{"id":"https://openalex.org/W4412605596","doi":"https://doi.org/10.1109/ipdps64566.2025.00087","title":"NM-SpMM: Accelerating Matrix Multiplication Using N:M Sparsity with GPGPU","display_name":"NM-SpMM: Accelerating Matrix Multiplication Using N:M Sparsity with GPGPU","publication_year":2025,"publication_date":"2025-06-03","ids":{"openalex":"https://openalex.org/W4412605596","doi":"https://doi.org/10.1109/ipdps64566.2025.00087"},"language":"en","primary_location":{"id":"doi:10.1109/ipdps64566.2025.00087","is_oa":false,"landing_page_url":"https://doi.org/10.1109/ipdps64566.2025.00087","pdf_url":null,"source":null,"license":null,"license_id":null,"version":"publishedVersion","is_accepted":true,"is_published":true,"raw_source_name":"2025 IEEE International Parallel and Distributed Processing Symposium (IPDPS)","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/A5101810188","display_name":"Cong Ma","orcid":"https://orcid.org/0000-0002-9787-6273"},"institutions":[{"id":"https://openalex.org/I4210145761","display_name":"Shenzhen Institutes of Advanced Technology","ror":"https://ror.org/04gh4er46","country_code":"CN","type":"facility","lineage":["https://openalex.org/I19820366","https://openalex.org/I4210145761"]}],"countries":["CN"],"is_corresponding":false,"raw_author_name":"Cong Ma","raw_affiliation_strings":["Shenzhen Institutes of Advanced Technology,Chinese Academy of Sciences,China"],"raw_orcid":null,"affiliations":[{"raw_affiliation_string":"Shenzhen Institutes of Advanced Technology,Chinese Academy of Sciences,China","institution_ids":["https://openalex.org/I4210145761"]}]},{"author_position":"middle","author":{"id":"https://openalex.org/A5051506217","display_name":"Du Wu","orcid":"https://orcid.org/0000-0002-4002-0837"},"institutions":[{"id":"https://openalex.org/I4210129730","display_name":"RIKEN Center for Computational Science","ror":"https://ror.org/03r519674","country_code":"JP","type":"facility","lineage":["https://openalex.org/I4210110652","https://openalex.org/I4210129730"]}],"countries":["JP"],"is_corresponding":false,"raw_author_name":"Du Wu","raw_affiliation_strings":["RIKEN Center for Computational Science,Japan"],"raw_orcid":null,"affiliations":[{"raw_affiliation_string":"RIKEN Center for Computational Science,Japan","institution_ids":["https://openalex.org/I4210129730"]}]},{"author_position":"middle","author":{"id":"https://openalex.org/A5103173962","display_name":"Z. Y. Deng","orcid":"https://orcid.org/0000-0003-0440-3870"},"institutions":[{"id":"https://openalex.org/I4210145761","display_name":"Shenzhen Institutes of Advanced Technology","ror":"https://ror.org/04gh4er46","country_code":"CN","type":"facility","lineage":["https://openalex.org/I19820366","https://openalex.org/I4210145761"]}],"countries":["CN"],"is_corresponding":false,"raw_author_name":"Zhelang Deng","raw_affiliation_strings":["Shenzhen Institutes of Advanced Technology,Chinese Academy of Sciences,China"],"raw_orcid":null,"affiliations":[{"raw_affiliation_string":"Shenzhen Institutes of Advanced Technology,Chinese Academy of Sciences,China","institution_ids":["https://openalex.org/I4210145761"]}]},{"author_position":"middle","author":{"id":"https://openalex.org/A5101531055","display_name":"Jiang Chen","orcid":"https://orcid.org/0009-0000-6888-6643"},"institutions":[{"id":"https://openalex.org/I4210145761","display_name":"Shenzhen Institutes of Advanced Technology","ror":"https://ror.org/04gh4er46","country_code":"CN","type":"facility","lineage":["https://openalex.org/I19820366","https://openalex.org/I4210145761"]}],"countries":["CN"],"is_corresponding":false,"raw_author_name":"Jiang Chen","raw_affiliation_strings":["Shenzhen Institutes of Advanced Technology,Chinese Academy of Sciences,China"],"raw_orcid":null,"affiliations":[{"raw_affiliation_string":"Shenzhen Institutes of Advanced Technology,Chinese Academy of