{"id":"https://openalex.org/W2093843662","doi":"https://doi.org/10.1109/cgo.2013.6494986","title":"Performance upper bound analysis and optimization of SGEMM on Fermi and Kepler GPUs","display_name":"Performance upper bound analysis and optimization of SGEMM on Fermi and Kepler GPUs","publication_year":2013,"publication_date":"2013-02-01","ids":{"openalex":"https://openalex.org/W2093843662","doi":"https://doi.org/10.1109/cgo.2013.6494986","mag":"2093843662"},"language":"en","primary_location":{"id":"doi:10.1109/cgo.2013.6494986","is_oa":false,"landing_page_url":"https://doi.org/10.1109/cgo.2013.6494986","pdf_url":null,"source":null,"license":null,"license_id":null,"version":"publishedVersion","is_accepted":true,"is_published":true,"raw_source_name":"Proceedings of the 2013 IEEE/ACM International Symposium on Code Generation and Optimization (CGO)","raw_type":"proceedings-article"},"type":"preprint","indexed_in":["crossref"],"open_access":{"is_oa":true,"oa_status":"green","oa_url":"https://inria.hal.science/hal-00789958","any_repository_has_fulltext":true},"authorships":[{"author_position":"first","author":{"id":"https://openalex.org/A5103287856","display_name":"Junjie Lai","orcid":null},"institutions":[{"id":"https://openalex.org/I1326498283","display_name":"Institut national de recherche en informatique et en automatique","ror":"https://ror.org/02kvxyf05","country_code":"FR","type":"funder","lineage":["https://openalex.org/I1326498283"]}],"countries":["FR"],"is_corresponding":true,"raw_author_name":"Junjie Lai","raw_affiliation_strings":["INRIA, France","[INRIA-Sophia Antipolis, France]"],"affiliations":[{"raw_affiliation_string":"INRIA, France","institution_ids":["https://openalex.org/I1326498283"]},{"raw_affiliation_string":"[INRIA-Sophia Antipolis, France]","institution_ids":["https://openalex.org/I1326498283"]}]},{"author_position":"last","author":{"id":"https://openalex.org/A5075222745","display_name":"Andr\u00e9 Seznec","orcid":"https://orcid.org/0000-0002-3058-6503"},"institutions":[{"id":"https://openalex.org/I1326498283","display_name":"Institut national de recherche en informatique et en automatique","ror":"https://ror.org/02kvxyf05","country_code":"FR","type":"funder","lineage":["https://openalex.org/I1326498283"]}],"countries":["FR"],"is_corresponding":false,"raw_author_name":"A. Seznec","raw_affiliation_strings":["INRIA, France","[INRIA-Sophia Antipolis, France]"],"affiliations":[{"raw_affiliation_string":"INRIA, France","institution_ids":["https://openalex.org/I1326498283"]},{"raw_affiliation_string":"[INRIA-Sophia Antipolis, France]","institution_ids":["https://openalex.org/I1326498283"]}]}],"institutions":[],"countries_distinct_count":1,"institutions_distinct_count":2,"corresponding_author_ids":["https://openalex.org/A5103287856"],"corresponding_institution_ids":["https://openalex.org/I1326498283"],"apc_list":null,"apc_paid":null,"fwci":7.9129,"has_fulltext":false,"cited_by_count":86,"citation_normalized_percentile":{"value":0.97622739,"is_in_top_1_percent":false,"is_in_top_10_percent":true},"cited_by_percentile_year":{"min":89,"max":100},"biblio":{"volume":null,"issue":null,"first_page":"1","last_page":"10"},"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":0.996999979019165,"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/T10715","display_name":"Distributed and Parallel Computing Systems","score":0.9959999918937683,"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/cuda","display_name":"CUDA","score":0.8542906641960144},{"id":"https://openalex.org/keywords/computer-science","display_name":"Computer