{"id":"https://openalex.org/W4399426405","doi":"https://doi.org/10.1109/access.2024.3411473","title":"How Much Can We Gain From Tensor Kernel Fusion on GPUs?","display_name":"How Much Can We Gain From Tensor Kernel Fusion on GPUs?","publication_year":2024,"publication_date":"2024-01-01","ids":{"openalex":"https://openalex.org/W4399426405","doi":"https://doi.org/10.1109/access.2024.3411473"},"language":"en","primary_location":{"id":"doi:10.1109/access.2024.3411473","is_oa":true,"landing_page_url":"https://doi.org/10.1109/access.2024.3411473","pdf_url":null,"source":{"id":"https://openalex.org/S2485537415","display_name":"IEEE Access","issn_l":"2169-3536","issn":["2169-3536"],"is_oa":true,"is_in_doaj":true,"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 Access","raw_type":"journal-article"},"type":"article","indexed_in":["crossref","doaj"],"open_access":{"is_oa":true,"oa_status":"gold","oa_url":"https://doi.org/10.1109/access.2024.3411473","any_repository_has_fulltext":true},"authorships":[{"author_position":"first","author":{"id":"https://openalex.org/A5059737143","display_name":"Wei Sun","orcid":"https://orcid.org/0000-0003-2873-6418"},"institutions":[{"id":"https://openalex.org/I83019370","display_name":"Eindhoven University of Technology","ror":"https://ror.org/02c2kyt77","country_code":"NL","type":"education","lineage":["https://openalex.org/I83019370"]}],"countries":["NL"],"is_corresponding":false,"raw_author_name":"Wei Sun","raw_affiliation_strings":["Electronic System Group, Eindhoven University of Technology, Eindhoven, The Netherlands"],"raw_orcid":"https://orcid.org/0000-0003-2873-6418","affiliations":[{"raw_affiliation_string":"Electronic System Group, Eindhoven University of Technology, Eindhoven, The Netherlands","institution_ids":["https://openalex.org/I83019370"]}]},{"author_position":"middle","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":["Physical and Computational Sciences Directorate, Pacific Northwest National Laboratory, Richland, WA, USA"],"raw_orcid":"https://orcid.org/0000-0003-3734-9137","affiliations":[{"raw_affiliation_string":"Physical and Computational Sciences Directorate, Pacific Northwest National Laboratory, Richland, WA, USA","institution_ids":["https://openalex.org/I142606810"]}]},{"author_position":"middle","author":{"id":"https://openalex.org/A5032436846","display_name":"Sander Stuijk","orcid":"https://orcid.org/0000-0002-2518-6847"},"institutions":[{"id":"https://openalex.org/I83019370","display_name":"Eindhoven University of Technology","ror":"https://ror.org/02c2kyt77","country_code":"NL","type":"education","lineage":["https://openalex.org/I83019370"]}],"countries":["NL"],"is_corresponding":false,"raw_author_name":"Sander Stuijk","raw_affiliation_strings":["Electronic System Group, Eindhoven University of Technology, Eindhoven, The Netherlands"],"raw_orcid":"https://orcid.org/0000-0002-2518-6847","affiliations":[{"raw_affiliation_string":"Electronic System Group, Eindhoven University of Technology, Eindhoven, The Netherlands","institution_ids":["https://openalex.org/I83019370"]}]},{"author_position":"last","author":{"id":"https://openalex.org/A5081768631","display_name":"Henk Corporaal","orcid":"https://orcid.org/0000-0003-4506-5732"},"institutions":[{"id":"https://openalex.org/I83019370","display_name":"Eindhoven University of Technology","ror":"https://ror.org/02c2kyt77","country_code":"NL","type":"education","lineage":["https://openalex.org/I83019370"]}],"countries":["NL"],"is_corresponding":false,"raw_author_name":"Henk Corporaal","raw_affiliation_strings":["Electronic System Group, Eindhoven University of Technology, Eindhoven, The Netherlands"],"raw_orcid":"https://orcid.org/0000-0003-4506-5732","affiliations":[{"raw_affiliation_string":"Electronic System Group, Eindhoven University of Technology, Eindhoven, The Netherlands","institution_ids":["https://openalex.org/I83019370"]}]}],"institutions":[],"countries_distinct_count":2,"institutions_distinct_count":2,"corresponding_author_ids":[],"corresponding_institution_ids":[],"apc_list":{"value":1850,"currency":"USD","value_usd":1850},"apc_paid":{"value":1850,"currency":"USD","value_usd":1850},"fwci":1.9752,"has_fulltext":true,"cited_by_count":5,"citation_normalized_percentile":{"value":0.86859187,"is_in_top_1_percent":false,"is_in_top_10_percent":false},"cited_by_percentile_year":{"min":90,"max":99},"biblio":{"volume":"12","issue":null,"first_page":"126135","last_page":"126144"},"is_retracted":false,"is_paratext":false,"is_xpac":false,"primary_topic":{"id":"https://openalex.org/T10054","display_name":"Parallel