5407-2905"},"institutions":[{"id":"https://openalex.org/I2250955327","display_name":"Huawei Technologies (China)","ror":"https://ror.org/00cmhce21","country_code":"CN","type":"company","lineage":["https://openalex.org/I2250955327"]}],"countries":["CN"],"is_corresponding":false,"raw_author_name":"Shaoteng Liu","raw_affiliation_strings":["Huawei Technologies Co. Ltd.,Shenzhen,China","Huawei Technologies Co. Ltd., Shenzhen, China"],"raw_orcid":null,"affiliations":[{"raw_affiliation_string":"Huawei Technologies Co. Ltd.,Shenzhen,China","institution_ids":["https://openalex.org/I2250955327"]},{"raw_affiliation_string":"Huawei Technologies Co. Ltd., Shenzhen, China","institution_ids":["https://openalex.org/I2250955327"]}]},{"author_position":"last","author":{"id":"https://openalex.org/A5075806425","display_name":"Jun Wu","orcid":"https://orcid.org/0000-0001-7090-8653"},"institutions":[{"id":"https://openalex.org/I24943067","display_name":"Fudan University","ror":"https://ror.org/013q1eq08","country_code":"CN","type":"education","lineage":["https://openalex.org/I24943067"]}],"countries":["CN"],"is_corresponding":false,"raw_author_name":"Jun Wu","raw_affiliation_strings":["School of Computer Science and Technology, Fudan University,Shanghai,China","School of Computer Science and Technology, Fudan University, Shanghai, China"],"raw_orcid":null,"affiliations":[{"raw_affiliation_string":"School of Computer Science and Technology, Fudan University,Shanghai,China","institution_ids":["https://openalex.org/I24943067"]},{"raw_affiliation_string":"School of Computer Science and Technology, Fudan University, Shanghai, China","institution_ids":["https://openalex.org/I24943067"]}]}],"institutions":[],"countries_distinct_count":1,"institutions_distinct_count":2,"corresponding_author_ids":[],"corresponding_institution_ids":[],"apc_list":null,"apc_paid":null,"fwci":0.9246,"has_fulltext":false,"cited_by_count":11,"citation_normalized_percentile":{"value":0.8342039,"is_in_top_1_percent":false,"is_in_top_10_percent":false},"cited_by_percentile_year":{"min":94,"max":98},"biblio":{"volume":null,"issue":null,"first_page":"350","last_page":"359"},"is_retracted":false,"is_paratext":false,"is_xpac":false,"primary_topic":{"id":"https://openalex.org/T10036","display_name":"Advanced Neural Network Applications","score":0.991599977016449,"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":0.991599977016449,"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/T12303","display_name":"Tensor decomposition and applications","score":0.9904999732971191,"subfield":{"id":"https://openalex.org/subfields/2605","display_name":"Computational Mathematics"},"field":{"id":"https://openalex.org/fields/26","display_name":"Mathematics"},"domain":{"id":"https://openalex.org/domains/3","display_name":"Physical Sciences"}},{"id":"https://openalex.org/T10054","display_name":"Parallel Computing and Optimization Techniques","score":0.9861999750137329,"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"}}],"keywords":[{"id":"https://openalex.org/keywords/computer-science","display_name":"Computer science","score":0.7968449592590332},{"id":"https://openalex.org/keywords/speedup","display_name":"Speedup","score":0.7848179340362549},{"id":"https://openalex.org/keywords/scheduling","display_name":"Scheduling (production processes)","score":0.6599645614624023},{"id":"https://openalex.org/keywords/granularity","display_name":"Granularity","score":0.6333301067352295},{"id":"https://openalex.org/keywords/computation","display_name":"Computation","score":0.6221250891685486},{"id":"https://openalex.org/keywords/network-packet","display_name":"Network packet","score":0.5186033248901367},{"id":"https://openalex.org/keywords/parallel-computing","display_name":"Parallel computing","score":0.5115073919296265},{"id":"https://openalex.org/keywords/mercury","display_name":"Mercury (programming language)","score":0.4986741542816162},{"id":"https://openalex.org/keywords/transport-layer","display_name":"Transport layer","score":0.45992952585220337},{"id":"https://openalex.org/keywords/distributed-computing","display_name":"Distributed computing","score":0.4583742618560791},{"id":"https://openalex.org/keywords/partition","display_name":"Partition (number theory)","score":0.4268642067909241},{"id":"https://openalex.org/keywords/computer-network","display_name":"Computer network","score":0.4143141806125641},{"id":"https://openalex.org/keywords/algorithm","display_name":"Algorithm","score":0.2234863042831421},{"id":"https://openalex.org/keywords/layer","display_name":"Layer (electronics)","score":0.17296147346496582},{"id":"https://openalex.org/keywords/mathematical-optimization","display_name":"Mathematical