{"id":"https://openalex.org/W4386857147","doi":"https://doi.org/10.1145/3595178","title":"Cache Optimization and Performance Modeling of Batched, Small, and Rectangular Matrix Multiplication on Intel, AMD, and Fujitsu Processors","display_name":"Cache Optimization and Performance Modeling of Batched, Small, and Rectangular Matrix Multiplication on Intel, AMD, and Fujitsu Processors","publication_year":2023,"publication_date":"2023-09-19","ids":{"openalex":"https://openalex.org/W4386857147","doi":"https://doi.org/10.1145/3595178"},"language":"en","primary_location":{"id":"doi:10.1145/3595178","doi":"https://doi.org/10.1145/3595178","is_oa":true,"landing_page_url":"https://doi.org/10.1145/3595178","pdf_url":"https://dl.acm.org/doi/pdf/10.1145/3595178","source":{"id":"https://openalex.org/S163326656","display_name":"ACM Transactions on Mathematical Software","issn_l":"0098-3500","issn":["0098-3500","1557-7295"],"is_oa":false,"is_in_doaj":false,"is_core":true,"listed_in":["anvur-scientific","ccf-b","cwts-core","dongbi-b","jufo-3","norway-2","poland-200","russia-white-list-1"],"host_organization":"https://openalex.org/P4310319798","host_organization_name":"Association for Computing Machinery","host_organization_lineage":["https://openalex.org/P4310319798"],"host_organization_lineage_names":["Association for Computing Machinery"],"type":"journal"},"license":null,"license_id":null,"version":"publishedVersion","is_accepted":true,"is_published":true,"raw_source_name":"ACM Transactions on Mathematical Software","raw_type":"journal-article"},"volume":null,"type":"article","is_traditional":true,"indexed_in":["arxiv","crossref","datacite"],"open_access":{"is_oa":true,"oa_status":"bronze","oa_url":"https://dl.acm.org/doi/pdf/10.1145/3595178","any_repository_has_fulltext":true},"authorships":[{"author_position":"first","author":{"id":"https://openalex.org/A5065317117","display_name":"Sameer Deshmukh","orcid":"https://orcid.org/0000-0002-5615-1399","observed_orcids":["https://orcid.org/0000-0002-5615-1399"]},"institutions":[{"id":"https://openalex.org/I114531698","display_name":"Tokyo Institute of Technology","ror":"https://ror.org/0112mx960","country_code":"JP","type":"education","lineage":["https://openalex.org/I114531698","https://openalex.org/I4400009020"]}],"countries":["JP"],"is_corresponding":false,"raw_author_name":"Sameer Deshmukh","raw_affiliation_strings":["School of Computing, Tokyo Institute of Technology, AIST, Japan"],"raw_orcid":"https://orcid.org/0000-0002-5615-1399","affiliations":[{"raw_affiliation_string":"School of Computing, Tokyo Institute of Technology, AIST, Japan","institution_ids":["https://openalex.org/I114531698"]}]},{"author_position":"middle","author":{"id":"https://openalex.org/A5024747717","display_name":"Rio Yokota","orcid":"https://orcid.org/0000-0001-7573-7873","observed_orcids":["https://orcid.org/0000-0001-7573-7873"]},"institutions":[{"id":"https://openalex.org/I114531698","display_name":"Tokyo Institute of Technology","ror":"https://ror.org/0112mx960","country_code":"JP","type":"education","lineage":["https://openalex.org/I114531698","https://openalex.org/I4400009020"]}],"countries":["JP"],"is_corresponding":false,"raw_author_name":"Rio Yokota","raw_affiliation_strings":["School of Computing, Tokyo Institute of Technology, AIST, Japan"],"raw_orcid":"https://orcid.org/0000-0001-7573-7873","affiliations":[{"raw_affiliation_string":"School of Computing, Tokyo Institute of Technology, AIST, Japan","institution_ids":["https://openalex.org/I114531698"]}]},{"author_position":"last","author":{"id":"https://openalex.org/A5010055736","display_name":"George Bosilca","orcid":"https://orcid.org/0000-0003-2411-8495","observed_orcids":["https://orcid.org/0000-0003-2411-8495"]},"institutions":[{"id":"https://openalex.org/I75027704","display_name":"University of Tennessee at