{"id":"https://openalex.org/W2903901007","doi":"https://doi.org/10.1109/iiswc.2018.8573483","title":"Tartan: Evaluating Modern GPU Interconnect via a Multi-GPU Benchmark Suite","display_name":"Tartan: Evaluating Modern GPU Interconnect via a Multi-GPU Benchmark Suite","publication_year":2018,"publication_date":"2018-09-01","ids":{"openalex":"https://openalex.org/W2903901007","doi":"https://doi.org/10.1109/iiswc.2018.8573483","mag":"2903901007"},"language":"en","primary_location":{"id":"doi:10.1109/iiswc.2018.8573483","is_oa":false,"landing_page_url":"https://doi.org/10.1109/iiswc.2018.8573483","pdf_url":null,"source":null,"license":null,"license_id":null,"version":"publishedVersion","is_accepted":true,"is_published":true,"raw_source_name":"2018 IEEE International Symposium on Workload Characterization (IISWC)","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/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":false,"raw_author_name":"Ang Li","raw_affiliation_strings":["Pacific Northwest National Laboratory"],"raw_orcid":null,"affiliations":[{"raw_affiliation_string":"Pacific Northwest National Laboratory","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":["Pacific Northwest National Laboratory"],"raw_orcid":null,"affiliations":[{"raw_affiliation_string":"Pacific Northwest National Laboratory","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/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":"Jieyang Chen","raw_affiliation_strings":["Pacific Northwest National Laboratory"],"raw_orcid":null,"affiliations":[{"raw_affiliation_string":"Pacific Northwest National Laboratory","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/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":"Xu Liu","raw_affiliation_strings":["Pacific Northwest National Laboratory"],"raw_orcid":null,"affiliations":[{"raw_affiliation_string":"Pacific Northwest National Laboratory","institution_ids":["https://openalex.org/I142606810"]}]},{"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 Tallent","raw_affiliation_strings":["Pacific Northwest National Laboratory"],"raw_orcid":null,"affiliations":[{"raw_affiliation_string":"Pacific Northwest National Laboratory","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 Barker","raw_affiliation_strings":["Pacific Northwest National Laboratory"],"raw_orcid":null,"affiliations":[{"raw_affiliation_string":"Pacific Northwest National Laboratory","institution_ids":["https://openalex.org/I142606810"]}]}],"institutions":[],"countries_distinct_count":1,"institutions_distinct_count":1,"corresponding_author_ids":[],"corresponding_institution_ids":["https://openalex.org/I142606810"],"apc_list":null,"apc_paid":null,"fwci":null,"has_fulltext":false,"cited_by_count":68,"citation_normalized_percentile":null,"cited_by_percentile_year":null,"biblio":{"volume":null,"issue":null,"first_page":"191","last_page":"202"},"is_retracted":false,"is_paratext":false,"is_xpac":false,"primary_topic":{"id":"https://openalex.org/T10054","display_name":"Parallel Computing and Optimization Techniques","score":1.0,"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":1.0,"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":0.9994000196456909,"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/T10502","display_name":"Advanced Memory and Neural Computing","score":0.9987999796867371,"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/remote-direct-memory-access","display_name":"Remote direct memory access","score":0.9289866089820862},{"id":"https://openalex.org/keywords/computer-science","display_name":"Computer science","score":0.862430214881897},{"id":"https://openalex.org/keywords/infiniband","display_name":"InfiniBand","score":0.7568522691726685},{"id":"https://openalex.org/keywords/pci-express","display_name":"PCI Express","score":0.7423945665359497},{"id":"https://openalex.org/keywords/suite","display_name":"Suite","score":0.7329561710357666},{"id":"https://openalex.org/keywords/benchmark","display_name":"Benchmark (surveying)","score":0.7224792242050171},{"id":"https://openalex.org/keywords/parallel-computing","display_name":"Parallel computing","score":0.608102560043335},{"id":"https://openalex.org/keywords/node","display_name":"Node (physics)","score":0.4975462257862091},{"id":"https://openalex.org/keywords/programming-paradigm","display_name":"Programming paradigm","score":0.41908949613571167},{"id":"https://openalex.org/keywords/supercomputer","display_name":"Supercomputer","score":0.4130033552646637},{"id":"https://openalex.org/keywords/computer-architecture","display_name":"Computer architecture","score":0.36194294691085815},{"id":"https://openalex.org/keywords/computational-science","display_name":"Computational