{"id":"https://openalex.org/W2926767350","doi":"https://doi.org/10.1109/tpds.2019.2928289","title":"Evaluating Modern GPU Interconnect: PCIe, NVLink, NV-SLI, NVSwitch and GPUDirect","display_name":"Evaluating Modern GPU Interconnect: PCIe, NVLink, NV-SLI, NVSwitch and GPUDirect","publication_year":2019,"publication_date":"2019-07-15","ids":{"openalex":"https://openalex.org/W2926767350","doi":"https://doi.org/10.1109/tpds.2019.2928289","mag":"2926767350"},"language":"en","primary_location":{"id":"doi:10.1109/tpds.2019.2928289","is_oa":false,"landing_page_url":"https://doi.org/10.1109/tpds.2019.2928289","pdf_url":null,"source":{"id":"https://openalex.org/S97130795","display_name":"IEEE Transactions on Parallel and Distributed Systems","issn_l":"1045-9219","issn":["1045-9219","1558-2183","2161-9883"],"is_oa":false,"is_in_doaj":false,"is_core":true,"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":"journal"},"license":null,"license_id":null,"version":"publishedVersion","is_accepted":true,"is_published":true,"raw_source_name":"IEEE Transactions on Parallel and Distributed Systems","raw_type":"journal-article"},"type":"article","indexed_in":["arxiv","crossref"],"open_access":{"is_oa":true,"oa_status":"green","oa_url":"https://arxiv.org/pdf/1903.04611","any_repository_has_fulltext":true},"authorships":[{"author_position":"first","author":{"id":"https://openalex.org/A5100413657","display_name":"Ang Li","orcid":"https://orcid.org/0000-0003-3734-9137"},"institutions":[{"id":"https://openalex.org/I142606810","display_name":"Pacific Northwest National Laboratory","ror":"https://ror.org/05h992307","country_code":"US","type":"facility","lineage":["https://openalex.org/I1325736334","https://openalex.org/I1330989302","https://openalex.org/I142606810","https://openalex.org/I39565521"]}],"countries":["US"],"is_corresponding":true,"raw_author_name":"Ang Li","raw_affiliation_strings":["High-Performance Computing Group, Pacific Northwest National Laboratory (PNNL), Richland, USA"],"affiliations":[{"raw_affiliation_string":"High-Performance Computing Group, Pacific Northwest National Laboratory (PNNL), Richland, USA","institution_ids":["https://openalex.org/I142606810"]}]},{"author_position":"middle","author":{"id":"https://openalex.org/A5043209884","display_name":"Shuaiwen Leon Song","orcid":"https://orcid.org/0000-0002-8402-1436"},"institutions":[{"id":"https://openalex.org/I142606810","display_name":"Pacific Northwest National Laboratory","ror":"https://ror.org/05h992307","country_code":"US","type":"facility","lineage":["https://openalex.org/I1325736334","https://openalex.org/I1330989302","https://openalex.org/I142606810","https://openalex.org/I39565521"]}],"countries":["US"],"is_corresponding":false,"raw_author_name":"Shuaiwen Leon Song","raw_affiliation_strings":["High-Performance Computing Group, Pacific Northwest National Laboratory (PNNL), Richland, USA"],"affiliations":[{"raw_affiliation_string":"High-Performance Computing Group, Pacific Northwest National Laboratory (PNNL), Richland, USA","institution_ids":["https://openalex.org/I142606810"]}]},{"author_position":"middle","author":{"id":"https://openalex.org/A5063649910","display_name":"Jieyang Chen","orcid":"https://orcid.org/0000-0002-1905-9171"},"institutions":[{"id":"https://openalex.org/I1289243028","display_name":"Oak Ridge National Laboratory","ror":"https://ror.org/01qz5mb56","country_code":"US","type":"facility","lineage":["https://openalex.org/I1289243028","https://openalex.org/I1330989302","https://openalex.org/I39565521","https://openalex.org/I4210159294"]}],"countries":["US"],"is_corresponding":false,"raw_author_name":"Jieyang Chen","raw_affiliation_strings":["Computer Science and Mathematics Department, Oak Ridge National Laboratory, Oak Ridge, USA"],"affiliations":[{"raw_affiliation_string":"Computer Science and Mathematics Department, Oak Ridge National Laboratory, Oak