science","score":0.7001000642776489},{"id":"https://openalex.org/keywords/upper-and-lower-bounds","display_name":"Upper and lower bounds","score":0.6338629722595215},{"id":"https://openalex.org/keywords/benchmarking","display_name":"Benchmarking","score":0.6070820689201355},{"id":"https://openalex.org/keywords/parallel-computing","display_name":"Parallel computing","score":0.5999699831008911},{"id":"https://openalex.org/keywords/fermi-gamma-ray-space-telescope","display_name":"Fermi Gamma-ray Space Telescope","score":0.5637715458869934},{"id":"https://openalex.org/keywords/set","display_name":"Set (abstract data type)","score":0.5257171988487244},{"id":"https://openalex.org/keywords/kepler","display_name":"Kepler","score":0.5103940367698669},{"id":"https://openalex.org/keywords/performance-improvement","display_name":"Performance improvement","score":0.4782211482524872},{"id":"https://openalex.org/keywords/code","display_name":"Code (set theory)","score":0.45259925723075867},{"id":"https://openalex.org/keywords/instruction-set","display_name":"Instruction set","score":0.44733864068984985},{"id":"https://openalex.org/keywords/computational-science","display_name":"Computational science","score":0.4326454997062683},{"id":"https://openalex.org/keywords/physics","display_name":"Physics","score":0.19124329090118408},{"id":"https://openalex.org/keywords/mathematics","display_name":"Mathematics","score":0.10938325524330139},{"id":"https://openalex.org/keywords/astrophysics","display_name":"Astrophysics","score":0.07264229655265808}],"concepts":[{"id":"https://openalex.org/C2778119891","wikidata":"https://www.wikidata.org/wiki/Q477690","display_name":"CUDA","level":2,"score":0.8542906641960144},{"id":"https://openalex.org/C41008148","wikidata":"https://www.wikidata.org/wiki/Q21198","display_name":"Computer science","level":0,"score":0.7001000642776489},{"id":"https://openalex.org/C77553402","wikidata":"https://www.wikidata.org/wiki/Q13222579","display_name":"Upper and lower bounds","level":2,"score":0.6338629722595215},{"id":"https://openalex.org/C86251818","wikidata":"https://www.wikidata.org/wiki/Q816754","display_name":"Benchmarking","level":2,"score":0.6070820689201355},{"id":"https://openalex.org/C173608175","wikidata":"https://www.wikidata.org/wiki/Q232661","display_name":"Parallel computing","level":1,"score":0.5999699831008911},{"id":"https://openalex.org/C186769553","wikidata":"https://www.wikidata.org/wiki/Q726648","display_name":"Fermi Gamma-ray Space Telescope","level":2,"score":0.5637715458869934},{"id":"https://openalex.org/C177264268","wikidata":"https://www.wikidata.org/wiki/Q1514741","display_name":"Set (abstract data type)","level":2,"score":0.5257171988487244},{"id":"https://openalex.org/C207963374","wikidata":"https://www.wikidata.org/wiki/Q47592","display_name":"Kepler","level":3,"score":0.5103940367698669},{"id":"https://openalex.org/C2778915421","wikidata":"https://www.wikidata.org/wiki/Q3643177","display_name":"Performance improvement","level":2,"score":0.4782211482524872},{"id":"https://openalex.org/C2776760102","wikidata":"https://www.wikidata.org/wiki/Q5139990","display_name":"Code (set theory)","level":3,"score":0.45259925723075867},{"id":"https://openalex.org/C202491316","wikidata":"https://www.wikidata.org/wiki/Q272683","display_name":"Instruction set","level":2,"score":0.44733864068984985},{"id":"https://openalex.org/C459310","wikidata":"https://www.wikidata.org/wiki/Q117801","display_name":"Computational