optimization","score":0.12139329314231873},{"id":"https://openalex.org/keywords/chemistry","display_name":"Chemistry","score":0.09412083029747009},{"id":"https://openalex.org/keywords/operating-system","display_name":"Operating system","score":0.08588188886642456}],"concepts":[{"id":"https://openalex.org/C41008148","wikidata":"https://www.wikidata.org/wiki/Q21198","display_name":"Computer science","level":0,"score":0.7968449592590332},{"id":"https://openalex.org/C68339613","wikidata":"https://www.wikidata.org/wiki/Q1549489","display_name":"Speedup","level":2,"score":0.7848179340362549},{"id":"https://openalex.org/C206729178","wikidata":"https://www.wikidata.org/wiki/Q2271896","display_name":"Scheduling (production processes)","level":2,"score":0.6599645614624023},{"id":"https://openalex.org/C177774035","wikidata":"https://www.wikidata.org/wiki/Q1246948","display_name":"Granularity","level":2,"score":0.6333301067352295},{"id":"https://openalex.org/C45374587","wikidata":"https://www.wikidata.org/wiki/Q12525525","display_name":"Computation","level":2,"score":0.6221250891685486},{"id":"https://openalex.org/C158379750","wikidata":"https://www.wikidata.org/wiki/Q214111","display_name":"Network packet","level":2,"score":0.5186033248901367},{"id":"https://openalex.org/C173608175","wikidata":"https://www.wikidata.org/wiki/Q232661","display_name":"Parallel computing","level":1,"score":0.5115073919296265},{"id":"https://openalex.org/C2777777548","wikidata":"https://www.wikidata.org/wiki/Q206040","display_name":"Mercury (programming language)","level":2,"score":0.4986741542816162},{"id":"https://openalex.org/C7345512","wikidata":"https://www.wikidata.org/wiki/Q209372","display_name":"Transport layer","level":3,"score":0.45992952585220337},{"id":"https://openalex.org/C120314980","wikidata":"https://www.wikidata.org/wiki/Q180634","display_name":"Distributed computing","level":1,"score":0.4583742618560791},{"id":"https://openalex.org/C42812","wikidata":"https://www.wikidata.org/wiki/Q1082910","display_name":"Partition (number theory)","level":2,"score":0.4268642067909241},{"id":"https://openalex.org/C31258907","wikidata":"https://www.wikidata.org/wiki/Q1301371","display_name":"Computer network","level":1,"score":0.4143141806125641},{"id":"https://openalex.org/C11413529","wikidata":"https://www.wikidata.org/wiki/Q8366","display_name":"Algorithm","level":1,"score":0.2234863042831421},{"id":"https://openalex.org/C2779227376","wikidata":"https://www.wikidata.org/wiki/Q6505497","display_name":"Layer (electronics)","level":2,"score":0.17296147346496582},{"id":"https://openalex.org/C126255220","wikidata":"https://www.wikidata.org/wiki/Q141495","display_name":"Mathematical optimization","level":1,"score":0.12139329314231873},{"id":"https://openalex.org/C185592680","wikidata":"https://www.wikidata.org/wiki/Q2329","display_name":"Chemistry","level":0,"score":0.09412083029747009},{"id":"https://openalex.org/C111919701","wikidata":"https://www.wikidata.org/wiki/Q9135","display_name":"Operating system","level":1,"score":0.08588188886642456},{"id":"https://openalex.org/C178790620","wikidata":"https://www.wikidata.org/wiki/Q11351","display_name":"Organic chemistry","level":1,"score":0.0},{"id":"https://openalex.org/C114614502","wikidata":"https://www.wikidata.org/wiki/Q76592","display_name":"Combinatorics","level":1,"score":0.0},{"id":"https://openalex.org/C33923547","wikidata":"https://www.wikidata.org/wiki/Q395","display_name":"Mathematics","level":0,"score":0.0},{"id":"https://openalex.org/C199360897","wikidata":"https://www.wikidata.org/wiki/Q9143","display_name":"Programming language","level":1,"score":0.0}],"mesh":[],"locations_count":1,"locations":[{"id":"doi:10.1109/infocom48880.2022.9796820","is_oa":false,"landing_page_url":"https://doi.org/10.1109/infocom48880.2022.9796820","pdf_url":null,"source":null,"license":null,"license_id":null,"version":"publishedVersion","is_accepted":true,"is_published":true,"raw_source_name":"IEEE INFOCOM 2022 - IEEE Conference on Computer Communications","raw_type":"proceedings-article"}],"best_oa_location":null,"sustainable_development_goals":[],"awards":[],"funders":[{"id":"https://openalex.org/F4320337504","display_name":"Research and