{"id":"https://openalex.org/W3177007352","doi":"https://doi.org/10.1109/tpds.2021.3093239","title":"Tensorox: Accelerating GPU Applications via Neural Approximation on Unused Tensor Cores","display_name":"Tensorox: Accelerating GPU Applications via Neural Approximation on Unused Tensor Cores","publication_year":2021,"publication_date":"2021-06-29","ids":{"openalex":"https://openalex.org/W3177007352","doi":"https://doi.org/10.1109/tpds.2021.3093239","mag":"3177007352"},"language":"en","primary_location":{"id":"doi:10.1109/tpds.2021.3093239","is_oa":false,"landing_page_url":"https://doi.org/10.1109/tpds.2021.3093239","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":["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/A5065637540","display_name":"Nhut-Minh Ho","orcid":"https://orcid.org/0000-0002-3864-8027"},"institutions":[{"id":"https://openalex.org/I165932596","display_name":"National University of Singapore","ror":"https://ror.org/01tgyzw49","country_code":"SG","type":"education","lineage":["https://openalex.org/I165932596"]}],"countries":["SG"],"is_corresponding":false,"raw_author_name":"Nhut-Minh Ho","raw_affiliation_strings":["National University of Singapore, Singapore, Singapore"],"raw_orcid":"https://orcid.org/0000-0002-3864-8027","affiliations":[{"raw_affiliation_string":"National University of Singapore, Singapore, Singapore","institution_ids":["https://openalex.org/I165932596"]}]},{"author_position":"last","author":{"id":"https://openalex.org/A5023989495","display_name":"Weng\u2010Fai Wong","orcid":"https://orcid.org/0000-0002-4281-2053"},"institutions":[{"id":"https://openalex.org/I165932596","display_name":"National University of Singapore","ror":"https://ror.org/01tgyzw49","country_code":"SG","type":"education","lineage":["https://openalex.org/I165932596"]}],"countries":["SG"],"is_corresponding":false,"raw_author_name":"Weng-Fai Wong","raw_affiliation_strings":["National University of Singapore, Singapore, Singapore"],"raw_orcid":"https://orcid.org/0000-0002-4281-2053","affiliations":[{"raw_affiliation_string":"National University of Singapore, Singapore, Singapore","institution_ids":["https://openalex.org/I165932596"]}]}],"institutions":[],"countries_distinct_count":1,"institutions_distinct_count":1,"corresponding_author_ids":[],"corresponding_institution_ids":["https://openalex.org/I165932596"],"apc_list":null,"apc_paid":null,"fwci":1.8156,"has_fulltext":false,"cited_by_count":9,"citation_normalized_percentile":{"value":0.84752858,"is_in_top_1_percent":false,"is_in_top_10_percent":false},"cited_by_percentile_year":{"min":89,"max":97},"biblio":{"volume":"33","issue":"2","first_page":"429","last_page":"443"},"is_retracted":false,"is_paratext":false,"is_xpac":false,"primary_topic":{"id":"https://openalex.org/T10054","display_name":"Parallel Computing and Optimization Techniques","score":0.9983999729156494,"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":0.9983999729156494,"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/T12303","display_name":"Tensor decomposition and applications","score":0.9842000007629395,"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/T11181","display_name":"Advanced Data Storage Technologies","score":0.9764999747276306,"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/computer-science","display_name":"Computer science","score":0.8321852684020996},{"id":"https://openalex.org/keywords/tensor","display_name":"Tensor (intrinsic definition)","score":0.7208353877067566},{"id":"https://openalex.org/keywords/floating-point","display_name":"Floating point","score":0.6182986497879028},{"id":"https://openalex.org/keywords/artificial-neural-network","display_name":"Artificial neural network","score":0.5800914168357849},{"id":"https://openalex.org/keywords/parallel-computing","display_name":"Parallel computing","score":0.556240439414978},{"id":"https://openalex.org/keywords/deep-learning","display_name":"Deep