Computing and Optimization Techniques","score":0.9972000122070312,"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":0.9972000122070312,"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":0.9929999709129333,"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/T10036","display_name":"Advanced Neural Network Applications","score":0.9904000163078308,"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"}}],"keywords":[{"id":"https://openalex.org/keywords/computer-science","display_name":"Computer science","score":0.647416889667511},{"id":"https://openalex.org/keywords/kernel","display_name":"Kernel (algebra)","score":0.5802175402641296},{"id":"https://openalex.org/keywords/fusion","display_name":"Fusion","score":0.5229119658470154},{"id":"https://openalex.org/keywords/tensor","display_name":"Tensor (intrinsic definition)","score":0.4947595000267029},{"id":"https://openalex.org/keywords/parallel-computing","display_name":"Parallel computing","score":0.33702027797698975},{"id":"https://openalex.org/keywords/mathematics","display_name":"Mathematics","score":0.15388697385787964},{"id":"https://openalex.org/keywords/pure-mathematics","display_name":"Pure mathematics","score":0.08691221475601196}],"concepts":[{"id":"https://openalex.org/C41008148","wikidata":"https://www.wikidata.org/wiki/Q21198","display_name":"Computer science","level":0,"score":0.647416889667511},{"id":"https://openalex.org/C74193536","wikidata":"https://www.wikidata.org/wiki/Q574844","display_name":"Kernel (algebra)","level":2,"score":0.5802175402641296},{"id":"https://openalex.org/C158525013","wikidata":"https://www.wikidata.org/wiki/Q2593739","display_name":"Fusion","level":2,"score":0.5229119658470154},{"id":"https://openalex.org/C155281189","wikidata":"https://www.wikidata.org/wiki/Q3518150","display_name":"Tensor (intrinsic definition)","level":2,"score":0.4947595000267029},{"id":"https://openalex.org/C173608175","wikidata":"https://www.wikidata.org/wiki/Q232661","display_name":"Parallel computing","level":1,"score":0.33702027797698975},{"id":"https://openalex.org/C33923547","wikidata":"https://www.wikidata.org/wiki/Q395","display_name":"Mathematics","level":0,"score":0.15388697385787964},{"id":"https://openalex.org/C202444582","wikidata":"https://www.wikidata.org/wiki/Q837863","display_name":"Pure mathematics","level":1,"score":0.08691221475601196},{"id":"https://openalex.org/C138885662","wikidata":"https://www.wikidata.org/wiki/Q5891","display_name":"Philosophy","level":0,"score":0.0},{"id":"https://openalex.org/C41895202","wikidata":"https://www.wikidata.org/wiki/Q8162","display_name":"Linguistics","level":1,"score":0.0}],"mesh":[],"locations_count":6,"locations":[{"id":"doi:10.1109/access.2024.3411473","is_oa":true,"landing_page_url":"https://doi.org/10.1109/access.2024.3411473","pdf_url":null,"source":{"id":"https://openalex.org/S2485537415","display_name":"IEEE Access","issn_l":"2169-3536","issn":["2169-3536"],"is_oa":true,"is_in_doaj":true,"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 Access","raw_type":"journal-article"},{"id":"pmh:oai:pure.tue.nl:openaire/c3951112-e7d6-4849-9ba7-27d97524fb93","is_oa":true,"landing_page_url":"https://research.tue.nl/en/publications/c3951112-e7d6-4849-9ba7-27d97524fb93","pdf_url":"https://pure.tue.nl/ws/files/342691262/How_Much_Can_We_Gain_From_Tensor_Kernel_Fusion_on_GPUs.pdf","source":{"id":"https://openalex.org/S4406922641","display_name":"TU/e Research Portal","issn_l":null,"issn":null,"is_oa":false,"is_in_doaj":false,"is_core":false,"host_organization":null,"host_organization_name":null,"host_organization_lineage":[],"host_organization_lineage_names":[],"type":"repository"},"license":"cc-by-nc-nd","license_id":"https://openalex.org/licenses/cc-by-nc-nd","version":"submittedVersion","is_accepted":false,"is_published":false,"raw_source_name":"Sun, W, Li, A, Stuijk, S & Corporaal, H 2024, 'How much