Ridge, USA","institution_ids":["https://openalex.org/I1289243028"]}]},{"author_position":"middle","author":{"id":"https://openalex.org/A5100454305","display_name":"Jiajia Li","orcid":"https://orcid.org/0000-0003-1270-4147"},"institutions":[{"id":"https://openalex.org/I142606810","display_name":"Pacific Northwest National Laboratory","ror":"https://ror.org/05h992307","country_code":"US","type":"facility","lineage":["https://openalex.org/I1325736334","https://openalex.org/I1330989302","https://openalex.org/I142606810","https://openalex.org/I39565521"]}],"countries":["US"],"is_corresponding":false,"raw_author_name":"Jiajia Li","raw_affiliation_strings":["High-Performance Computing Group, Pacific Northwest National Laboratory (PNNL), Richland, USA"],"affiliations":[{"raw_affiliation_string":"High-Performance Computing Group, Pacific Northwest National Laboratory (PNNL), Richland, USA","institution_ids":["https://openalex.org/I142606810"]}]},{"author_position":"middle","author":{"id":"https://openalex.org/A5100331713","display_name":"Xu Liu","orcid":"https://orcid.org/0000-0002-8984-5485"},"institutions":[{"id":"https://openalex.org/I16285277","display_name":"William & Mary","ror":"https://ror.org/03hsf0573","country_code":"US","type":"education","lineage":["https://openalex.org/I16285277"]},{"id":"https://openalex.org/I267592682","display_name":"Williams (United States)","ror":"https://ror.org/007zhvp17","country_code":"US","type":"company","lineage":["https://openalex.org/I267592682"]}],"countries":["US"],"is_corresponding":false,"raw_author_name":"Xu Liu","raw_affiliation_strings":["Computer Science Department, College of William and Mary, Williamsburg, USA"],"affiliations":[{"raw_affiliation_string":"Computer Science Department, College of William and Mary, Williamsburg, USA","institution_ids":["https://openalex.org/I16285277","https://openalex.org/I267592682"]}]},{"author_position":"middle","author":{"id":"https://openalex.org/A5047746050","display_name":"Nathan R. Tallent","orcid":"https://orcid.org/0000-0003-4297-3057"},"institutions":[{"id":"https://openalex.org/I142606810","display_name":"Pacific Northwest National Laboratory","ror":"https://ror.org/05h992307","country_code":"US","type":"facility","lineage":["https://openalex.org/I1325736334","https://openalex.org/I1330989302","https://openalex.org/I142606810","https://openalex.org/I39565521"]}],"countries":["US"],"is_corresponding":false,"raw_author_name":"Nathan R. Tallent","raw_affiliation_strings":["High-Performance Computing Group, Pacific Northwest National Laboratory (PNNL), Richland, USA"],"affiliations":[{"raw_affiliation_string":"High-Performance Computing Group, Pacific Northwest National Laboratory (PNNL), Richland, USA","institution_ids":["https://openalex.org/I142606810"]}]},{"author_position":"last","author":{"id":"https://openalex.org/A5038062469","display_name":"Kevin Barker","orcid":"https://orcid.org/0000-0003-4947-0559"},"institutions":[{"id":"https://openalex.org/I142606810","display_name":"Pacific Northwest National Laboratory","ror":"https://ror.org/05h992307","country_code":"US","type":"facility","lineage":["https://openalex.org/I1325736334","https://openalex.org/I1330989302","https://openalex.org/I142606810","https://openalex.org/I39565521"]}],"countries":["US"],"is_corresponding":false,"raw_author_name":"Kevin J. Barker","raw_affiliation_strings":["High-Performance Computing Group, Pacific Northwest National Laboratory (PNNL), Richland, USA"],"affiliations":[{"raw_affiliation_string":"High-Performance Computing Group, Pacific Northwest National Laboratory (PNNL), Richland, USA","institution_ids":["https://openalex.org/I142606810"]}]}],"institutions":[],"countries_distinct_count":1,"institutions_distinct_count":7,"corresponding_author_ids":["https://openalex.org/A5100413657"],"corresponding_institution_ids":["https://openalex.org/I142606810"],"apc_list":null,"apc_paid":null,"fwci":28.3898,"has_fulltext":false,"cited_by_count":263,"citation_normalized_percentile":{"value":0.99911404,"is_in_top_1_percent":true,"is_in_top_10_percent":true},"cited_by_percentile_year":{"min":99,"max":100},"biblio":{"volume":"31","issue":"1","first_page":"94","last_page":"110"},"is_retracted":false,"is_paratext":false,"is_xpac":false,"primary_topic":{"id":"https://openalex.org/T10054","display_name":"Parallel Computing and Optimization Techniques","score":0.9998999834060669,"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"}},"topics":[{"id":"https://openalex.org/T10054","display_name":"Parallel Computing and Optimization Techniques","score":0.9998999834060669,"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/T11181","display_name":"Advanced Data Storage Technologies","score":0.9997000098228455,"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/T10829","display_name":"Interconnection Networks and Systems","score":0.9994999766349792,"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/pci-express","display_name":"PCI Express","score":0.88974928855896},{"id":"https://openalex.org/keywords/computer-science","display_name":"Computer science","score":0.8782692551612854},{"id":"https://openalex.org/keywords/parallel-computing","display_name":"Parallel computing","score":0.5018222332000732},{"id":"https://openalex.org/keywords/supercomputer","display_name":"Supercomputer","score":0.4824012517929077},{"id":"https://openalex.org/keywords/server","display_name":"Server","score":0.48165643215179443},{"id":"https://openalex.org/keywords/general-purpose-computing-on-graphics-processing-units","display_name":"General-purpose computing on graphics processing units","score":0.439981609582901},{"id":"https://openalex.org/keywords/computer-architecture","display_name":"Computer architecture","score":0.3887993097305298},{"id":"https://openalex.org/keywords/field-programmable-gate-array","display_name":"Field-programmable gate array","score":0.38304346799850464},{"id":"https://openalex.org/keywords/embedded-system","display_name":"Embedded system","score":0.32583409547805786},{"id":"https://openalex.org/keywords/operating-system","display_name":"Operating system","score":0.315640389919281},{"id":"https://openalex.org/keywords/graphics","display_name":"Graphics","score":0.10595890879631042}],"concepts":[{"id":"https://openalex.org/C64270927","wikidata":"https://www.wikidata.org/wiki/Q206924","display_name":"PCI Express","level":3,"score":0.88974928855896},{"id":"https://openalex.org/C41008148","wikidata":"https://www.wikidata.org/wiki/Q21198","display_name":"Computer science","level":0,"score":0.8782692551612854},{"id":"https://openalex.org/C173608175","wikidata":"https://www.wikidata.org/wiki/Q232661","display_name":"Parallel computing","level":1,"score":0.5018222332000732},{"id":"https://openalex.org/C83283714","wikidata":"https://www.wikidata.org/wiki/Q121117","display_name":"Supercomputer","level":2,"score":0.4824012517929077},{"id":"https://openalex.org/C93996380","wikidata":"https://www.wikidata.org/wiki/Q44127","display_name":"Server","level":2,"score":0.48165643215179443},{"id":"https://openalex.org/C50630238","wikidata":"https://www.wikidata.org/wiki/Q971505","display_name":"General-purpose computing on graphics processing units","level":3,"score":0.439981609582901},{"id":"https://openalex.org/C118524514","wikidata":"https://www.wikidata.org/wiki/Q173212","display_name":"Computer architecture","level":1,"score":0.3887993097305298},{"id":"https://openalex.org/C42935608","wikidata":"https://www.wikidata.org/wiki/Q190411","display_name":"Field-programmable gate array","level":2,"score":0.38304346799850464},{"id":"https://openalex.org/C149635348","wikidata":"https://www.wikidata.org/wiki/Q193040","display_name":"Embedded system","level":1,"score":0.32583409547805786},{"id":"https://openalex.org/C111919701","wikidata":"https://www.wikidata.org/wiki/Q9135","display_name":"Operating