learning","score":0.527091383934021},{"id":"https://openalex.org/keywords/exploit","display_name":"Exploit","score":0.5131619572639465},{"id":"https://openalex.org/keywords/graphics-processing-unit","display_name":"Graphics processing unit","score":0.4947955012321472},{"id":"https://openalex.org/keywords/function-approximation","display_name":"Function approximation","score":0.49457690119743347},{"id":"https://openalex.org/keywords/dimension","display_name":"Dimension (graph theory)","score":0.47964173555374146},{"id":"https://openalex.org/keywords/computational-science","display_name":"Computational science","score":0.4664912223815918},{"id":"https://openalex.org/keywords/scaling","display_name":"Scaling","score":0.4647727608680725},{"id":"https://openalex.org/keywords/function","display_name":"Function (biology)","score":0.4458235800266266},{"id":"https://openalex.org/keywords/matrix","display_name":"Matrix (chemical analysis)","score":0.4338535666465759},{"id":"https://openalex.org/keywords/central-processing-unit","display_name":"Central processing unit","score":0.41599202156066895},{"id":"https://openalex.org/keywords/algorithm","display_name":"Algorithm","score":0.3641078472137451},{"id":"https://openalex.org/keywords/artificial-intelligence","display_name":"Artificial intelligence","score":0.29214662313461304},{"id":"https://openalex.org/keywords/computer-hardware","display_name":"Computer hardware","score":0.14264869689941406},{"id":"https://openalex.org/keywords/mathematics","display_name":"Mathematics","score":0.10501191020011902}],"concepts":[{"id":"https://openalex.org/C41008148","wikidata":"https://www.wikidata.org/wiki/Q21198","display_name":"Computer science","level":0,"score":0.8321852684020996},{"id":"https://openalex.org/C155281189","wikidata":"https://www.wikidata.org/wiki/Q3518150","display_name":"Tensor (intrinsic definition)","level":2,"score":0.7208353877067566},{"id":"https://openalex.org/C84211073","wikidata":"https://www.wikidata.org/wiki/Q117879","display_name":"Floating point","level":2,"score":0.6182986497879028},{"id":"https://openalex.org/C50644808","wikidata":"https://www.wikidata.org/wiki/Q192776","display_name":"Artificial neural network","level":2,"score":0.5800914168357849},{"id":"https://openalex.org/C173608175","wikidata":"https://www.wikidata.org/wiki/Q232661","display_name":"Parallel computing","level":1,"score":0.556240439414978},{"id":"https://openalex.org/C108583219","wikidata":"https://www.wikidata.org/wiki/Q197536","display_name":"Deep learning","level":2,"score":0.527091383934021},{"id":"https://openalex.org/C165696696","wikidata":"https://www.wikidata.org/wiki/Q11287","display_name":"Exploit","level":2,"score":0.5131619572639465},{"id":"https://openalex.org/C2779851693","wikidata":"https://www.wikidata.org/wiki/Q183484","display_name":"Graphics processing unit","level":2,"score":0.4947955012321472},{"id":"https://openalex.org/C91873725","wikidata":"https://www.wikidata.org/wiki/Q3445816","display_name":"Function approximation","level":3,"score":0.49457690119743347},{"id":"https://openalex.org/C33676613","wikidata":"https://www.wikidata.org/wiki/Q13415176","display_name":"Dimension (graph theory)","level":2,"score":0.47964173555374146},{"id":"https://openalex.org/C459310","wikidata":"https://www.wikidata.org/wiki/Q117801","display_name":"Computational science","level":1,"score":0.4664912223815918},{"id":"https://openalex.org/C99844830","wikidata":"https://www.wikidata.org/wiki/Q102441924","display_name":"Scaling","level":2,"score":0.4647727608680725},{"id":"https://openalex.org/C14036430","wikidata":"https://www.wikidata.org/wiki/Q3736076","display_name":"Function (biology)","level":2,"score":0.4458235800266266},{"id":"https://openalex.org/C106487976","wikidata":"https://www.wikidata.org/wiki/Q685816","display_name":"Matrix (chemical analysis)","level":2,"score":0.4338535666465759},{"id":"https://openalex.org/C49154492","wikidata":"https://www.wikidata.org/wiki/Q5300","display_name":"Central