{"id":"https://openalex.org/W4409248726","doi":"https://doi.org/10.1109/hpca61900.2025.00075","title":"TB-STC: Transposable Block-wise N:M Structured Sparse Tensor Core","display_name":"TB-STC: Transposable Block-wise N:M Structured Sparse Tensor Core","publication_year":2025,"publication_date":"2025-03-01","ids":{"openalex":"https://openalex.org/W4409248726","doi":"https://doi.org/10.1109/hpca61900.2025.00075"},"language":"en","primary_location":{"id":"doi:10.1109/hpca61900.2025.00075","is_oa":false,"landing_page_url":"https://doi.org/10.1109/hpca61900.2025.00075","pdf_url":null,"source":null,"license":null,"license_id":null,"version":"publishedVersion","is_accepted":true,"is_published":true,"raw_source_name":"2025 IEEE International Symposium on High Performance Computer Architecture (HPCA)","raw_type":"proceedings-article"},"type":"article","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/A5100361691","display_name":"Jun Liu","orcid":"https://orcid.org/0000-0001-8300-8646"},"institutions":[{"id":"https://openalex.org/I183067930","display_name":"Shanghai Jiao Tong University","ror":"https://ror.org/0220qvk04","country_code":"CN","type":"education","lineage":["https://openalex.org/I183067930"]}],"countries":["CN"],"is_corresponding":true,"raw_author_name":"Jun Liu","raw_affiliation_strings":["Shanghai Jiao Tong University,China"],"affiliations":[{"raw_affiliation_string":"Shanghai Jiao Tong University,China","institution_ids":["https://openalex.org/I183067930"]}]},{"author_position":"middle","author":{"id":"https://openalex.org/A5026831784","display_name":"Shulin Zeng","orcid":"https://orcid.org/0000-0002-1030-3748"},"institutions":[{"id":"https://openalex.org/I99065089","display_name":"Tsinghua University","ror":"https://ror.org/03cve4549","country_code":"CN","type":"education","lineage":["https://openalex.org/I99065089"]}],"countries":["CN"],"is_corresponding":false,"raw_author_name":"Shulin Zeng","raw_affiliation_strings":["Tsinghua University,China"],"affiliations":[{"raw_affiliation_string":"Tsinghua University,China","institution_ids":["https://openalex.org/I99065089"]}]},{"author_position":"middle","author":{"id":"https://openalex.org/A5108792780","display_name":"Junbo Zhao","orcid":null},"institutions":[{"id":"https://openalex.org/I99065089","display_name":"Tsinghua University","ror":"https://ror.org/03cve4549","country_code":"CN","type":"education","lineage":["https://openalex.org/I99065089"]}],"countries":["CN"],"is_corresponding":false,"raw_author_name":"Junbo Zhao","raw_affiliation_strings":["Tsinghua University,China"],"affiliations":[{"raw_affiliation_string":"Tsinghua University,China","institution_ids":["https://openalex.org/I99065089"]}]},{"author_position":"middle","author":{"id":"https://openalex.org/A5108052606","display_name":"Li Ding","orcid":"https://orcid.org/0000-0003-2394-0778"},"institutions":[{"id":"https://openalex.org/I183067930","display_name":"Shanghai Jiao Tong University","ror":"https://ror.org/0220qvk04","country_code":"CN","type":"education","lineage":["https://openalex.org/I183067930"]}],"countries":["CN"],"is_corresponding":false,"raw_author_name":"Li Ding","raw_affiliation_strings":["Shanghai Jiao Tong University,China"],"affiliations":[{"raw_affiliation_string":"Shanghai Jiao Tong University,China","institution_ids":["https://openalex.org/I183067930"]}]},{"author_position":"middle","author":{"id":"https://openalex.org/A5100379682","display_name":"Zeyu Wang","orcid":"https://orcid.org/0000-0001-5374-6330"},"institutions":[{"id":"https://openalex.org/I183067930","display_name":"Shanghai Jiao Tong University","ror":"https://ror.org/0220qvk04","country_code":"CN","type":"education","lineage":["https://openalex.org/I183067930"]}],"countries":["CN"],"is_corresponding":false,"raw_author_name":"Zeyu Wang","raw_affiliation_strings":["Shanghai Jiao Tong University,China"],"affiliations":[{"raw_affiliation_string":"Shanghai