{"id":"https://openalex.org/W7125931368","doi":"https://doi.org/10.1145/3774934.3786456","title":"Characterizing Matrix Multiplication Units across General Parallel Patterns in Scientific Computing","display_name":"Characterizing Matrix Multiplication Units across General Parallel Patterns in Scientific Computing","publication_year":2026,"publication_date":"2026-01-28","ids":{"openalex":"https://openalex.org/W7125931368","doi":"https://doi.org/10.1145/3774934.3786456"},"language":null,"primary_location":{"id":"doi:10.1145/3774934.3786456","is_oa":false,"landing_page_url":"https://doi.org/10.1145/3774934.3786456","pdf_url":null,"source":null,"license":null,"license_id":null,"version":"publishedVersion","is_accepted":true,"is_published":true,"raw_source_name":"Proceedings of the 31st ACM SIGPLAN Annual Symposium on Principles and Practice of Parallel Programming","raw_type":"proceedings-article"},"type":"conference-paper","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/A5040032112","display_name":"Yuechen Lu","orcid":"https://orcid.org/0009-0008-6387-8116"},"institutions":[{"id":"https://openalex.org/I204553293","display_name":"China University of Petroleum, Beijing","ror":"https://ror.org/041qf4r12","country_code":"CN","type":"education","lineage":["https://openalex.org/I204553293"]}],"countries":["CN"],"is_corresponding":false,"raw_author_name":"Yuechen Lu","raw_affiliation_strings":["China University of Petroleum-Beijing, Beijing, China"],"raw_orcid":"https://orcid.org/0009-0008-6387-8116","affiliations":[{"raw_affiliation_string":"China University of Petroleum-Beijing, Beijing, China","institution_ids":["https://openalex.org/I204553293"]}]},{"author_position":"middle","author":{"id":"https://openalex.org/A5124096239","display_name":"Hongwei Zeng","orcid":null},"institutions":[{"id":"https://openalex.org/I204553293","display_name":"China University of Petroleum, Beijing","ror":"https://ror.org/041qf4r12","country_code":"CN","type":"education","lineage":["https://openalex.org/I204553293"]}],"countries":["CN"],"is_corresponding":false,"raw_author_name":"Hongwei Zeng","raw_affiliation_strings":["China University of Petroleum-Beijing, Beijing, China"],"raw_orcid":"https://orcid.org/0009-0005-7504-4758","affiliations":[{"raw_affiliation_string":"China University of Petroleum-Beijing, Beijing, China","institution_ids":["https://openalex.org/I204553293"]}]},{"author_position":"middle","author":{"id":"https://openalex.org/A5044183257","display_name":"Marc Casas","orcid":"https://orcid.org/0000-0003-4564-2093"},"institutions":[{"id":"https://openalex.org/I2799803557","display_name":"Barcelona Supercomputing Center","ror":"https://ror.org/05sd8tv96","country_code":"ES","type":"facility","lineage":["https://openalex.org/I2799803557","https://openalex.org/I9617848"]},{"id":"https://openalex.org/I9617848","display_name":"Universitat Polit\u00e8cnica de Catalunya","ror":"https://ror.org/03mb6wj31","country_code":"ES","type":"education","lineage":["https://openalex.org/I9617848"]}],"countries":["ES"],"is_corresponding":false,"raw_author_name":"Marc Casas","raw_affiliation_strings":["Barcelona Supercomputing Center, Barcelona, Spain","Universitat Polit\u00e8cnica de Catalunya, Barcelona, Spain"],"raw_orcid":"https://orcid.org/0000-0003-4564-2093","affiliations":[{"raw_affiliation_string":"Barcelona Supercomputing Center, Barcelona, Spain","institution_ids":["https://openalex.org/I9617848","https://openalex.org/I2799803557"]},{"raw_affiliation_string":"Universitat Polit\u00e8cnica de Catalunya, Barcelona, Spain","institution_ids":["https://openalex.org/I9617848"]}]},{"author_position":"last","author":{"id":"https://openalex.org/A5124088796","display_name":"Weifeng Liu","orcid":null},"institutions":[{"id":"https://openalex.org/I204553293","display_name":"China University of Petroleum, Beijing","ror":"https://ror.org/041qf4r12","country_code":"CN","type":"education","lineage":["https://openalex.org/I204553293"]}],"countries":["CN"],"is_corresponding":false,"raw_author_name":"Weifeng Liu","raw_affiliation_strings":["China University of Petroleum-Beijing, Beijing, China"],"raw_orcid":"https://orcid.org/0000-0002-2150-5759","affiliations":[{"raw_affiliation_string":"China University of Petroleum-Beijing, Beijing, China","institution_ids":["https://openalex.org/I204553293"]}]}],"institutions":[],"countries_distinct_count":2,"institutions_distinct_count":3,"corresponding_author_ids":[],"corresponding_institution_ids":[],"apc_list":null,"apc_paid":null,"fwci":null,"has_fulltext":false,"cited_by_count":0,"citation_normalized_percentile":null,"cited_by_percentile_year":null,"biblio":{"volume":null,"issue":null,"first_page":"687","last_page":"701"},"is_retracted":false,"is_paratext":false,"is_xpac":false,"primary_topic":{"id":"https://openalex.org/T10054","display_name":"Parallel