Development","ror":"https://ror.org/027s68j25"}],"has_content":{"grobid_xml":false,"pdf":false},"content_urls":null,"referenced_works_count":57,"referenced_works":["https://openalex.org/W1686810756","https://openalex.org/W2060393849","https://openalex.org/W2132737349","https://openalex.org/W2138243089","https://openalex.org/W2163605009","https://openalex.org/W2186615578","https://openalex.org/W2193413348","https://openalex.org/W2194775991","https://openalex.org/W2617766261","https://openalex.org/W2751527518","https://openalex.org/W2769644379","https://openalex.org/W2769986458","https://openalex.org/W2896457183","https://openalex.org/W2920397365","https://openalex.org/W2953384591","https://openalex.org/W2962758826","https://openalex.org/W2963748143","https://openalex.org/W2963786636","https://openalex.org/W2975712713","https://openalex.org/W2982475424","https://openalex.org/W3047002910","https://openalex.org/W3047357290","https://openalex.org/W3106914464","https://openalex.org/W3107997909","https://openalex.org/W3124352525","https://openalex.org/W3134069806","https://openalex.org/W3153854431","https://openalex.org/W3154219392","https://openalex.org/W3154560182","https://openalex.org/W3177554119","https://openalex.org/W3193250980","https://openalex.org/W4288357791","https://openalex.org/W4295312788","https://openalex.org/W4297775537","https://openalex.org/W4301239768","https://openalex.org/W6637373629","https://openalex.org/W6665801690","https://openalex.org/W6679393576","https://openalex.org/W6684191040","https://openalex.org/W6686509673","https://openalex.org/W6687566353","https://openalex.org/W6713134421","https://openalex.org/W6737664043","https://openalex.org/W6738460352","https://openalex.org/W6739693220","https://openalex.org/W6743499376","https://openalex.org/W6746081298","https://openalex.org/W6746200960","https://openalex.org/W6748645090","https://openalex.org/W6754818105","https://openalex.org/W6755207826","https://openalex.org/W6760703128","https://openalex.org/W6762211661","https://openalex.org/W6764309476","https://openalex.org/W6766978945","https://openalex.org/W6786422851","https://openalex.org/W6794216301"],"related_works":["https://openalex.org/W2058965144","https://openalex.org/W2164382479","https://openalex.org/W2146343568","https://openalex.org/W98480971","https://openalex.org/W2150291671","https://openalex.org/W2013643406","https://openalex.org/W2027972911","https://openalex.org/W2157978810","https://openalex.org/W2383090288","https://openalex.org/W2387846155"],"abstract_inverted_index":{"Communication":[0],"scheduling":[1,45,62,101],"is":[2,29,69],"crucial":[3],"to":[4,41,102,141],"improve":[5],"the":[6,19,43,65,72,77,95,99,103,107,113,117,126,138,143,192],"efficiency":[7],"of":[8,22],"training":[9],"large":[10],"deep":[11,24],"learning":[12],"models":[13,162],"with":[14,46,112,167],"data":[15],"parallelism,":[16],"in":[17,116,153,163],"which":[18],"transmission":[20],"order":[21],"layer-wise":[23],"neural":[25],"network":[26],"(DNN)":[27],"tensors":[28],"determined":[30],"for":[31],"a":[32,50,86],"better":[33],"computation-communication":[34],"overlap.":[35],"Prior":[36],"approaches":[37],"adopt":[38],"tensor":[39,57,194],"partitioning":[40,78,195],"enhance":[42],"priority":[44,100,115],"finer":[47],"granularity.":[48,109],"However,":[49],"startup":[51],"time":[52],"slot":[53],"inserted":[54],"before":[55],"each":[56],"partition":[58,67,94],"will":[59,120],"neutralize":[60],"this":[61,81],"gain.":[63],"Tuning":[64],"optimal":[66],"size":[68],"difficult":[70],"and":[71,131,147,155,186],"application-layer":[73],"solutions":[74],"cannot":[75],"eliminate":[76],"overhead.":[79],"In":[80],"paper,":[82],"we":[83],"propose":[84],"Mercury,":[85],"simple":[87],"transport":[88,104,139],"layer":[89,105,140],"scheduler":[90],"that":[91,173],"does":[92],"not":[93],"tensors,":[96],"but":[97],"moves":[98],"at":[106,137],"packet":[108],"The":[110],"packets":[111],"highest":[114],"Mercury":[118,124,152,174],"buffer":[119],"be":[121],"transmitted":[122],"first.":[123],"achieves":[125],"near-optimal":[127],"overlapping":[128],"between":[129],"communication":[130],"computation.":[132],"It":[133],"leverages":[134],"immediate":[135],"aggregation":[136],"enable":[142],"coincident":[144],"gradient":[145],"push":[146],"parameter":[148],"pull.":[149],"We":[150],"implement":[151],"MXNet":[154],"conduct":[156],"comprehensive":[157],"experiments":[158],"on":[159],"five":[160],"DNN":[161],"an":[164],"8-node":[165],"cluster":[166],"10Gbps":[168],"Ethernet.":[169],"Experimental":[170],"results":[171],"show":[172],"can":[175],"achieve":[176],"about":[177],"1.18":[178],"~":[179,188],"2.18":[180],"\u00d7":[181],"speedup":[182,190],"over":[183,191],"vanilla":[184],"MXNet,":[185],"1.08":[187],"2.04\u00d7":[189],"state-of-the-art":[193],"solution.":[196]},"counts_by_year":[{"year":2026,"cited_by_count":1},{"year":2025,"cited_by_count":2},{"year":2024,"cited_by_count":3},{"year":2023,"cited_by_count":5}],"updated_date":"2026-07-29T14:22:42.915294","created_date":"2025-10-10T00:00:00"}