science","score":0.32577061653137207},{"id":"https://openalex.org/keywords/operating-system","display_name":"Operating system","score":0.23771917819976807},{"id":"https://openalex.org/keywords/field-programmable-gate-array","display_name":"Field-programmable gate array","score":0.20000076293945312}],"concepts":[{"id":"https://openalex.org/C130795937","wikidata":"https://www.wikidata.org/wiki/Q2561570","display_name":"Remote direct memory access","level":2,"score":0.9289866089820862},{"id":"https://openalex.org/C41008148","wikidata":"https://www.wikidata.org/wiki/Q21198","display_name":"Computer science","level":0,"score":0.862430214881897},{"id":"https://openalex.org/C2781030343","wikidata":"https://www.wikidata.org/wiki/Q922437","display_name":"InfiniBand","level":2,"score":0.7568522691726685},{"id":"https://openalex.org/C64270927","wikidata":"https://www.wikidata.org/wiki/Q206924","display_name":"PCI Express","level":3,"score":0.7423945665359497},{"id":"https://openalex.org/C79581498","wikidata":"https://www.wikidata.org/wiki/Q1367530","display_name":"Suite","level":2,"score":0.7329561710357666},{"id":"https://openalex.org/C185798385","wikidata":"https://www.wikidata.org/wiki/Q1161707","display_name":"Benchmark (surveying)","level":2,"score":0.7224792242050171},{"id":"https://openalex.org/C173608175","wikidata":"https://www.wikidata.org/wiki/Q232661","display_name":"Parallel computing","level":1,"score":0.608102560043335},{"id":"https://openalex.org/C62611344","wikidata":"https://www.wikidata.org/wiki/Q1062658","display_name":"Node (physics)","level":2,"score":0.4975462257862091},{"id":"https://openalex.org/C34165917","wikidata":"https://www.wikidata.org/wiki/Q188267","display_name":"Programming paradigm","level":2,"score":0.41908949613571167},{"id":"https://openalex.org/C83283714","wikidata":"https://www.wikidata.org/wiki/Q121117","display_name":"Supercomputer","level":2,"score":0.4130033552646637},{"id":"https://openalex.org/C118524514","wikidata":"https://www.wikidata.org/wiki/Q173212","display_name":"Computer architecture","level":1,"score":0.36194294691085815},{"id":"https://openalex.org/C459310","wikidata":"https://www.wikidata.org/wiki/Q117801","display_name":"Computational science","level":1,"score":0.32577061653137207},{"id":"https://openalex.org/C111919701","wikidata":"https://www.wikidata.org/wiki/Q9135","display_name":"Operating system","level":1,"score":0.23771917819976807},{"id":"https://openalex.org/C42935608","wikidata":"https://www.wikidata.org/wiki/Q190411","display_name":"Field-programmable gate array","level":2,"score":0.20000076293945312},{"id":"https://openalex.org/C13280743","wikidata":"https://www.wikidata.org/wiki/Q131089","display_name":"Geodesy","level":1,"score":0.0},{"id":"https://openalex.org/C199360897","wikidata":"https://www.wikidata.org/wiki/Q9143","display_name":"Programming language","level":1,"score":0.0},{"id":"https://openalex.org/C95457728","wikidata":"https://www.wikidata.org/wiki/Q309","display_name":"History","level":0,"score":0.0},{"id":"https://openalex.org/C127413603","wikidata":"https://www.wikidata.org/wiki/Q11023","display_name":"Engineering","level":0,"score":0.0},{"id":"https://openalex.org/C166957645","wikidata":"https://www.wikidata.org/wiki/Q23498","display_name":"Archaeology","level":1,"score":0.0},{"id":"https://openalex.org/C205649164","wikidata":"https://www.wikidata.org/wiki/Q1071","display_name":"Geography","level":0,"score":0.0},{"id":"https://openalex.org/C66938386","wikidata":"https://www.wikidata.org/wiki/Q633538","display_name":"Structural engineering","level":1,"score":0.0}],"mesh":[],"locations_count":1,"locations":[{"id":"doi:10.1109/iiswc.2018.8573483","is_oa":false,"landing_page_url":"https://doi.org/10.1109/iiswc.2018.8573483","pdf_url":null,"source":null,"license":null,"license_id":null,"version":"publishedVersion","is_accepted":true,"is_published":true,"raw_source_name":"2018 IEEE International Symposium on Workload Characterization (IISWC)","raw_type":"proceedings-article"}],"best_oa_location":null,"sustainable_development_goals":[],"awards":[],"funders":[{"id":"https://openalex.org/F4320337144","display_name":"National Supercomputing Center, Korea Institute of Science and Technology