{"id":"https://openalex.org/W4297097055","doi":"https://doi.org/10.1109/hcs55958.2022.9895617","title":"VTA-NIC: Deep Learning Inference Serving in Network Interface Cards","display_name":"VTA-NIC: Deep Learning Inference Serving in Network Interface Cards","publication_year":2022,"publication_date":"2022-08-21","ids":{"openalex":"https://openalex.org/W4297097055","doi":"https://doi.org/10.1109/hcs55958.2022.9895617"},"language":"en","primary_location":{"id":"doi:10.1109/hcs55958.2022.9895617","is_oa":false,"landing_page_url":"https://doi.org/10.1109/hcs55958.2022.9895617","pdf_url":null,"source":null,"license":null,"license_id":null,"version":"publishedVersion","is_accepted":true,"is_published":true,"raw_source_name":"2022 IEEE Hot Chips 34 Symposium (HCS)","raw_type":"proceedings-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/A5069099139","display_name":"Kenji Tanaka","orcid":"https://orcid.org/0000-0003-2320-4535"},"institutions":[],"countries":[],"is_corresponding":true,"raw_author_name":"Kenji Tanaka","raw_affiliation_strings":["NTT Device Technology Labs"],"affiliations":[{"raw_affiliation_string":"NTT Device Technology Labs","institution_ids":[]}]},{"author_position":"middle","author":{"id":"https://openalex.org/A5077656627","display_name":"Yuki Arikawa","orcid":null},"institutions":[],"countries":[],"is_corresponding":false,"raw_author_name":"Yuki Arikawa","raw_affiliation_strings":["NTT Device Technology Labs"],"affiliations":[{"raw_affiliation_string":"NTT Device Technology Labs","institution_ids":[]}]},{"author_position":"middle","author":{"id":"https://openalex.org/A5036198872","display_name":"Kazutaka Morita","orcid":null},"institutions":[],"countries":[],"is_corresponding":false,"raw_author_name":"Kazutaka Morita","raw_affiliation_strings":["NTT Software Innovation Center"],"affiliations":[{"raw_affiliation_string":"NTT Software Innovation Center","institution_ids":[]}]},{"author_position":"middle","author":{"id":"https://openalex.org/A5101100781","display_name":"Tsuyoshi Ito","orcid":null},"institutions":[],"countries":[],"is_corresponding":false,"raw_author_name":"Tsuyoshi Ito","raw_affiliation_strings":["NTT Device Technology Labs"],"affiliations":[{"raw_affiliation_string":"NTT Device Technology Labs","institution_ids":[]}]},{"author_position":"middle","author":{"id":"https://openalex.org/A5005066302","display_name":"Takashi Uchida","orcid":"https://orcid.org/0000-0002-8257-7334"},"institutions":[{"id":"https://openalex.org/I4388482706","display_name":"ARES (United States)","ror":"https://ror.org/03pnvey90","country_code":null,"type":"company","lineage":["https://openalex.org/I4388482706"]},{"id":"https://openalex.org/I4210126820","display_name":"Fixstars Solutions (United States)","ror":"https://ror.org/035myfa14","country_code":"US","type":"company","lineage":["https://openalex.org/I4210126820"]}],"countries":["US"],"is_corresponding":false,"raw_author_name":"Takashi Uchida","raw_affiliation_strings":["Fixstars Corporation"],"affiliations":[{"raw_affiliation_string":"Fixstars Corporation","institution_ids":["https://openalex.org/I4210126820","https://openalex.org/I4388482706"]}]},{"author_position":"middle","author":{"id":"https://openalex.org/A5076106576","display_name":"Natsuko Saito","orcid":"https://orcid.org/0009-0005-4059-7491"},"institutions":[{"id":"https://openalex.org/I4210126820","display_name":"Fixstars Solutions (United States)","ror":"https://ror.org/035myfa14","country_code":"US","type":"company","lineage":["https://openalex.org/I4210126820"]},{"id":"https://openalex.org/I4388482706","display_name":"ARES (United States)","ror":"https://ror.org/03pnvey90","country_code":null,"type":"company","lineage":["https://openalex.org/I4388482706"]}],"countries":["US"],"is_corresponding":false,"raw_author_name":"Natsuko Saito","raw_affiliation_strings":["Fixstars Corporation"],"affiliations":[{"raw_affiliation_string":"Fixstars Corporation","institution_ids":["https://openalex.org/I4210126820","https://openalex.org/I4388482706"]}]},{"author_position":"middle","author":{"id":"https://openalex.org/A5015894637","display_name":"Shinya Kaji","orcid":null},"institutions":[{"id":"https://openalex.org/I4388482706","display_name":"ARES (United States)","ror":"https://ror.org/03pnvey90","country_code":null,"type":"company","lineage":["https://openalex.org/I4388482706"]},{"id":"https://openalex.org/I4210126820","display_name":"Fixstars Solutions (United States)","ror":"https://ror.org/035myfa14","country_code":"US","type":"company","lineage":["https://openalex.org/I4210126820"]}],"countries":["US"],"is_corresponding":false,"raw_author_name":"Shinya Kaji","raw_affiliation_strings":["Fixstars Corporation"],"affiliations":[{"raw_affiliation_string":"Fixstars Corporation","institution_ids":["https://openalex.org/I4210126820","https://openalex.org/I4388482706"]}]},{"author_position":"last","author":{"id":"https://openalex.org/A5037700663","display_name":"Takeshi Sakamoto","orcid":"https://orcid.org/0000-0001-6312-2249"},"institutions":[],"countries":[],"is_corresponding":false,"raw_author_name":"Takeshi Sakamoto","raw_affiliation_strings":["NTT Device Technology Labs"],"affiliations":[{"raw_affiliation_string":"NTT Device Technology Labs","institution_ids":[]}]}],"institutions":[],"countries_distinct_count":1,"institutions_distinct_count":8,"corresponding_author_ids":["https://openalex.org/A5069099139"],"corresponding_institution_ids":[],"apc_list":null,"apc_paid":null,"fwci":0.0,"has_fulltext":false,"cited_by_count":0,"citation_normalized_percentile":{"value":0.10013028,"is_in_top_1_percent":false,"is_in_top_10_percent":false},"cited_by_percentile_year":null,"biblio":{"volume":null,"issue":null,"first_page":"1","last_page":"16"},"is_retracted":false,"is_paratext":false,"is_xpac":false,"primary_topic":{"id":"https://openalex.org/T10054","display_name":"Parallel Computing and Optimization Techniques","score":0.9994000196456909,"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":0.9994000196456909,"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/T10829","display_name":"Interconnection Networks and Systems","score":0.9940000176429749,"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/T12326","display_name":"Network Packet Processing and Optimization","score":0.9934999942779541,"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":0.6568639874458313},{"id":"https://openalex.org/keywords/latency","display_name":"Latency (audio)","score":0.656779944896698},{"id":"https://openalex.org/keywords/inference","display_name":"Inference","score":0.47209638357162476},{"id":"https://openalex.org/keywords/microservices","display_name":"Microservices","score":0.4435116946697235},{"id":"https://openalex.org/keywords/operating-system","display_name":"Operating system","score":0.3811536729335785},{"id":"https://openalex.org/keywords/artificial-intelligence","display_name":"Artificial intelligence","score":0.3226587474346161},{"id":"https://openalex.org/keywords/cloud-computing","display_name":"Cloud computing","score":0.1670438051223755}],"concepts":[{"id":"https://openalex.org/C41008148","wikidata":"https://www.wikidata.org/wiki/Q21198","display_name":"Computer science","level":0,"score":0.6568639874458313},{"id":"https://openalex.org/C82876162","wikidata":"https://www.wikidata.org/wiki/Q17096504","display_name":"Latency (audio)","level":2,"score":0.656779944896698},{"id":"https://openalex.org/C2776214188","wikidata":"https://www.wikidata.org/wiki/Q408386","display_name":"Inference","level":2,"score":0.47209638357162476},{"id":"https://openalex.org/C2778505942","wikidata":"https://www.wikidata.org/wiki/Q18344624","display_name":"Microservices","level":3,"score":0.4435116946697235},{"id":"https://openalex.org/C111919701","wikidata":"https://www.wikidata.org/wiki/Q9135","display_name":"Operating system","level":1,"score":0.3811536729335785},{"id":"https://openalex.org/C154945302","wikidata":"https://www.wikidata.org/wiki/Q11660","display_name":"Artificial intelligence","level":1,"score":0.3226587474346161},{"id":"https://openalex.org/C79974875","wikidata":"https://www.wikidata.org/wiki/Q483639","display_name":"Cloud