science","level":1,"score":0.4326454997062683},{"id":"https://openalex.org/C121332964","wikidata":"https://www.wikidata.org/wiki/Q413","display_name":"Physics","level":0,"score":0.19124329090118408},{"id":"https://openalex.org/C33923547","wikidata":"https://www.wikidata.org/wiki/Q395","display_name":"Mathematics","level":0,"score":0.10938325524330139},{"id":"https://openalex.org/C44870925","wikidata":"https://www.wikidata.org/wiki/Q37547","display_name":"Astrophysics","level":1,"score":0.07264229655265808},{"id":"https://openalex.org/C162853370","wikidata":"https://www.wikidata.org/wiki/Q39809","display_name":"Marketing","level":1,"score":0.0},{"id":"https://openalex.org/C21547014","wikidata":"https://www.wikidata.org/wiki/Q1423657","display_name":"Operations management","level":1,"score":0.0},{"id":"https://openalex.org/C31972630","wikidata":"https://www.wikidata.org/wiki/Q844240","display_name":"Computer vision","level":1,"score":0.0},{"id":"https://openalex.org/C144133560","wikidata":"https://www.wikidata.org/wiki/Q4830453","display_name":"Business","level":0,"score":0.0},{"id":"https://openalex.org/C162324750","wikidata":"https://www.wikidata.org/wiki/Q8134","display_name":"Economics","level":0,"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/C134306372","wikidata":"https://www.wikidata.org/wiki/Q7754","display_name":"Mathematical analysis","level":1,"score":0.0},{"id":"https://openalex.org/C150846664","wikidata":"https://www.wikidata.org/wiki/Q7602306","display_name":"Stars","level":2,"score":0.0}],"mesh":[],"locations_count":2,"locations":[{"id":"doi:10.1109/cgo.2013.6494986","is_oa":false,"landing_page_url":"https://doi.org/10.1109/cgo.2013.6494986","pdf_url":null,"source":null,"license":null,"license_id":null,"version":"publishedVersion","is_accepted":true,"is_published":true,"raw_source_name":"Proceedings of the 2013 IEEE/ACM International Symposium on Code Generation and Optimization (CGO)","raw_type":"proceedings-article"},{"id":"pmh:oai:HAL:hal-00789958v1","is_oa":true,"landing_page_url":"https://inria.hal.science/hal-00789958","pdf_url":null,"source":{"id":"https://openalex.org/S4306402512","display_name":"HAL (Le Centre pour la Communication Scientifique Directe)","issn_l":null,"issn":null,"is_oa":false,"is_in_doaj":false,"is_core":false,"host_organization":"https://openalex.org/I1294671590","host_organization_name":"Centre National de la Recherche Scientifique","host_organization_lineage":["https://openalex.org/I1294671590"],"host_organization_lineage_names":[],"type":"repository"},"license":"other-oa","license_id":"https://openalex.org/licenses/other-oa","version":"submittedVersion","is_accepted":false,"is_published":false,"raw_source_name":"CGO '13 - 2013 International Symposium on Code Generation and Optimization, Feb 2013, Shenzhen, China","raw_type":"Conference papers"}],"best_oa_location":{"id":"pmh:oai:HAL:hal-00789958v1","is_oa":true,"landing_page_url":"https://inria.hal.science/hal-00789958","pdf_url":null,"source":{"id":"https://openalex.org/S4306402512","display_name":"HAL (Le Centre pour la Communication Scientifique Directe)","issn_l":null,"issn":null,"is_oa":false,"is_in_doaj":false,"is_core":false,"host_organization":"https://openalex.org/I1294671590","host_organization_name":"Centre National de la Recherche Scientifique","host_organization_lineage":["https://openalex.org/I1294671590"],"host_organization_lineage_names":[],"type":"repository"},"license":"other-oa","license_id":"https://openalex.org/licenses/other-oa","version":"submittedVersion","is_accepted":false,"is_published":false,"raw_source_name":"CGO '13 - 2013 International Symposium on Code Generation and Optimization, Feb 2013, Shenzhen, China","raw_type":"Conference