Sciences,China","institution_ids":["https://openalex.org/I4210145761"]}]},{"author_position":"middle","author":{"id":"https://openalex.org/A5113312900","display_name":"Xiaowen Huang","orcid":"https://orcid.org/0000-0002-1918-1177"},"institutions":[{"id":"https://openalex.org/I180726961","display_name":"Shenzhen University","ror":"https://ror.org/01vy4gh70","country_code":"CN","type":"education","lineage":["https://openalex.org/I180726961"]}],"countries":["CN"],"is_corresponding":false,"raw_author_name":"Xiaowen Huang","raw_affiliation_strings":["Shenzhen University,China"],"raw_orcid":null,"affiliations":[{"raw_affiliation_string":"Shenzhen University,China","institution_ids":["https://openalex.org/I180726961"]}]},{"author_position":"middle","author":{"id":"https://openalex.org/A5056275783","display_name":"Jintao Meng","orcid":"https://orcid.org/0000-0002-6208-4102"},"institutions":[{"id":"https://openalex.org/I4210145761","display_name":"Shenzhen Institutes of Advanced Technology","ror":"https://ror.org/04gh4er46","country_code":"CN","type":"facility","lineage":["https://openalex.org/I19820366","https://openalex.org/I4210145761"]}],"countries":["CN"],"is_corresponding":false,"raw_author_name":"Jintao Meng","raw_affiliation_strings":["Shenzhen Institutes of Advanced Technology,Chinese Academy of Sciences,China"],"raw_orcid":null,"affiliations":[{"raw_affiliation_string":"Shenzhen Institutes of Advanced Technology,Chinese Academy of Sciences,China","institution_ids":["https://openalex.org/I4210145761"]}]},{"author_position":"middle","author":{"id":"https://openalex.org/A5101131384","display_name":"Wenxi Zhu","orcid":"https://orcid.org/0009-0001-4933-4672"},"institutions":[{"id":"https://openalex.org/I2250653659","display_name":"Tencent (China)","ror":"https://ror.org/00hhjss72","country_code":"CN","type":"company","lineage":["https://openalex.org/I2250653659"]}],"countries":["CN"],"is_corresponding":false,"raw_author_name":"Wenxi Zhu","raw_affiliation_strings":["Tencent AI Lab,China"],"raw_orcid":null,"affiliations":[{"raw_affiliation_string":"Tencent AI Lab,China","institution_ids":["https://openalex.org/I2250653659"]}]},{"author_position":"middle","author":{"id":"https://openalex.org/A5102768672","display_name":"Bingqiang Wang","orcid":"https://orcid.org/0009-0004-4964-3258"},"institutions":[{"id":"https://openalex.org/I4210136793","display_name":"Peng Cheng Laboratory","ror":"https://ror.org/03qdqbt06","country_code":"CN","type":"facility","lineage":["https://openalex.org/I4210136793"]}],"countries":["CN"],"is_corresponding":false,"raw_author_name":"Bingqiang Wang","raw_affiliation_strings":["Peng Cheng Laboratory,China"],"raw_orcid":null,"affiliations":[{"raw_affiliation_string":"Peng Cheng Laboratory,China","institution_ids":["https://openalex.org/I4210136793"]}]},{"author_position":"middle","author":{"id":"https://openalex.org/A5102803813","display_name":"Amelie Chi Zhou","orcid":"https://orcid.org/0000-0002-9773-9332"},"institutions":[{"id":"https://openalex.org/I141568987","display_name":"Hong Kong Baptist University","ror":"https://ror.org/0145fw131","country_code":"HK","type":"education","lineage":["https://openalex.org/I141568987"]}],"countries":["HK"],"is_corresponding":false,"raw_author_name":"Amelie Chi Zhou","raw_affiliation_strings":["Hong Kong Baptist University,China"],"raw_orcid":null,"affiliations":[{"raw_affiliation_string":"Hong Kong Baptist University,China","institution_ids":["https://openalex.org/I141568987"]}]},{"author_position":"middle","author":{"id":"https://openalex.org/A5100338479","display_name":"Peng Chen","orcid":"https://orcid.org/0000-0003-1244-3151"},"institutions":[{"id":"https://openalex.org/I4210129730","display_name":"RIKEN Center for Computational Science","ror":"https://ror.org/03r519674","country_code":"JP","type":"facility","lineage":["https://openalex.org/I4210110652","https://openalex.org/I4210129730"]}],"countries":["JP"],"is_corresponding":false,"raw_author_name":"Peng Chen","raw_affiliation_strings":["RIKEN