computing","level":2,"score":0.1670438051223755},{"id":"https://openalex.org/C76155785","wikidata":"https://www.wikidata.org/wiki/Q418","display_name":"Telecommunications","level":1,"score":0.0}],"mesh":[],"locations_count":1,"locations":[{"id":"doi:10.1109/hcs55958.2022.9895617","is_oa":false,"landing_page_url":"https://doi.org/10.1109/hcs55958.2022.9895617","pdf_url":null,"source":null,"license":null,"license_id":null,"version":"publishedVersion","is_accepted":true,"is_published":true,"raw_source_name":"2022 IEEE Hot Chips 34 Symposium (HCS)","raw_type":"proceedings-article"}],"best_oa_location":null,"sustainable_development_goals":[{"display_name":"Affordable and clean energy","id":"https://metadata.un.org/sdg/7","score":0.7099999785423279}],"awards":[],"funders":[],"has_content":{"pdf":false,"grobid_xml":false},"content_urls":null,"referenced_works_count":3,"referenced_works":["https://openalex.org/W2794670651","https://openalex.org/W3033527233","https://openalex.org/W4223585153"],"related_works":["https://openalex.org/W4210334847","https://openalex.org/W3181741639","https://openalex.org/W4367181468","https://openalex.org/W2767473057","https://openalex.org/W4225586032","https://openalex.org/W3003367851","https://openalex.org/W1492892507","https://openalex.org/W3119944216","https://openalex.org/W3128807919","https://openalex.org/W3176411177"],"abstract_inverted_index":{"VTA-NIC":[0,54],"Chip":[1],"ArchitectureWe":[2],"aim":[3],"to":[4,87,169],"achieve":[5],"DL":[6],"inference":[7],"serving":[8,192],"(DLIS)":[9,99],"without":[10],"CPU":[11],"interference.We":[12],"integrate":[13],"hardware":[14],"data":[15,129,131],"paths":[16],"as":[17],"a":[18,23,112],"NIC":[19],"(Network":[20],"Interface":[21],"Card),":[22],"REST":[24],"API":[25],"parser/deparser":[26],"and":[27,188],"multiple":[28],"VTAs":[29],"(Versatile":[30],"Tensor":[31],"Accelerators).ConfigurationVTA-NICProcess":[32],"node16":[33],"nm":[34],"FinFET":[35],"@Xilinx":[36],"FPGANumber":[37],"of":[38,53,60,70,102,147,160,180,190],"Cores8":[39],"VTA":[40],"CoresCore":[41],"Frequency213":[42],"MHzMACs":[43],"per":[44],"core169Memory":[45],"Throughput19.2GB/s":[46],"(DDR4-2400)Number":[47],"precisionINT8PerformancePower":[48],"EfficiencyThe":[49],"DLIS":[50,149,164,182],"power":[51],"efficiency":[52],"is":[55,81,85,100,109,194],"6.1x":[56],"better":[57],"than":[58],"that":[59],"GPU":[61],"(Nvidia":[62],"V100).Tail":[63],"LatencyAt":[64],"high":[65],"load,":[66],"the":[67,78,88,136,148,158,176,191],"tail":[68,79,165,186],"latency":[69,80,150,166],"heterogeneous":[71],"systems":[72],"unexpectedly":[73],"increases.":[74],"With":[75],"our":[76,143],"chip,":[77],"predictable":[82],"since":[83],"it":[84],"proportional":[86],"load.BackgroundRecently,":[89],"web":[90],"applications":[91],"are":[92,121],"often":[93],"built":[94],"on":[95],"microservices.DL":[96],"Inference":[97,204],"Serving":[98],"one":[101],"those":[103],"microservices":[104],"<sup":[105,116,139,172,197],"xmlns:mml=\"http://www.w3.org/1998/Math/MathML\"":[106,117,140,173,198],"xmlns:xlink=\"http://www.w3.org/1999/xlink\">1</sup>":[107],".DLIS":[108],"provisioned":[110],"with":[111],"special":[113],"accelerator":[114],"instance":[115],"xmlns:xlink=\"http://www.w3.org/1999/xlink\">2</sup>":[118],".The":[119],"microservices/instances":[120],"loosely":[122],"coupled":[123],"via":[124,132],"APIs.BackgroundAccelerator":[125],"instances":[126],"risk":[127],"inefficient":[128],"movement.1.Moving":[130],"host":[133,161],"processors":[134,162],"decreases":[135],"accelerator's":[137],"utilization.":[138],"xmlns:xlink=\"http://www.w3.org/1999/xlink\">3</sup>":[141],"a.In":[142],"preliminary":[144],"experiments,":[145],"half":[146,189],"was":[151],"caused":[152],"by":[153,167],"moving":[154],"data.2.Under":[155],"high-load":[156],"conditions,":[157],"interference":[159],"degrades":[163],"up":[168],"100":[170],"times":[171],"xmlns:xlink=\"http://www.w3.org/1999/xlink\">4</sup>":[174],".a.In":[175],"real":[177],"cloud,":[178],"9%":[179],"light":[181],"tasks":[183],"suffer":[184],"server":[185],"latency,":[187],"time":[193],"waiting":[195],"time.":[196],"xmlns:xlink=\"http://www.w3.org/1999/xlink\">5</sup>":[199],"Model:":[200],"ResNet-18@TensorRTPrecision:":[201],"INT8System:":[202],"Triton":[203],"ServerAccelerator:":[205],"Nvidia":[206],"V100":[207]},"counts_by_year":[],"updated_date":"2025-11-06T03:46:38.306776","created_date":"2025-10-10T00:00:00"}