Knoxville","ror":"https://ror.org/020f3ap87","country_code":"US","type":"education","lineage":["https://openalex.org/I75027704"]}],"countries":["US"],"is_corresponding":false,"raw_author_name":"George Bosilca","raw_affiliation_strings":["Innovative Computing Laboratory, University of Tennessee at Knoxville, USA"],"raw_orcid":"https://orcid.org/0000-0003-2411-8495","affiliations":[{"raw_affiliation_string":"Innovative Computing Laboratory, University of Tennessee at Knoxville, USA","institution_ids":["https://openalex.org/I75027704"]}]}],"institutions":[],"countries_distinct_count":2,"institutions_distinct_count":2,"corresponding_author_ids":[],"corresponding_institution_ids":[],"apc_list":null,"apc_paid":null,"fwci":0.3413,"has_fulltext":true,"cited_by_count":2,"citation_normalized_percentile":{"value":0.525,"is_in_top_1_percent":false,"is_in_top_10_percent":false},"cited_by_percentile_year":{"min":89,"max":95},"biblio":{"volume":"49","issue":"3","first_page":"1","last_page":"29"},"is_retracted":false,"is_paratext":false,"is_xpac":false,"primary_topic":{"id":"https://openalex.org/T10054","display_name":"Parallel Computing and Optimization Techniques","score":0.7979000210762024,"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":0.7979000210762024,"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/T10792","display_name":"Matrix Theory and Algorithms","score":0.1965000033378601,"subfield":{"id":"https://openalex.org/subfields/1703","display_name":"Computational Theory and Mathematics"},"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/T11697","display_name":"Numerical Methods and Algorithms","score":0.0027000000700354576,"subfield":{"id":"https://openalex.org/subfields/1703","display_name":"Computational Theory and Mathematics"},"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/rectangular-matrix-multiplication","display_name":"rectangular matrix multiplication","score":0.8610000014305115},{"id":"https://openalex.org/keywords/cache-optimization","display_name":"cache optimization","score":0.8159999847412109},{"id":"https://openalex.org/keywords/performance-modeling","display_name":"performance modeling","score":0.7540000081062317},{"id":"https://openalex.org/keywords/low-rank-approximation","display_name":"low-rank approximation","score":0.7739999890327454},{"id":"https://openalex.org/keywords/fujitsu-a64fx","display_name":"Fujitsu A64FX","score":0.8799999952316284},{"id":"https://openalex.org/keywords/intel-xeon","display_name":"Intel Xeon","score":0.6510000228881836},{"id":"https://openalex.org/keywords/avx-512","display_name":"AVX-512","score":0.7699999809265137}],"concepts":[{"id":"https://openalex.org/C41008148","wikidata":"https://www.wikidata.org/wiki/Q21198","display_name":"Computer science","level":0,"score":0.8397505283355713},{"id":"https://openalex.org/C173608175","wikidata":"https://www.wikidata.org/wiki/Q232661","display_name":"Parallel computing","level":1,"score":0.7929824590682983},{"id":"https://openalex.org/C17349429","wikidata":"https://www.wikidata.org/wiki/Q1049914","display_name":"Matrix multiplication","level":3,"score":0.6527460813522339},{"id":"https://openalex.org/C150552126","wikidata":"https://www.wikidata.org/wiki/Q339387","display_name":"SIMD","level":2,"score":0.645162045955658},{"id":"https://openalex.org/C2780595030","wikidata":"https://www.wikidata.org/wiki/Q3860309","display_name":"Multiplication (music)","level":2,"score":0.5359616279602051},{"id":"https://openalex.org/C115537543","wikidata":"https://www.wikidata.org/wiki/Q165596","display_name":"Cache","level":2,"score":0.4989893436431885},{"id":"https://openalex.org/C2777210771","wikidata":"https://www.wikidata.org/wiki/Q4927124","display_name":"Block (permutation group theory)","level":2,"score":0.45102542638778687},{"id":"https://openalex.org/C123213974","wikidata":"https://www.wikidata.org/wiki/Q833089","display_name":"LU