Computing and Optimization Techniques","score":0.7936000227928162,"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":0.7936000227928162,"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":0.03519999980926514,"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/T11181","display_name":"Advanced Data Storage Technologies","score":0.02879999950528145,"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/matrix-multiplication","display_name":"Matrix multiplication","score":0.6291999816894531},{"id":"https://openalex.org/keywords/suite","display_name":"Suite","score":0.626800000667572},{"id":"https://openalex.org/keywords/key","display_name":"Key (lock)","score":0.5942000150680542},{"id":"https://openalex.org/keywords/multiplication","display_name":"Multiplication (music)","score":0.5446000099182129},{"id":"https://openalex.org/keywords/benchmark","display_name":"Benchmark (surveying)","score":0.5080999732017517},{"id":"https://openalex.org/keywords/range","display_name":"Range (aeronautics)","score":0.45579999685287476},{"id":"https://openalex.org/keywords/supercomputer","display_name":"Supercomputer","score":0.4415999948978424}],"concepts":[{"id":"https://openalex.org/C41008148","wikidata":"https://www.wikidata.org/wiki/Q21198","display_name":"Computer science","level":0,"score":0.7663999795913696},{"id":"https://openalex.org/C17349429","wikidata":"https://www.wikidata.org/wiki/Q1049914","display_name":"Matrix multiplication","level":3,"score":0.6291999816894531},{"id":"https://openalex.org/C79581498","wikidata":"https://www.wikidata.org/wiki/Q1367530","display_name":"Suite","level":2,"score":0.626800000667572},{"id":"https://openalex.org/C26517878","wikidata":"https://www.wikidata.org/wiki/Q228039","display_name":"Key (lock)","level":2,"score":0.5942000150680542},{"id":"https://openalex.org/C2780595030","wikidata":"https://www.wikidata.org/wiki/Q3860309","display_name":"Multiplication (music)","level":2,"score":0.5446000099182129},{"id":"https://openalex.org/C173608175","wikidata":"https://www.wikidata.org/wiki/Q232661","display_name":"Parallel computing","level":1,"score":0.5167999863624573},{"id":"https://openalex.org/C185798385","wikidata":"https://www.wikidata.org/wiki/Q1161707","display_name":"Benchmark (surveying)","level":2,"score":0.5080999732017517},{"id":"https://openalex.org/C80444323","wikidata":"https://www.wikidata.org/wiki/Q2878974","display_name":"Theoretical computer science","level":1,"score":0.4652000069618225},{"id":"https://openalex.org/C204323151","wikidata":"https://www.wikidata.org/wiki/Q905424","display_name":"Range (aeronautics)","level":2,"score":0.45579999685287476},{"id":"https://openalex.org/C83283714","wikidata":"https://www.wikidata.org/wiki/Q121117","display_name":"Supercomputer","level":2,"score":0.4415999948978424},{"id":"https://openalex.org/C459310","wikidata":"https://www.wikidata.org/wiki/Q117801","display_name":"Computational science","level":1,"score":0.3779999911785126},{"id":"https://openalex.org/C23375383","wikidata":"https://www.wikidata.org/wiki/Q176499","display_name":"Unconventional computing","level":2,"score":0.3610999882221222},{"id":"https://openalex.org/C106487976","wikidata":"https://www.wikidata.org/wiki/Q685816","display_name":"Matrix (chemical analysis)","level":2,"score":0.3531000018119812},{"id":"https://openalex.org/C106515295","wikidata":"https://www.wikidata.org/wiki/Q26806595","display_name":"Parallel processing","level":2,"score":0.33180001378059387},{"id":"https://openalex.org/C94124525","wikidata":"https://www.wikidata.org/wiki/Q912550","display_name":"Categorization","level":2,"score":0.2743000090122223},{"id":"https://openalex.org/C2984917352","wikidata":"https://www.wikidata.org/wiki/Q12772819","display_name":"Scientific discovery","level":2,"score":0.26499998569488525},{"id":"https://openalex.org/C162319229","wikidata":"https://www.wikidata.org/wiki/Q175263","display_name":"Data structure","level":2,"score":0.25929999351501465}],"mesh":[],"locations_count":1,"locations":[{"id":"doi:10.1145/3774934.3786456","is_oa":false,"landing_page_url":"https://doi.org/10.1145/3774934.3786456","pdf_url":null,"source":null,"license":null,"license_id":null,"version":"publishedVersion","is_accepted":true,"is_published":true,"raw_source_name":"Proceedings of the 31st ACM SIGPLAN Annual Symposium on Principles and Practice of Parallel Programming","raw_type":"proceedings-article"}],"best_oa_location":null,"sustainable_development_goals":[{"id":"https://metadata.un.org/sdg/7","score":0.6079866290092468,"display_name":"Affordable