papers"},"sustainable_development_goals":[],"awards":[],"funders":[],"has_content":{"pdf":false,"grobid_xml":false},"content_urls":null,"referenced_works_count":16,"referenced_works":["https://openalex.org/W1978642402","https://openalex.org/W1979527452","https://openalex.org/W1981896890","https://openalex.org/W2002555321","https://openalex.org/W2004711550","https://openalex.org/W2081040934","https://openalex.org/W2090593986","https://openalex.org/W2098220211","https://openalex.org/W2130336316","https://openalex.org/W2142769604","https://openalex.org/W2153492376","https://openalex.org/W2167334577","https://openalex.org/W2622728018","https://openalex.org/W4253068647","https://openalex.org/W6644889242","https://openalex.org/W6673666955"],"related_works":["https://openalex.org/W3062287","https://openalex.org/W2380390332","https://openalex.org/W2577502644","https://openalex.org/W2742145873","https://openalex.org/W4245975140","https://openalex.org/W2532592438","https://openalex.org/W1977763331","https://openalex.org/W2062253548","https://openalex.org/W4225414539","https://openalex.org/W1920227747"],"abstract_inverted_index":{"In":[0],"this":[1,121],"paper,":[2],"we":[3,26,170],"present":[4],"an":[5,24],"approach":[6,89],"to":[7,47,63,88,185],"estimate":[8],"GPU":[9,112],"applications'":[10],"performance":[11,31,98,108,138,152,160,169,178],"upper":[12,179],"bound":[13],"based":[14],"on":[15,38,109,115,129,148,163],"algorithm":[16],"analysis":[17,122],"and":[18,41,60,73,113,123],"assembly":[19,127,188],"code":[20],"level":[21],"benchmarking.":[22],"As":[23],"example,":[25],"analyze":[27],"the":[28,49,66,74,79,82,90,105,125,157,167,176,192],"potential":[29],"peak":[30,92,97,107],"of":[32,51,68,78,99,104,156,161,175],"SGEMM":[33,59,87,100,132,162],"(Single-precision":[34],"General":[35],"Matrix":[36],"Multiply)":[37],"Fermi":[39,69,111],"(GF110)":[40],"Kepler":[42,117],"(GK104)":[43],"GPUs.":[44],"We":[45,181],"try":[46],"answer":[48],"question":[50],"how":[52,184],"much":[53],"optimization":[54],"space":[55],"is":[56,101,153,172],"left":[57],"for":[58,86,145],"why.":[61],"According":[62],"our":[64,131],"analysis,":[65],"nature":[67],"(Kepler)":[70],"instruction":[71],"set":[72],"limited":[75],"issue":[76],"throughput":[77],"schedulers":[80],"are":[81],"main":[83],"limitation":[84],"factors":[85],"theoretical":[91,106],"performance.":[93],"The":[94,150],"estimated":[95,158,177],"upper-bound":[96,159],"around":[102,154,173],"82.5%":[103],"GTX580":[110],"57.6%":[114],"GTX680":[116],"GPU.":[118],"Guided":[119],"by":[120],"using":[124],"native":[126,187],"language,":[128],"average,":[130],"implementations":[133],"achieve":[134,171],"about":[135],"5%":[136],"better":[137],"than":[139],"CUBLAS":[140],"in":[141,191],"CUDA":[142,193],"4.1":[143],"SDK":[144],"large":[146],"matrices":[147],"GTX580.":[149,164],"achieved":[151],"90%":[155],"On":[165],"GTX680,":[166],"best":[168],"77.3%":[174],"bound.":[180],"also":[182],"describe":[183],"use":[186],"language":[189],"directly":[190],"runtime":[194],"source":[195],"code.":[196]},"counts_by_year":[{"year":2025,"cited_by_count":1},{"year":2024,"cited_by_count":4},{"year":2023,"cited_by_count":3},{"year":2022,"cited_by_count":3},{"year":2021,"cited_by_count":3},{"year":2020,"cited_by_count":10},{"year":2019,"cited_by_count":12},{"year":2018,"cited_by_count":14},{"year":2017,"cited_by_count":11},{"year":2016,"cited_by_count":12},{"year":2015,"cited_by_count":7},{"year":2014,"cited_by_count":5},{"year":2013,"cited_by_count":1}],"updated_date":"2026-04-04T16:13:02.066488","created_date":"2025-10-10T00:00:00"}