decomposition","level":4,"score":0.44028499722480774},{"id":"https://openalex.org/C157764524","wikidata":"https://www.wikidata.org/wiki/Q1383412","display_name":"Throughput","level":3,"score":0.42817223072052},{"id":"https://openalex.org/C139352143","wikidata":"https://www.wikidata.org/wiki/Q82571","display_name":"Linear algebra","level":2,"score":0.4195341467857361},{"id":"https://openalex.org/C45374587","wikidata":"https://www.wikidata.org/wiki/Q12525525","display_name":"Computation","level":2,"score":0.4144169092178345},{"id":"https://openalex.org/C111919701","wikidata":"https://www.wikidata.org/wiki/Q9135","display_name":"Operating system","level":1,"score":0.2499121129512787},{"id":"https://openalex.org/C11413529","wikidata":"https://www.wikidata.org/wiki/Q8366","display_name":"Algorithm","level":1,"score":0.20919585227966309},{"id":"https://openalex.org/C42355184","wikidata":"https://www.wikidata.org/wiki/Q1361088","display_name":"Matrix decomposition","level":3,"score":0.20647424459457397},{"id":"https://openalex.org/C33923547","wikidata":"https://www.wikidata.org/wiki/Q395","display_name":"Mathematics","level":0,"score":0.08434361219406128},{"id":"https://openalex.org/C121332964","wikidata":"https://www.wikidata.org/wiki/Q413","display_name":"Physics","level":0,"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/C84114770","wikidata":"https://www.wikidata.org/wiki/Q46344","display_name":"Quantum","level":2,"score":0.0},{"id":"https://openalex.org/C555944384","wikidata":"https://www.wikidata.org/wiki/Q249","display_name":"Wireless","level":2,"score":0.0},{"id":"https://openalex.org/C62520636","wikidata":"https://www.wikidata.org/wiki/Q944","display_name":"Quantum mechanics","level":1,"score":0.0},{"id":"https://openalex.org/C158693339","wikidata":"https://www.wikidata.org/wiki/Q190524","display_name":"Eigenvalues and eigenvectors","level":2,"score":0.0},{"id":"https://openalex.org/C114614502","wikidata":"https://www.wikidata.org/wiki/Q76592","display_name":"Combinatorics","level":1,"score":0.0}],"mesh":[],"locations_count":3,"locations":[{"id":"doi:10.1145/3595178","doi":"https://doi.org/10.1145/3595178","is_oa":true,"landing_page_url":"https://doi.org/10.1145/3595178","pdf_url":"https://dl.acm.org/doi/pdf/10.1145/3595178","source":{"id":"https://openalex.org/S163326656","display_name":"ACM Transactions on Mathematical Software","issn_l":"0098-3500","issn":["0098-3500","1557-7295"],"is_oa":false,"is_in_doaj":false,"is_core":true,"listed_in":["anvur-scientific","ccf-b","cwts-core","dongbi-b","jufo-3","norway-2","poland-200","russia-white-list-1"],"host_organization":"https://openalex.org/P4310319798","host_organization_name":"Association for Computing Machinery","host_organization_lineage":["https://openalex.org/P4310319798"],"host_organization_lineage_names":["Association for Computing Machinery"],"type":"journal"},"license":null,"license_id":null,"version":"publishedVersion","is_accepted":true,"is_published":true,"raw_source_name":"ACM Transactions on Mathematical Software","raw_type":"journal-article"},{"id":"pmh:oai:arXiv.org:2311.07602","doi":null,"is_oa":true,"landing_page_url":"http://arxiv.org/abs/2311.07602","pdf_url":"https://arxiv.org/pdf/2311.07602","source":{"id":"https://openalex.org/S4306400194","display_name":"arXiv (Cornell University)","issn_l":"2331-8422","issn":["2331-8422"],"is_oa":true,"is_in_doaj":false,"is_core":false,"listed_in":[],"host_organization":"https://openalex.org/I205783295","host_organization_name":"Cornell University","host_organization_lineage":["https://openalex.org/I205783295"],"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":"text"},{"id":"doi:10.48550/arxiv.2311.07602","doi":"https://doi.org/10.48550/arxiv.2311.07602","is_oa":true,"landing_page_url":"https://doi.org/10.48550/arxiv.2311.07602","pdf_url":null,"source":{"id":"https://openalex.org/S4306400194","display_name":"arXiv (Cornell