can we gain from Tensor Kernel Fusion on GPUs?', IEEE Access, vol. 12, 10551817, pp. 126135-126144. https://doi.org/10.1109/ACCESS.2024.3411473","raw_type":"info:eu-repo/semantics/article"},{"id":"pmh:oai:doaj.org/article:c28dd8fd91494bd4a7834ab6e0dfcf1d","is_oa":true,"landing_page_url":"https://doaj.org/article/c28dd8fd91494bd4a7834ab6e0dfcf1d","pdf_url":null,"source":{"id":"https://openalex.org/S4306401280","display_name":"DOAJ (DOAJ: Directory of Open Access Journals)","issn_l":null,"issn":null,"is_oa":false,"is_in_doaj":false,"is_core":false,"host_organization":null,"host_organization_name":null,"host_organization_lineage":[],"host_organization_lineage_names":[],"type":"repository"},"license":"cc-by-sa","license_id":"https://openalex.org/licenses/cc-by-sa","version":"submittedVersion","is_accepted":false,"is_published":false,"raw_source_name":"IEEE Access, Vol 12, Pp 126135-126144 (2024)","raw_type":"article"},{"id":"pmh:oai:osti.gov:2446540","is_oa":true,"landing_page_url":"https://www.osti.gov/biblio/2446540","pdf_url":null,"source":{"id":"https://openalex.org/S4306402487","display_name":"OSTI OAI (U.S. Department of Energy Office of Scientific and Technical Information)","issn_l":null,"issn":null,"is_oa":false,"is_in_doaj":false,"is_core":false,"host_organization":"https://openalex.org/I139351228","host_organization_name":"Office of Scientific and Technical Information","host_organization_lineage":["https://openalex.org/I139351228"],"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":null},{"id":"pmh:oai:osti.gov:2446541","is_oa":true,"landing_page_url":"https://www.osti.gov/biblio/2446541","pdf_url":null,"source":{"id":"https://openalex.org/S4306402487","display_name":"OSTI OAI (U.S. Department of Energy Office of Scientific and Technical Information)","issn_l":null,"issn":null,"is_oa":false,"is_in_doaj":false,"is_core":false,"host_organization":"https://openalex.org/I139351228","host_organization_name":"Office of Scientific and Technical Information","host_organization_lineage":["https://openalex.org/I139351228"],"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":null},{"id":"pmh:oai:osti.gov:2475359","is_oa":true,"landing_page_url":"https://www.osti.gov/biblio/2475359","pdf_url":null,"source":{"id":"https://openalex.org/S4306402487","display_name":"OSTI OAI (U.S. Department of Energy Office of Scientific and Technical Information)","issn_l":null,"issn":null,"is_oa":false,"is_in_doaj":false,"is_core":false,"host_organization":"https://openalex.org/I139351228","host_organization_name":"Office of Scientific and Technical Information","host_organization_lineage":["https://openalex.org/I139351228"],"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":null}],"best_oa_location":{"id":"doi:10.1109/access.2024.3411473","is_oa":true,"landing_page_url":"https://doi.org/10.1109/access.2024.3411473","pdf_url":null,"source":{"id":"https://openalex.org/S2485537415","display_name":"IEEE Access","issn_l":"2169-3536","issn":["2169-3536"],"is_oa":true,"is_in_doaj":true,"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 Access","raw_type":"journal-article"},"sustainable_development_goals":[{"id":"https://metadata.un.org/sdg/7","display_name":"Affordable and clean energy","score":0.5099999904632568}],"awards":[{"id":"https://openalex.org/G1791724952","display_name":null,"funder_award_id":"78284","funder_id":"https://openalex.org/F4320337506","funder_display_name":"Advanced Scientific Computing Research"},{"id":"https://openalex.org/G2014121706","display_name":null,"funder_award_id":"ZERO-ARM P3","funder_id":"https://openalex.org/F4320321800","funder_display_name":"Nederlandse Organisatie voor Wetenschappelijk Onderzoek"},{"id":"https://openalex.org/G3620080081","display_name":null,"funder_award_id":"78284","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/F4320321800","display_name":"Nederlandse Organisatie voor Wetenschappelijk Onderzoek","ror":"https://ror.org/04jsz6e67"},{"id":"https://openalex.org/F4320332359","display_name":"Office of Science","ror":"https://ror.org/00mmn6b08"},{"id":"https://openalex.org/F4320337506","display_name":"Advanced Scientific Computing