processing unit","level":2,"score":0.41599202156066895},{"id":"https://openalex.org/C11413529","wikidata":"https://www.wikidata.org/wiki/Q8366","display_name":"Algorithm","level":1,"score":0.3641078472137451},{"id":"https://openalex.org/C154945302","wikidata":"https://www.wikidata.org/wiki/Q11660","display_name":"Artificial intelligence","level":1,"score":0.29214662313461304},{"id":"https://openalex.org/C9390403","wikidata":"https://www.wikidata.org/wiki/Q3966","display_name":"Computer hardware","level":1,"score":0.14264869689941406},{"id":"https://openalex.org/C33923547","wikidata":"https://www.wikidata.org/wiki/Q395","display_name":"Mathematics","level":0,"score":0.10501191020011902},{"id":"https://openalex.org/C202444582","wikidata":"https://www.wikidata.org/wiki/Q837863","display_name":"Pure mathematics","level":1,"score":0.0},{"id":"https://openalex.org/C38652104","wikidata":"https://www.wikidata.org/wiki/Q3510521","display_name":"Computer security","level":1,"score":0.0},{"id":"https://openalex.org/C2524010","wikidata":"https://www.wikidata.org/wiki/Q8087","display_name":"Geometry","level":1,"score":0.0},{"id":"https://openalex.org/C78458016","wikidata":"https://www.wikidata.org/wiki/Q840400","display_name":"Evolutionary biology","level":1,"score":0.0},{"id":"https://openalex.org/C192562407","wikidata":"https://www.wikidata.org/wiki/Q228736","display_name":"Materials science","level":0,"score":0.0},{"id":"https://openalex.org/C86803240","wikidata":"https://www.wikidata.org/wiki/Q420","display_name":"Biology","level":0,"score":0.0},{"id":"https://openalex.org/C159985019","wikidata":"https://www.wikidata.org/wiki/Q181790","display_name":"Composite material","level":1,"score":0.0}],"mesh":[],"locations_count":1,"locations":[{"id":"doi:10.1109/tpds.2021.3093239","is_oa":false,"landing_page_url":"https://doi.org/10.1109/tpds.2021.3093239","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"}],"best_oa_location":null,"sustainable_development_goals":[{"score":0.6100000143051147,"id":"https://metadata.un.org/sdg/9","display_name":"Industry, innovation and infrastructure"}],"awards":[{"id":"https://openalex.org/G3122060537","display_name":null,"funder_award_id":"T1-251RES1818","funder_id":"https://openalex.org/F4320320751","funder_display_name":"Ministry of Education - Singapore"},{"id":"https://openalex.org/G8928973550","display_name":null,"funder_award_id":"MOE2016-T2-2-150","funder_id":"https://openalex.org/F4320320751","funder_display_name":"Ministry of Education - Singapore"}],"funders":[{"id":"https://openalex.org/F4320320751","display_name":"Ministry of Education - Singapore","ror":"https://ror.org/01kcva023"}],"has_content":{"grobid_xml":false,"pdf":false},"content_urls":null,"referenced_works_count":65,"referenced_works":["https://openalex.org/W1522301498","https://openalex.org/W1791348790","https://openalex.org/W1841592590","https://openalex.org/W2026764611","https://openalex.org/W2030118646","https://openalex.org/W2081368694","https://openalex.org/W2101234009","https://openalex.org/W2142883190","https://openalex.org/W2170881177","https://openalex.org/W2172211804","https://openalex.org/W2172246201","https://openalex.org/W2187230075","https://openalex.org/W2194775991","https://openalex.org/W2233797083","https://openalex.org/W2244005500","https://openalex.org/W2265166184","https://openalex.org/W2276892413","https://openalex.org/W2322230929","https://openalex.org/W2554131156","https://openalex.org/W2606722458","https://openalex.org/W2625483679","https://openalex.org/W2757821154","https://openalex.org/W2762967556","https://openalex.org/W2766338242","https://openalex.org/W2771937810","https://openalex.org/W2777482433","https://openalex.org/W2796649226","https://openalex.org/W2799131675","https://openalex.org/W2803113791","https://openalex.org/W2888857701","https://openalex.org/W2895305554","https://openalex.org/W2899771611","https://openalex.org/W2901078150","https