University)","issn_l":"2331-8422","issn":["2331-8422"],"is_oa":true,"is_in_doaj":false,"is_core":false,"listed_in":[],"host_organization":"https://openalex.org/I205783295","host_organization_name":"Cornell University","host_organization_lineage":["https://openalex.org/I205783295"],"host_organization_lineage_names":[],"type":"repository"},"license":null,"license_id":null,"version":null,"is_accepted":false,"is_published":false,"raw_source_name":null,"raw_type":"Preprint"}],"best_oa_location":{"id":"doi:10.1145/3595178","doi":"https://doi.org/10.1145/3595178","is_oa":true,"landing_page_url":"https://doi.org/10.1145/3595178","pdf_url":"https://dl.acm.org/doi/pdf/10.1145/3595178","source":{"id":"https://openalex.org/S163326656","display_name":"ACM Transactions on Mathematical Software","issn_l":"0098-3500","issn":["0098-3500","1557-7295"],"is_oa":false,"is_in_doaj":false,"is_core":true,"listed_in":["anvur-scientific","ccf-b","cwts-core","dongbi-b","jufo-3","norway-2","poland-200","russia-white-list-1"],"host_organization":"https://openalex.org/P4310319798","host_organization_name":"Association for Computing Machinery","host_organization_lineage":["https://openalex.org/P4310319798"],"host_organization_lineage_names":["Association for Computing Machinery"],"type":"journal"},"license":null,"license_id":null,"version":"publishedVersion","is_accepted":true,"is_published":true,"raw_source_name":"ACM Transactions on Mathematical Software","raw_type":"journal-article"},"sustainable_development_goals":[],"study_designs":[],"awards":[{"id":"https://openalex.org/G5835976563","display_name":"Linear Solvers for Machine Learning Hardware","funder_award_id":"18H03248","funder_id":"https://openalex.org/F4320334764","funder_display_name":"Japan Society for the Promotion of Science"},{"id":"https://openalex.org/G5938650130","display_name":"Construction of numerical linear algebra based on lattice H-matrices and its high-performance implementation on modern architectures","funder_award_id":"21H03447","funder_id":"https://openalex.org/F4320334764","funder_display_name":"Japan Society for the Promotion of Science"},{"id":"https://openalex.org/G720495953","display_name":"Application of Unconventional Linear Algebra Techniques to Continuous Learning in Supergiant Neural Networks","funder_award_id":"20K20624","funder_id":"https://openalex.org/F4320334764","funder_display_name":"Japan Society for the Promotion of Science"}],"funders":[{"id":"https://openalex.org/F4320311508","display_name":"National Institute of Advanced Industrial Science and Technology","ror":"https://ror.org/01703db54"},{"id":"https://openalex.org/F4320322704","display_name":"Tokyo Institute of Technology","ror":"https://ror.org/0112mx960"},{"id":"https://openalex.org/F4320334764","display_name":"Japan Society for the Promotion of Science","ror":"https://ror.org/00hhkn466"},{"id":"https://openalex.org/F4320335125","display_name":"RIKEN","ror":"https://ror.org/01sjwvz98"}],"has_content":{"pdf":true,"grobid_xml":true},"content_urls":{"pdf":"https://content.openalex.org/works/W4386857147.pdf","grobid_xml":"https://content.openalex.org/works/W4386857147.grobid-xml"},"referenced_works_count":61,"referenced_works":["https://openalex.org/W201315547","https://openalex.org/W248911749","https://openalex.org/W1492914374","https://openalex.org/W1547507450","https://openalex.org/W1977230955","https://openalex.org/W2010747199","https://openalex.org/W2029891792","https://openalex.org/W2042390223","https://openalex.org/W2043275593","https://openalex.org/W2047656763","https://openalex.org/W2052196304","https://openalex.org/W2073061372","https://openalex.org/W2087440962","https://openalex.org/W2089212609","https://openalex.org/W2099059741","https://openalex.org/W2117756735","https://openalex.org/W2130289795","https://openalex.org/W2162630236","https://openalex.org/W2164705534","https://openalex.org/W2252007067","https://openalex.org/W2268122618","https://openalex.org/W2298381282","https://openalex.org