Research","ror":"https://ror.org/0012c7r22"}],"has_content":{"grobid_xml":false,"pdf":false},"content_urls":null,"referenced_works_count":19,"referenced_works":["https://openalex.org/W2002555321","https://openalex.org/W2194775991","https://openalex.org/W2896457183","https://openalex.org/W3047196046","https://openalex.org/W3085139254","https://openalex.org/W3129181244","https://openalex.org/W3137147200","https://openalex.org/W4205146836","https://openalex.org/W4309672181","https://openalex.org/W4385245566","https://openalex.org/W4387436180","https://openalex.org/W4400409880","https://openalex.org/W6751349269","https://openalex.org/W6755207826","https://openalex.org/W6766978945","https://openalex.org/W6778883912","https://openalex.org/W6790521546","https://openalex.org/W6791940793","https://openalex.org/W6838322825"],"related_works":["https://openalex.org/W4391375266","https://openalex.org/W2899084033","https://openalex.org/W2748952813","https://openalex.org/W2390279801","https://openalex.org/W4391913857","https://openalex.org/W2358668433","https://openalex.org/W4396701345","https://openalex.org/W2376932109","https://openalex.org/W2001405890","https://openalex.org/W4396696052"],"abstract_inverted_index":{"Kernel":[0],"fusion":[1,153,178],"is":[2,105,130,165,175],"a":[3,22,87,106,119,132,215,232],"crucial":[4],"optimization":[5],"technique":[6],"for":[7,36,208,241],"GPU":[8,72,121],"applications,":[9,128],"particularly":[10],"deep":[11,84],"neural":[12],"networks,":[13],"where":[14],"it":[15,129,161,166],"involves":[16],"combining":[17,101,114],"multiple":[18],"consecutive":[19],"kernels":[20,46,73,117],"into":[21,75],"single":[23,120],"larger":[24],"kernel.":[25],"This":[26,56],"approach":[27,141],"aims":[28],"to":[29,61,109],"enhance":[30],"performance":[31,239],"by":[32],"reducing":[33],"the":[34,59,197,220,257],"need":[35],"slow":[37],"off-chip":[38],"memory":[39,52],"accesses.":[40],"Instead,":[41],"intermediate":[42],"results":[43],"between":[44],"successive":[45],"are":[47],"stored":[48],"in":[49,118,126,261],"faster":[50],"on-chip":[51],"like":[53],"shared":[54],"memory.":[55],"strategy":[57,218],"has":[58,123],"potential":[60],"not":[62,131,167],"only":[63],"boost":[64],"performance,":[65],"but":[66],"also":[67],"reduce":[68],"energy":[69],"consumption.":[70],"Typically,":[71],"fall":[74],"two":[76,115],"categories:":[77],"tensor":[78,88,116,151,176,228],"operations":[79],"and":[80,137,163,188,210,243],"element":[81,93],"operations.":[82],"In":[83,171],"learning,":[85],"fusing":[86],"operation":[89,94],"kernel":[90,95,122,152,177,199,229],"with":[91,103,201,231,247],"an":[92],"that":[96,207,252],"follows":[97],"it,":[98],"such":[99],"as":[100,256],"convolution":[102],"ReLU,":[104],"common":[107],"practice":[108],"achieve":[110],"improved":[111],"performance.":[112],"While":[113],"shown":[124],"benefits":[125,254],"certain":[127],"straightforward":[133],"task.":[134],"The":[135],"advantages":[136,149],"limitations":[138,159],"of":[139],"this":[140],"remain":[142],"unclear,":[143],"prompting":[144],"several":[145],"questions:":[146],"1)":[147],"What":[148,158],"does":[150,160],"offer":[154],"on":[155,191],"GPGPUs?":[156],"2)":[157],"have":[162],"why":[164],"widely":[168],"adopted?":[169],"3)":[170],"what":[172],"practical":[173],"scenarios":[174],"beneficial?":[179],"To":[180],"address":[181],"these":[182,253],"questions,":[183],"we":[184,236,250],"conducted":[185],"both":[186],"analytical":[187],"experimental":[189,204],"studies":[190],"Nvidia":[192],"Tensor":[193],"Core":[194],"GPUs,":[195],"using":[196],"CUTLASS":[198],"library":[200],"extensions.":[202],"Our":[203],"findings":[205],"revealed":[206],"tall":[209,242],"narrow":[211,244],"matrix":[212,245,258],"multiplications,":[213],"employing":[214],"1D":[216,233],"tiling":[217,224,234],"outperforms":[219],"commonly":[221],"used":[222],"2D":[223],"strategy.":[225],"By":[226],"comparing":[227],"fusions":[230],"baseline,":[235],"demonstrated":[237],"significant":[238],"gains":[240],"multiplications":[246],"fusion.":[248],"However,":[249],"observe":[251],"diminish":[255],"sizes":[259],"increase":[260],"width.":[262]},"counts_by_year":[{"year":2026,"cited_by_count":3},{"year":2025,"cited_by_count":1},{"year":2024,"cited_by_count":1}],"updated_date":"2025-12-26T23:08:49.675405","created_date":"2025-10-10T00:00:00"}