://openalex.org/W2901549770","https://openalex.org/W2917710824","https://openalex.org/W2919115771","https://openalex.org/W2920798074","https://openalex.org/W2923228966","https://openalex.org/W2945383908","https://openalex.org/W2950713841","https://openalex.org/W2952851876","https://openalex.org/W2963122961","https://openalex.org/W2963363373","https://openalex.org/W2963374099","https://openalex.org/W2963403868","https://openalex.org/W2963709416","https://openalex.org/W2963840672","https://openalex.org/W2964121744","https://openalex.org/W2999044305","https://openalex.org/W3001472289","https://openalex.org/W3104528661","https://openalex.org/W4236269878","https://openalex.org/W4385245566","https://openalex.org/W6631190155","https://openalex.org/W6638783484","https://openalex.org/W6675354045","https://openalex.org/W6685324067","https://openalex.org/W6696085341","https://openalex.org/W6739901393","https://openalex.org/W6746372235","https://openalex.org/W6750448596","https://openalex.org/W6751528251","https://openalex.org/W6756040250","https://openalex.org/W6760272376","https://openalex.org/W6761083414"],"related_works":["https://openalex.org/W2598944200","https://openalex.org/W1936382095","https://openalex.org/W2559348759","https://openalex.org/W2992303576","https://openalex.org/W4317790246","https://openalex.org/W2124331456","https://openalex.org/W4318602434","https://openalex.org/W2470372437","https://openalex.org/W4386530547","https://openalex.org/W764538942"],"abstract_inverted_index":{"Driven":[0],"by":[1,177],"the":[2,51,75,79,91,94,108,115,133,142,146,151,165,174,178],"demands":[3],"of":[4,78,93,107,145,148],"deep":[5,62],"learning,":[6],"many":[7],"hardware":[8],"accelerators,":[9],"including":[10],"GPUs,":[11],"have":[12,29],"begun":[13],"to":[14,20,81,103,161,164,180],"include":[15],"specialized":[16],"tensor":[17,41,53,98],"processing":[18],"units":[19],"accelerate":[21],"matrix":[22,35],"operations.":[23],"However,":[24],"general-purpose":[25],"GPU":[26],"applications":[27],"that":[28,49,127],"little":[30],"or":[31],"no":[32],"large":[33],"dense":[34],"operations":[36,99],"cannot":[37],"benefit":[38],"from":[39,159],"these":[40],"units.":[42],"This":[43],"article":[44],"proposes":[45],"Tensorox,":[46],"a":[47,67],"framework":[48],"exploits":[50],"half-precision":[52],"cores":[54],"available":[55],"on":[56,74],"recent":[57],"GPUs":[58,102],"for":[59,141],"approximable,":[60],"non":[61],"learning":[63],"applications.":[64,186],"In":[65],"essence,":[66],"shallow":[68],"neural":[69,110],"network":[70,111],"is":[71,90,128],"trained":[72],"based":[73],"input-output":[76],"mapping":[77],"function":[80],"be":[82],"approximated.":[83],"The":[84],"key":[85],"innovation":[86],"in":[87,100,112,167,184],"our":[88,121],"implementation":[89],"use":[92],"small":[95],"and":[96,118],"dimension-restricted":[97],"Nvidia":[101],"run":[104],"multiple":[105],"instances":[106],"approximation":[109,122,179],"parallel.":[113],"With":[114],"proper":[116],"scaling":[117],"training":[119],"methods,":[120],"yielded":[123],"an":[124],"overall":[125],"accuracy":[126],"higher":[129],"than":[130],"na\u00efvely":[131],"running":[132],"original":[134,166],"programs":[135],"with":[136],"half-precision.":[137],"Furthermore,":[138],"Tensorox":[139],"allows":[140],"runtime":[143],"adjustment":[144],"degree":[147],"approximation.":[149],"For":[150],"10":[152,182],"benchmarks":[153],"we":[154,156],"tested,":[155],"achieved":[157],"speedups":[158],"2\u00d7":[160],"112\u00d7":[162],"compared":[163],"single":[168],"precision":[169],"floating":[170],"point,":[171],"while":[172],"maintaining":[173],"error":[175],"caused":[176],"below":[181],"percent":[183],"most":[185]},"counts_by_year":[{"year":2026,"cited_by_count":1},{"year":2024,"cited_by_count":3},{"year":2023,"cited_by_count":2},{"year":2022,"cited_by_count":2},{"year":2021,"cited_by_count":1}],"updated_date":"2025-11-06T03:46:38.306776","created_date":"2025-10-10T00:00:00"}