Information","ror":null}],"has_content":{"grobid_xml":false,"pdf":false},"content_urls":null,"referenced_works_count":53,"referenced_works":["https://openalex.org/W348433680","https://openalex.org/W1847196819","https://openalex.org/W1967701350","https://openalex.org/W1979527452","https://openalex.org/W1980759071","https://openalex.org/W1988307172","https://openalex.org/W1997162567","https://openalex.org/W2007511350","https://openalex.org/W2025388037","https://openalex.org/W2041356909","https://openalex.org/W2067280503","https://openalex.org/W2070940137","https://openalex.org/W2080592089","https://openalex.org/W2091599698","https://openalex.org/W2098019196","https://openalex.org/W2109341366","https://openalex.org/W2128120785","https://openalex.org/W2149234156","https://openalex.org/W2265680019","https://openalex.org/W2273440736","https://openalex.org/W2345717615","https://openalex.org/W2410071687","https://openalex.org/W2469975815","https://openalex.org/W2483772785","https://openalex.org/W2557367833","https://openalex.org/W2559678521","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/W2767210925","https://openalex.org/W2768065515","https://openalex.org/W2777078856","https://openalex.org/W2778540992","https://openalex.org/W2785822536","https://openalex.org/W2795994638","https://openalex.org/W2891890103","https://openalex.org/W2964324519","https://openalex.org/W4229654654","https://openalex.org/W4293318097","https://openalex.org/W6694513646","https://openalex.org/W6705026367","https://openalex.org/W6714836252","https://openalex.org/W6719908723","https://openalex.org/W6739622702","https://openalex.org/W6746239207","https://openalex.org/W6747195836","https://openalex.org/W6765060121","https://openalex.org/W7015949093"],"related_works":["https://openalex.org/W1978254186","https://openalex.org/W4245217724","https://openalex.org/W4390917331","https://openalex.org/W1984788566","https://openalex.org/W3135214639","https://openalex.org/W1966974913","https://openalex.org/W1901618855","https://openalex.org/W2937592512","https://openalex.org/W2132218766","https://openalex.org/W619618554"],"abstract_inverted_index":{"High":[0],"performance":[1,51,168],"multi-GPU":[2,61,71,83,171,193],"computing":[3],"becomes":[4,52],"an":[5],"inevitable":[6],"trend":[7],"due":[8],"to":[9,99,195,215,221],"the":[10,29,42,56,78,101,179,216,225,233],"ever-increasing":[11],"demand":[12],"on":[13,34,48,133],"computation":[14],"capability":[15],"in":[16,70,170],"emerging":[17],"domains":[18],"such":[19,202],"as":[20,125,127,203],"deep":[21,32],"learning,":[22],"big":[23],"data":[24],"and":[25,41,92,113,152,237],"planet-scale":[26],"applications.":[27,94],"However,":[28],"lack":[30],"of":[31,45,58,105,141],"understanding":[33],"how":[35],"modern":[36,106],"GPUs":[37],"can":[38,185],"be":[39,186,222],"connected":[40],"actual":[43],"impact":[44],"state-of-the-art":[46],"interconnect":[47,211],"multiGPU":[49],"application":[50],"a":[53,59,82],"hurdle.":[54],"Additionally,":[55],"absence":[57],"practical":[60],"benchmark":[62,84,230],"suite":[63,85,231],"poses":[64],"further":[65],"obstacles":[66],"for":[67,167,191,206],"conducting":[68],"research":[69],"era.":[72],"In":[73],"this":[74],"paper,":[75],"we":[76,136],"fill":[77],"gap":[79],"by":[80,148,159],"proposing":[81],"named":[86],"Tartan,":[87],"which":[88],"contains":[89],"microbenchmarks,":[90],"scale-up":[91,192],"scale-out":[93,208],"We":[95],"then":[96],"apply":[97],"Tartan":[98,229],"evaluate":[100],"four":[102,138],"latest":[103],"types":[104,140,145],"GPU":[107],"interconnects,":[108],"i.e.,":[109],"PCI-e,":[110],"NVLink-V1,":[111],"NVLink-V2":[112],"InfiniBand":[114],"with":[115],"GPUDirect-RDMA":[116,219],"from":[117,198],"two":[118],"recently":[119],"released":[120],"NVIDIA":[121],"super":[122],"AI":[123],"platforms":[124],"well":[126],"ORNL's":[128],"exascale":[129],"prototype":[130],"system.":[131],"Based":[132],"empirical":[134],"evaluation,":[135],"observe":[137],"new":[139],"NUMA":[142],"effects:":[143],"three":[144],"are":[146,164,235],"triggered":[147],"NVLink's":[149],"topology,":[150],"connectivity":[151],"routing,":[153],"while":[154,205],"one":[155],"type":[156],"is":[157,189,212],"caused":[158],"PCI-e":[160],"(i.e.,":[161],"anti-locality).":[162],"They":[163],"very":[165],"important":[166],"tuning":[169],"environment.":[172],"Our":[173],"evaluation":[174],"results":[175],"show":[176],"that,":[177],"unless":[178],"current":[180],"CPU-GPU":[181],"master-slave":[182],"programming":[183],"model":[184],"replaced,":[187],"it":[188],"difficult":[190],"applications":[194],"really":[196],"benefit":[197],"faster":[199],"intra-node":[200],"interconnects":[201],"NVLinks;":[204],"inter-node":[207],"applications,":[209],"although":[210],"more":[213],"crucial":[214],"overall":[217],"performance,":[218],"appears":[220],"not":[223],"always":[224],"optimal":[226],"choice.":[227],"The":[228],"including":[232],"microbenchmarks":[234],"opensource":[236],"available":[238],"athttp://github.com/uuudown/Tartan.":[239]},"counts_by_year":[{"year":2025,"cited_by_count":8},{"year":2024,"cited_by_count":5},{"year":2023,"cited_by_count":5},{"year":2022,"cited_by_count":5},{"year":2021,"cited_by_count":16},{"year":2020,"cited_by_count":15},{"year":2019,"cited_by_count":14}],"updated_date":"2026-07-14T23:27:15.235271","created_date":"2025-10-10T00:00:00"}