system","level":1,"score":0.315640389919281},{"id":"https://openalex.org/C21442007","wikidata":"https://www.wikidata.org/wiki/Q1027879","display_name":"Graphics","level":2,"score":0.10595890879631042}],"mesh":[],"locations_count":2,"locations":[{"id":"doi:10.1109/tpds.2019.2928289","is_oa":false,"landing_page_url":"https://doi.org/10.1109/tpds.2019.2928289","pdf_url":null,"source":{"id":"https://openalex.org/S97130795","display_name":"IEEE Transactions on Parallel and Distributed Systems","issn_l":"1045-9219","issn":["1045-9219","1558-2183","2161-9883"],"is_oa":false,"is_in_doaj":false,"is_core":true,"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":"journal"},"license":null,"license_id":null,"version":"publishedVersion","is_accepted":true,"is_published":true,"raw_source_name":"IEEE Transactions on Parallel and Distributed Systems","raw_type":"journal-article"},{"id":"pmh:oai:arXiv.org:1903.04611","is_oa":true,"landing_page_url":"http://arxiv.org/abs/1903.04611","pdf_url":"https://arxiv.org/pdf/1903.04611","source":{"id":"https://openalex.org/S4306400194","display_name":"arXiv (Cornell University)","issn_l":null,"issn":null,"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":null,"license_id":null,"version":"submittedVersion","is_accepted":false,"is_published":false,"raw_source_name":null,"raw_type":"text"}],"best_oa_location":{"id":"pmh:oai:arXiv.org:1903.04611","is_oa":true,"landing_page_url":"http://arxiv.org/abs/1903.04611","pdf_url":"https://arxiv.org/pdf/1903.04611","source":{"id":"https://openalex.org/S4306400194","display_name":"arXiv (Cornell University)","issn_l":null,"issn":null,"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":null,"license_id":null,"version":"submittedVersion","is_accepted":false,"is_published":false,"raw_source_name":null,"raw_type":"text"},"sustainable_development_goals":[],"awards":[{"id":"https://openalex.org/G5200347334","display_name":null,"funder_award_id":"DE-AC05-76RL01830","funder_id":"https://openalex.org/F4320306084","funder_display_name":"U.S. Department of Energy"},{"id":"https://openalex.org/G8679700530","display_name":null,"funder_award_id":"66150","funder_id":"https://openalex.org/F4320306084","funder_display_name":"U.S. Department of Energy"}],"funders":[{"id":"https://openalex.org/F4320306084","display_name":"U.S. Department of Energy","ror":"https://ror.org/01bj3aw27"},{"id":"https://openalex.org/F4320332369","display_name":"National Nuclear Security Administration","ror":"https://ror.org/03sk1we31"},{"id":"https://openalex.org/F4320338354","display_name":"Pacific Northwest National Laboratory","ror":"https://ror.org/05h992307"}],"has_content":{"grobid_xml":false,"pdf":false},"content_urls":null,"referenced_works_count":48,"referenced_works":["https://openalex.org/W1967701350","https://openalex.org/W1980759071","https://openalex.org/W1988307172","https://openalex.org/W2007511350","https://openalex.org/W2025388037","https://openalex.org/W2041356909","https://openalex.org/W2056639008","https://openalex.org/W2067280503","https://openalex.org/W2070940137","https://openalex.org/W2091599698","https://openalex.org/W2098019196","https://openalex.org/W2109341366","https://openalex.org/W2345717615","https://openalex.org/W2410071687","https://openalex.org/W2469975815","https://openalex.org/W2483772785","https://openalex.org/W2557367833","https://openalex.org/W2560878429","https://openalex.org/W2605251767","https://openalex.org/W2612387305","https://openalex.org/W2622263826","https://openalex.org/W2733525365","https://openalex.org/W2734556049","https://openalex.org/W2740001873","https://openalex.org/W2753515784","https://openalex.org/W2763018387","https://openalex.org/W2767068499","https://openalex.org/W2767210925","https://openalex.org/W2768065515","https://openalex.org/W2777078856","https://openalex.org/