Center for Computational Science,Japan"],"raw_orcid":null,"affiliations":[{"raw_affiliation_string":"RIKEN Center for Computational Science,Japan","institution_ids":["https://openalex.org/I4210129730"]}]},{"author_position":"middle","author":{"id":"https://openalex.org/A5083508620","display_name":"Minwen Deng","orcid":"https://orcid.org/0009-0005-8122-3023"},"institutions":[{"id":"https://openalex.org/I2250653659","display_name":"Tencent (China)","ror":"https://ror.org/00hhjss72","country_code":"CN","type":"company","lineage":["https://openalex.org/I2250653659"]}],"countries":["CN"],"is_corresponding":false,"raw_author_name":"Minwen Deng","raw_affiliation_strings":["Tencent AI Lab,China"],"raw_orcid":null,"affiliations":[{"raw_affiliation_string":"Tencent AI Lab,China","institution_ids":["https://openalex.org/I2250653659"]}]},{"author_position":"middle","author":{"id":"https://openalex.org/A5037823825","display_name":"Yanjie Wei","orcid":"https://orcid.org/0000-0002-4791-7540"},"institutions":[{"id":"https://openalex.org/I4210145761","display_name":"Shenzhen Institutes of Advanced Technology","ror":"https://ror.org/04gh4er46","country_code":"CN","type":"facility","lineage":["https://openalex.org/I19820366","https://openalex.org/I4210145761"]}],"countries":["CN"],"is_corresponding":false,"raw_author_name":"Yanjie Wei","raw_affiliation_strings":["Shenzhen Institutes of Advanced Technology,Chinese Academy of Sciences,China"],"raw_orcid":null,"affiliations":[{"raw_affiliation_string":"Shenzhen Institutes of Advanced Technology,Chinese Academy of Sciences,China","institution_ids":["https://openalex.org/I4210145761"]}]},{"author_position":"middle","author":{"id":"https://openalex.org/A5101952645","display_name":"Shengzhong Feng","orcid":"https://orcid.org/0000-0002-6225-2815"},"institutions":[{"id":"https://openalex.org/I4210142539","display_name":"Guangdong Institute of Intelligent Manufacturing","ror":"https://ror.org/049jpjz09","country_code":"CN","type":"facility","lineage":["https://openalex.org/I4210142539"]}],"countries":["CN"],"is_corresponding":false,"raw_author_name":"Shengzhong Feng","raw_affiliation_strings":["Guangdong Institute of Intelligence Science and Technolog,China"],"raw_orcid":null,"affiliations":[{"raw_affiliation_string":"Guangdong Institute of Intelligence Science and Technolog,China","institution_ids":["https://openalex.org/I4210142539"]}]},{"author_position":"last","author":{"id":"https://openalex.org/A5068003384","display_name":"Yi Pan","orcid":"https://orcid.org/0000-0003-1978-475X"},"institutions":[{"id":"https://openalex.org/I4210145761","display_name":"Shenzhen Institutes of Advanced Technology","ror":"https://ror.org/04gh4er46","country_code":"CN","type":"facility","lineage":["https://openalex.org/I19820366","https://openalex.org/I4210145761"]}],"countries":["CN"],"is_corresponding":false,"raw_author_name":"Yi Pan","raw_affiliation_strings":["Shenzhen Institutes of Advanced Technology,Chinese Academy of Sciences,China"],"raw_orcid":null,"affiliations":[{"raw_affiliation_string":"Shenzhen Institutes of Advanced Technology,Chinese Academy of Sciences,China","institution_ids":["https://openalex.org/I4210145761"]}]}],"institutions":[],"countries_distinct_count":3,"institutions_distinct_count":7,"corresponding_author_ids":[],"corresponding_institution_ids":[],"apc_list":null,"apc_paid":null,"fwci":1.5151,"has_fulltext":false,"cited_by_count":1,"citation_normalized_percentile":{"value":0.78835453,"is_in_top_1_percent":false,"is_in_top_10_percent":false},"cited_by_percentile_year":{"min":94,"max":97},"biblio":{"volume":null,"issue":null,"first_page":"926","last_page":"937"},"is_retracted":false,"is_paratext":false,"is_xpac":false,"primary_topic":{"id":"https://openalex.org/T10054","display_name":"Parallel Computing and Optimization Techniques","score":0.9965000152587891,"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":0.9965000152587891,"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/T11612","display_name":"Stochastic