Jiao Tong University,China","institution_ids":["https://openalex.org/I183067930"]}]},{"author_position":"middle","author":{"id":"https://openalex.org/A5101545422","display_name":"Jinhao Li","orcid":"https://orcid.org/0009-0001-1535-3664"},"institutions":[{"id":"https://openalex.org/I183067930","display_name":"Shanghai Jiao Tong University","ror":"https://ror.org/0220qvk04","country_code":"CN","type":"education","lineage":["https://openalex.org/I183067930"]}],"countries":["CN"],"is_corresponding":false,"raw_author_name":"Jinhao Li","raw_affiliation_strings":["Shanghai Jiao Tong University,China"],"affiliations":[{"raw_affiliation_string":"Shanghai Jiao Tong University,China","institution_ids":["https://openalex.org/I183067930"]}]},{"author_position":"middle","author":{"id":"https://openalex.org/A5103138440","display_name":"Zhenhua Zhu","orcid":"https://orcid.org/0009-0007-9259-7180"},"institutions":[{"id":"https://openalex.org/I99065089","display_name":"Tsinghua University","ror":"https://ror.org/03cve4549","country_code":"CN","type":"education","lineage":["https://openalex.org/I99065089"]}],"countries":["CN"],"is_corresponding":false,"raw_author_name":"Zhenhua Zhu","raw_affiliation_strings":["Tsinghua University,China"],"affiliations":[{"raw_affiliation_string":"Tsinghua University,China","institution_ids":["https://openalex.org/I99065089"]}]},{"author_position":"middle","author":{"id":"https://openalex.org/A5086217226","display_name":"Xuefei Ning","orcid":"https://orcid.org/0000-0003-2209-8312"},"institutions":[{"id":"https://openalex.org/I99065089","display_name":"Tsinghua University","ror":"https://ror.org/03cve4549","country_code":"CN","type":"education","lineage":["https://openalex.org/I99065089"]}],"countries":["CN"],"is_corresponding":false,"raw_author_name":"Xuefei Ning","raw_affiliation_strings":["Tsinghua University,China"],"affiliations":[{"raw_affiliation_string":"Tsinghua University,China","institution_ids":["https://openalex.org/I99065089"]}]},{"author_position":"middle","author":{"id":"https://openalex.org/A5100374175","display_name":"Chen Zhang","orcid":"https://orcid.org/0000-0003-2762-2726"},"institutions":[{"id":"https://openalex.org/I183067930","display_name":"Shanghai Jiao Tong University","ror":"https://ror.org/0220qvk04","country_code":"CN","type":"education","lineage":["https://openalex.org/I183067930"]}],"countries":["CN"],"is_corresponding":false,"raw_author_name":"Chen Zhang","raw_affiliation_strings":["Shanghai Jiao Tong University,China"],"affiliations":[{"raw_affiliation_string":"Shanghai Jiao Tong University,China","institution_ids":["https://openalex.org/I183067930"]}]},{"author_position":"middle","author":{"id":"https://openalex.org/A5100445061","display_name":"Yu Wang","orcid":"https://orcid.org/0000-0001-6108-5157"},"institutions":[{"id":"https://openalex.org/I99065089","display_name":"Tsinghua University","ror":"https://ror.org/03cve4549","country_code":"CN","type":"education","lineage":["https://openalex.org/I99065089"]}],"countries":["CN"],"is_corresponding":false,"raw_author_name":"Yu Wang","raw_affiliation_strings":["Tsinghua University,China"],"affiliations":[{"raw_affiliation_string":"Tsinghua University,China","institution_ids":["https://openalex.org/I99065089"]}]},{"author_position":"last","author":{"id":"https://openalex.org/A5015946486","display_name":"Guohao Dai","orcid":"https://orcid.org/0000-0003-0849-3252"},"institutions":[{"id":"https://openalex.org/I183067930","display_name":"Shanghai Jiao Tong University","ror":"https://ror.org/0220qvk04","country_code":"CN","type":"education","lineage":["https://openalex.org/I183067930"]}],"countries":["CN"],"is_corresponding":false,"raw_author_name":"Guohao Dai","raw_affiliation_strings":["Shanghai Jiao Tong University,China"],"affiliations":[{"raw_affiliation_string":"Shanghai Jiao Tong