W2778540992","https://openalex.org/W2783292547","https://openalex.org/W2785822536","https://openalex.org/W2795994638","https://openalex.org/W2891890103","https://openalex.org/W2900824371","https://openalex.org/W2903901007","https://openalex.org/W2920668770","https://openalex.org/W2953200226","https://openalex.org/W3103894541","https://openalex.org/W4229654654","https://openalex.org/W4237978577","https://openalex.org/W4241741271","https://openalex.org/W4247353671","https://openalex.org/W6739622702","https://openalex.org/W6747195836","https://openalex.org/W6748629087","https://openalex.org/W6756488470"],"related_works":["https://openalex.org/W4385894176","https://openalex.org/W2347371119","https://openalex.org/W2612768808","https://openalex.org/W3131402800","https://openalex.org/W2388965158","https://openalex.org/W2366201982","https://openalex.org/W2808480228","https://openalex.org/W2593071546","https://openalex.org/W2805641962","https://openalex.org/W1564576805"],"abstract_inverted_index":{"High":[0],"performance":[1,52,186,212],"multi-GPU":[2,50,155,185],"computing":[3],"becomes":[4],"an":[5,101,150],"inevitable":[6],"trend":[7],"due":[8],"to":[9],"the":[10,29,42,61,112,158,173],"ever-increasing":[11],"demand":[12],"on":[13,34,49,68,111,166],"computation":[14],"capability":[15],"in":[16,153,182,198],"emerging":[17],"domains":[18],"such":[19],"as":[20,98,100,170,172,208,210],"deep":[21,32],"learning,":[22],"big":[23],"data":[24],"and":[25,41,80,86,95,134,196,205],"planet-scale":[26],"simulations.":[27],"However,":[28],"lack":[30],"of":[31,45,72,121],"understanding":[33],"how":[35],"modern":[36,73],"GPUs":[37],"can":[38,162,179],"be":[39,180],"connected":[40],"real":[43],"impact":[44,165],"state-of-the-art":[46],"interconnect":[47],"technology":[48],"application":[51,151],"become":[53],"a":[54,65,154,199],"hurdle.":[55],"In":[56],"this":[57],"paper,":[58],"we":[59,115],"fill":[60],"gap":[62],"by":[63,130,140],"conducting":[64],"thorough":[66],"evaluation":[67,178],"five":[69],"latest":[70],"types":[71,120],"GPU":[74,122,160,167,192],"interconnects:":[75],"PCIe,":[76],"NVLink-V1,":[77],"NVLink-V2,":[78],"NVLink-SLI":[79],"NVSwitch,":[81],"from":[82],"six":[83],"high-end":[84],"servers":[85],"HPC":[87,206],"platforms:":[88],"NVIDIA":[89,106],"P100-DGX-1,":[90],"V100-DGX-1,":[91],"DGX-2,":[92],"OLCF's":[93],"SummitDev":[94],"Summit":[96],"supercomputers,":[97],"well":[99,171,209],"SLI-linked":[102],"system":[103],"with":[104],"two":[105],"Turing":[107],"RTX-2080":[108],"GPUs.":[109],"Based":[110],"empirical":[113],"evaluation,":[114],"have":[116],"observed":[117],"four":[118],"new":[119],"communication":[123,168],"network":[124],"NUMA":[125],"effects:":[126],"three":[127],"are":[128,189],"triggered":[129],"NVLink's":[131],"topology,":[132],"connectivity":[133],"routing,":[135],"while":[136],"one":[137],"is":[138],"caused":[139],"PCIe":[141],"chipset":[142],"design":[143],"issue.":[144],"These":[145],"observations":[146],"indicate":[147],"that,":[148],"for":[149,191],"running":[152],"node,":[156],"choosing":[157],"right":[159],"combination":[161],"impose":[163],"considerable":[164],"efficiency,":[169],"application's":[174],"overall":[175],"performance.":[176],"Our":[177],"leveraged":[181],"building":[183],"practical":[184],"models,":[187],"which":[188],"vital":[190],"task":[193],"allocation,":[194],"scheduling":[195],"migration":[197],"shared":[200],"environment":[201],"(e.g.,":[202],"AI":[203],"cloud":[204],"centers),":[207],"communication-oriented":[211],"tuning.":[213]},"counts_by_year":[{"year":2026,"cited_by_count":11},{"year":2025,"cited_by_count":63},{"year":2024,"cited_by_count":42},{"year":2023,"cited_by_count":32},{"year":2022,"cited_by_count":30},{"year":2021,"cited_by_count":48},{"year":2020,"cited_by_count":28},{"year":2019,"cited_by_count":9}],"updated_date":"2026-04-13T07:58:08.660418","created_date":"2019-04-11T00:00:00"}