Gradient Optimization Techniques","score":0.9573000073432922,"subfield":{"id":"https://openalex.org/subfields/1702","display_name":"Artificial Intelligence"},"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":0.9557999968528748,"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/general-purpose-computing-on-graphics-processing-units","display_name":"General-purpose computing on graphics processing units","score":0.746333122253418},{"id":"https://openalex.org/keywords/computer-science","display_name":"Computer science","score":0.6664144992828369},{"id":"https://openalex.org/keywords/matrix-multiplication","display_name":"Matrix multiplication","score":0.6121144890785217},{"id":"https://openalex.org/keywords/parallel-computing","display_name":"Parallel computing","score":0.606752872467041},{"id":"https://openalex.org/keywords/multiplication","display_name":"Multiplication (music)","score":0.5798251628875732},{"id":"https://openalex.org/keywords/computational-science","display_name":"Computational science","score":0.38490402698516846},{"id":"https://openalex.org/keywords/computer-graphics","display_name":"Computer graphics (images)","score":0.17787933349609375},{"id":"https://openalex.org/keywords/mathematics","display_name":"Mathematics","score":0.16564780473709106},{"id":"https://openalex.org/keywords/physics","display_name":"Physics","score":0.1354939043521881},{"id":"https://openalex.org/keywords/graphics","display_name":"Graphics","score":0.124762624502182}],"concepts":[{"id":"https://openalex.org/C50630238","wikidata":"https://www.wikidata.org/wiki/Q971505","display_name":"General-purpose computing on graphics processing units","level":3,"score":0.746333122253418},{"id":"https://openalex.org/C41008148","wikidata":"https://www.wikidata.org/wiki/Q21198","display_name":"Computer science","level":0,"score":0.6664144992828369},{"id":"https://openalex.org/C17349429","wikidata":"https://www.wikidata.org/wiki/Q1049914","display_name":"Matrix multiplication","level":3,"score":0.6121144890785217},{"id":"https://openalex.org/C173608175","wikidata":"https://www.wikidata.org/wiki/Q232661","display_name":"Parallel computing","level":1,"score":0.606752872467041},{"id":"https://openalex.org/C2780595030","wikidata":"https://www.wikidata.org/wiki/Q3860309","display_name":"Multiplication (music)","level":2,"score":0.5798251628875732},{"id":"https://openalex.org/C459310","wikidata":"https://www.wikidata.org/wiki/Q117801","display_name":"Computational science","level":1,"score":0.38490402698516846},{"id":"https://openalex.org/C121684516","wikidata":"https://www.wikidata.org/wiki/Q7600677","display_name":"Computer graphics (images)","level":1,"score":0.17787933349609375},{"id":"https://openalex.org/C33923547","wikidata":"https://www.wikidata.org/wiki/Q395","display_name":"Mathematics","level":0,"score":0.16564780473709106},{"id":"https://openalex.org/C121332964","wikidata":"https://www.wikidata.org/wiki/Q413","display_name":"Physics","level":0,"score":0.1354939043521881},{"id":"https://openalex.org/C21442007","wikidata":"https://www.wikidata.org/wiki/Q1027879","display_name":"Graphics","level":2,"score":0.124762624502182},{"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/C84114770","wikidata":"https://www.wikidata.org/wiki/Q46344","display_name":"Quantum","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":1,"locations":[{"id":"doi:10.1109/ipdps64566.2025.00087","is_oa":false,"landing_page_url":"https://doi.org/10.1109/ipdps64566.2025.00087","pdf_url":null,"source":null,"license":null,"license_id":null,"version":"publishedVersion","is_accepted":true,"is_published":true,"raw_source_name":"2025 IEEE International Parallel and Distributed Processing Symposium (IPDPS)","raw_type":"proceedings-article"}],"best_oa_location":null,"sustainable_development_goals":[],"awards":[],"funders":[],"has_content":{"pdf":false,"grobid_xml":false},"content_urls":null,"referenced_works_count":28,"referenced_works":["https://openalex.org/W1761022139","https://openalex.org/W2156150815","https://openalex.org/W2566079294","https://openalex.org/W2899244816","https://openalex.org/W2915106038","https://openalex.org/W3034733718","https://openalex.org/W3130660608","https://openalex.org/W3132616766","https://openalex.org/W3208099998","https://openalex.org/W4293023340","https://openalex.org/W4312348045","https://openalex.org/W4320036900","https://openalex.org/W4320063610","https://openalex.org/W4386902741","https://openalex.org/W4388667495","https://openalex.org/W4402727041","https://openalex.org/W6683826617","https://openalex.org/W6739901393","https://openalex.org/W6755207826","https://openalex.org/W6778883912","https://openalex.org/W6790503700","https://openalex.org/W6790943123","https://openalex.org/W6803607599","https://openalex.org/W6803712692","https://openalex.org/W6848451824","https://openalex.org/W6849185048","https://openalex.org/W6849672336","https://openalex.org/W6852962002"],"related_works":["https://openalex.org/W3099313426","https://openalex.org/W4287593139","https://openalex.org/W752783541","https://openalex.org/W1506547947","https://openalex.org/W2995605830","https://openalex.org/W4206811032","https://openalex.org/W2086123442","https://openalex.org/W1925544630","https://openalex.org/W2596457687","https://openalex.org/W3212757063"],"abstract_inverted_index":{"Deep":[0],"learning":[1],"demonstrates":[2],"effectiveness":[3],"across":[4,129],"a":[5,73,98,103],"wide":[6],"range":[7],"of":[8,16],"tasks.":[9],"However,":[10],"the":[11,160,166],"dense":[12,44,155],"and":[13,56,64,114,148,177],"over-parameterized":[14],"nature":[15],"these":[17,69],"models":[18],"results":[19,134],"in":[20,168],"significant":[21],"resource":[22],"consumption":[23],"during":[24],"deployment.":[25],"In":[26],"response":[27],"to":[28,106,124,150,171],"this":[29],"issue,":[30],"weight":[31],"pruning,":[32],"particularly":[33],"through$N:":[34],"M$sparsity":[35,49,89],"matrix":[36],"multiplication,":[37],"offers":[38],"an":[39,51,86],"efficient":[40,87],"solution":[41],"by":[42],"transforming":[43],"operations":[45],"into":[46],"semisparse":[47],"ones.$N:":[48],"provides":[50],"option":[52],"for":[53,145],"balancing":[54],"performance":[55,76,94,128],"model":[57,78],"accuracy,":[58],"but":[59],"introduces":[60],"more":[61],"complex":[62],"programming":[63],"optimization":[65,105,113],"challenges.":[66],"To":[67],"address":[68],"issues,":[70],"we":[71],"design":[72,116],"systematic":[74],"top-down":[75],"analysis":[77],"for$N:":[79],"M$sparsity.":[80],"Meanwhile,":[81],"NM-SpMM":[82,96,137,173],"is":[83,138,174],"proposed":[84],"as":[85,102,119],"general$N:":[88,146],"implementation.":[90],"Based":[91],"on":[92],"our":[93],"analysis,":[95],"employs":[97],"hierarchical":[99],"blocking":[100],"mechanism":[101],"general":[104],"enhance":[107],"data":[108],"locality,":[109],"while":[110],"memory":[111],"access":[112],"pipeline":[115],"are":[117],"introduced":[118],"sparsity-aware":[120],"optimization,":[121],"allowing":[122],"it":[123],"achieve":[125],"close-to-theoretical":[126],"peak":[127],"different":[130],"sparsity":[131],"levels.":[132],"Experimental":[133],"show":[135],"that":[136],"2.1x":[139],"faster":[140,152],"than":[141,153],"nmSPARSE":[142],"(the":[143],"state-of-the-art":[144],"M$sparsity)":[147],"1.4\u00d7":[149],"6.3\u00d7":[151],"cuBLAS's":[154],"GEMM":[156],"operations,":[157],"closely":[158],"approaching":[159],"theoretical":[161],"maximum":[162],"speedup":[163],"resulting":[164],"from":[165],"reduction":[167],"computation":[169],"due":[170],"sparsity.":[172],"open":[175],"source":[176],"publicly":[178],"available":[179],"at":[180],"https://github.com/M-H482/NM-SpMM.":[181]},"counts_by_year":[{"year":2026,"cited_by_count":1}],"updated_date":"2026-07-29T14:22:42.915294","created_date":"2025-10-10T00:00:00"}