and clean energy"}],"awards":[],"funders":[],"has_content":{"grobid_xml":false,"pdf":false},"content_urls":null,"referenced_works_count":77,"referenced_works":["https://openalex.org/W1983545421","https://openalex.org/W1996875888","https://openalex.org/W2009654791","https://openalex.org/W2035080386","https://openalex.org/W2039417226","https://openalex.org/W2080592089","https://openalex.org/W2128853364","https://openalex.org/W2143011780","https://openalex.org/W2149234156","https://openalex.org/W2151587416","https://openalex.org/W2152839228","https://openalex.org/W2162283062","https://openalex.org/W2169631286","https://openalex.org/W2429898963","https://openalex.org/W2461193710","https://openalex.org/W2606722458","https://openalex.org/W2735793669","https://openalex.org/W2762661956","https://openalex.org/W2766111638","https://openalex.org/W2791673912","https://openalex.org/W2901549770","https://openalex.org/W2951135776","https://openalex.org/W2982236015","https://openalex.org/W3004913014","https://openalex.org/W3016542674","https://openalex.org/W3091804697","https://openalex.org/W3093933627","https://openalex.org/W3117765988","https://openalex.org/W3123639931","https://openalex.org/W3130554079","https://openalex.org/W3139307480","https://openalex.org/W3140386354","https://openalex.org/W3174843332","https://openalex.org/W3175644086","https://openalex.org/W3177007352","https://openalex.org/W3187908937","https://openalex.org/W3205226039","https://openalex.org/W3205883294","https://openalex.org/W3207730444","https://openalex.org/W4205744550","https://openalex.org/W4220980777","https://openalex.org/W4221106024","https://openalex.org/W4282975155","https://openalex.org/W4283029140","https://openalex.org/W4285327580","https://openalex.org/W4293023340","https://openalex.org/W4297097347","https://openalex.org/W4297097382","https://openalex.org/W4297097464","https://openalex.org/W4309672181","https://openalex.org/W4318541538","https://openalex.org/W4321446155","https://openalex.org/W4321636620","https://openalex.org/W4321636621","https://openalex.org/W4327911434","https://openalex.org/W4360831840","https://openalex.org/W4388661983","https://openalex.org/W4389166750","https://openalex.org/W4391827239","https://openalex.org/W4391987273","https://openalex.org/W4392884779","https://openalex.org/W4400409888","https://openalex.org/W4400681470","https://openalex.org/W4401408710","https://openalex.org/W4401408779","https://openalex.org/W4402475781","https://openalex.org/W4405755887","https://openalex.org/W4405756173","https://openalex.org/W4405756205","https://openalex.org/W4405778707","https://openalex.org/W4408029437","https://openalex.org/W4408029507","https://openalex.org/W4408029514","https://openalex.org/W4408049835","https://openalex.org/W4408049841","https://openalex.org/W4416199163","https://openalex.org/W7106821440"],"related_works":[],"abstract_inverted_index":{"Matrix":[0],"multiplication":[1],"units":[2],"(MMUs)":[3],"in":[4,21,30,96,117],"modern":[5],"parallel":[6,78],"processors":[7],"enable":[8],"efficient":[9],"execution":[10],"of":[11,46,76,115],"tiled":[12],"matrix":[13],"multiplications":[14],"at":[15],"varying":[16],"precisions.":[17],"While":[18],"their":[19,28],"effectiveness":[20],"AI":[22],"workloads":[23],"has":[24],"been":[25],"well":[26],"demonstrated,":[27],"utility":[29],"scientific":[31,47,97,119],"computing":[32,48],"lacks":[33],"systematic":[34],"analysis.":[35],"In":[36],"this":[37,63],"work,":[38],"we":[39,65,105],"characterize":[40],"MMUs":[41,116],"across":[42],"a":[43,68],"broad":[44],"range":[45],"patterns":[49,85],"by":[50],"evaluating":[51],"performance,":[52],"power":[53],"consumption,":[54],"numerical":[55],"precision,":[56],"and":[57,89,113,127],"memory":[58],"access":[59],"behavior.":[60],"To":[61],"support":[62],"analysis,":[64],"develop":[66],"Cubie,":[67],"comprehensive":[69],"benchmark":[70],"suite":[71],"comprising":[72],"ten":[73],"MMU-optimized":[74],"kernels":[75],"key":[77,108],"patterns.":[79],"We":[80],"also":[81],"categorize":[82],"MMU":[83,92],"utilization":[84],"into":[86],"four":[87],"quadrants":[88],"identify":[90],"the":[91,111],"limitations":[93],"that":[94],"arise":[95],"computing.":[98],"Through":[99],"detailed":[100],"comparisons":[101],"with":[102],"vector":[103],"units,":[104],"provide":[106],"nine":[107],"observations":[109],"on":[110],"behavior":[112],"implications":[114],"general":[118],"workloads,":[120],"offering":[121],"valuable":[122],"insights":[123],"for":[124],"architecture,":[125],"algorithm,":[126],"application":[128],"researchers.":[129]},"counts_by_year":[],"updated_date":"2026-07-14T23:27:15.235271","created_date":"2026-01-29T00:00:00"}