/W2412152731","https://openalex.org/W2505064766","https://openalex.org/W2508318196","https://openalex.org/W2516525699","https://openalex.org/W2531348778","https://openalex.org/W2546792675","https://openalex.org/W2548341626","https://openalex.org/W2601355684","https://openalex.org/W2613264175","https://openalex.org/W2741554038","https://openalex.org/W2767612671","https://openalex.org/W2799990610","https://openalex.org/W2886857292","https://openalex.org/W2891241286","https://openalex.org/W2949953997","https://openalex.org/W2950007006","https://openalex.org/W2950444855","https://openalex.org/W2951894856","https://openalex.org/W2954478959","https://openalex.org/W2954744747","https://openalex.org/W2989919385","https://openalex.org/W2994914294","https://openalex.org/W2995911532","https://openalex.org/W2997776301","https://openalex.org/W3009472344","https://openalex.org/W3009819974","https://openalex.org/W3028132625","https://openalex.org/W3044796228","https://openalex.org/W3089579782","https://openalex.org/W3183017198","https://openalex.org/W3189349474","https://openalex.org/W3190673640","https://openalex.org/W4205602523","https://openalex.org/W4214826206","https://openalex.org/W4232958218","https://openalex.org/W4244254628","https://openalex.org/W4247963494","https://openalex.org/W4248316278","https://openalex.org/W4256588782"],"related_works":["https://openalex.org/W2741554038","https://openalex.org/W2531348778","https://openalex.org/W2896652512","https://openalex.org/W7106783669","https://openalex.org/W3086738126","https://openalex.org/W1512243664","https://openalex.org/W4414195034","https://openalex.org/W2793489890","https://openalex.org/W2084379367","https://openalex.org/W2601355684"],"abstract_inverted_index":{"Factorization":[0],"and":[1,6,29,90,115,141,147,171],"multiplication":[2],"of":[3,14,20,32,58,70,81,93,105,113,163],"dense":[4,73],"matrices":[5,108],"tensors":[7],"are":[8],"critical,":[9],"yet":[10],"extremely":[11],"expensive":[12],"pieces":[13],"the":[15,27,68,82,94,129,135,161],"scientific":[16],"toolbox.":[17],"Careful":[18],"use":[19],"low":[21],"rank":[22],"approximation":[23],"can":[24,97],"drastically":[25],"reduce":[26],"computation":[28,69],"memory":[30],"requirements":[31],"these":[33,103],"operations.":[34],"In":[35],"addition":[36],"to":[37,50,66,156],"a":[38,91,110],"lower":[39],"arithmetic":[40],"complexity,":[41],"such":[42],"methods":[43],"can,":[44],"by":[45],"their":[46],"structure,":[47],"be":[48],"designed":[49],"efficiently":[51],"exploit":[52],"modern":[53],"hardware":[54],"architectures.":[55],"The":[56],"majority":[57],"existing":[59],"work":[60],"relies":[61],"on":[62,122],"batched":[63,106],"BLAS":[64],"libraries":[65,166],"handle":[67],"many":[71],"small":[72],"matrices.":[74],"We":[75,118],"show":[76,148],"that":[77,149],"through":[78],"careful":[79],"analysis":[80],"cache":[83],"utilization,":[84],"register":[85],"accumulation":[86],"using":[87,125,132,139,145],"SIMD":[88],"registers":[89],"redesign":[92],"implementation,":[95],"one":[96],"achieve":[98],"significantly":[99],"higher":[100],"throughput":[101,162],"for":[102,167],"types":[104],"low-rank":[107],"across":[109],"large":[111],"range":[112],"block":[114],"batch":[116],"sizes.":[117,173],"test":[119],"our":[120,150],"algorithm":[121],"three":[123],"CPUs":[124],"diverse":[126],"ISAs":[127],"\u2013":[128],"Fujitsu":[130],"A64FX":[131],"ARM":[133],"SVE,":[134],"Intel":[136],"Xeon":[137],"6148":[138],"AVX-512,":[140],"AMD":[142],"EPYC":[143],"7502":[144],"AVX-2,":[146],"new":[151],"batching":[152],"methodology":[153],"is":[154],"able":[155],"obtain":[157],"more":[158],"than":[159],"twice":[160],"vendor":[164],"optimized":[165],"all":[168],"CPU":[169],"architectures":[170],"problem":[172]},"counts_by_year":[{"year":2026,"cited_by_count":1},{"year":2025,"cited_by_count":1}],"updated_date":"2026-11-10T11:39:43.021295","created_date":"2023-09-20T00:00:00"}