University,China","institution_ids":["https://openalex.org/I183067930"]}]}],"institutions":[],"countries_distinct_count":1,"institutions_distinct_count":11,"corresponding_author_ids":["https://openalex.org/A5100361691"],"corresponding_institution_ids":["https://openalex.org/I183067930"],"apc_list":null,"apc_paid":null,"fwci":10.2326,"has_fulltext":false,"cited_by_count":4,"citation_normalized_percentile":{"value":0.97917773,"is_in_top_1_percent":false,"is_in_top_10_percent":true},"cited_by_percentile_year":{"min":95,"max":100},"biblio":{"volume":null,"issue":null,"first_page":"949","last_page":"962"},"is_retracted":false,"is_paratext":false,"is_xpac":false,"primary_topic":{"id":"https://openalex.org/T10054","display_name":"Parallel Computing and Optimization Techniques","score":0.9872000217437744,"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":0.9872000217437744,"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/T13182","display_name":"Quantum-Dot Cellular Automata","score":0.9631999731063843,"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/T12303","display_name":"Tensor decomposition and applications","score":0.9563999772071838,"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"}}],"keywords":[{"id":"https://openalex.org/keywords/tensor","display_name":"Tensor (intrinsic definition)","score":0.5516760945320129},{"id":"https://openalex.org/keywords/block","display_name":"Block (permutation group theory)","score":0.5446594953536987},{"id":"https://openalex.org/keywords/computer-science","display_name":"Computer science","score":0.5346649289131165},{"id":"https://openalex.org/keywords/transposable-element","display_name":"Transposable element","score":0.5021622180938721},{"id":"https://openalex.org/keywords/core","display_name":"Core (optical fiber)","score":0.4248383939266205},{"id":"https://openalex.org/keywords/mathematics","display_name":"Mathematics","score":0.2689470648765564},{"id":"https://openalex.org/keywords/combinatorics","display_name":"Combinatorics","score":0.23989185690879822},{"id":"https://openalex.org/keywords/pure-mathematics","display_name":"Pure mathematics","score":0.12565824389457703},{"id":"https://openalex.org/keywords/genetics","display_name":"Genetics","score":0.11865547299385071},{"id":"https://openalex.org/keywords/biology","display_name":"Biology","score":0.0981931984424591},{"id":"https://openalex.org/keywords/genome","display_name":"Genome","score":0.0841352641582489},{"id":"https://openalex.org/keywords/telecommunications","display_name":"Telecommunications","score":0.07786226272583008}],"concepts":[{"id":"https://openalex.org/C155281189","wikidata":"https://www.wikidata.org/wiki/Q3518150","display_name":"Tensor (intrinsic definition)","level":2,"score":0.5516760945320129},{"id":"https://openalex.org/C2777210771","wikidata":"https://www.wikidata.org/wiki/Q4927124","display_name":"Block (permutation group theory)","level":2,"score":0.5446594953536987},{"id":"https://openalex.org/C41008148","wikidata":"https://www.wikidata.org/wiki/Q21198","display_name":"Computer science","level":0,"score":0.5346649289131165},{"id":"https://openalex.org/C4918238","wikidata":"https://www.wikidata.org/wiki/Q121438","display_name":"Transposable element","level":4,"score":0.5021622180938721},{"id":"https://openalex.org/C2164484","wikidata":"https://www.wikidata.org/wiki/Q5170150","display_name":"Core (optical fiber)","level":2,"score":0.4248383939266205},{"id":"https://openalex.org/C33923547","wikidata":"https://www.wikidata.org/wiki/Q395","display_name":"Mathematics","level":0,"score":0.2689470648765564},{"id":"https://openalex.org/C114614502","wikidata":"https://www.wikidata.org/wiki/Q76592","display_name":"Combinatorics","level":1,"score":0.23989185690879822},{"id":"https://openalex.org/C202444582","wikidata":"https://www.wikidata.org/wiki/Q837863","display_name":"Pure mathematics","level":1,"score":0.12565824389457703},{"id":"https://openalex.org/C54355233","wikidata":"https://www.wikidata.org/wiki/Q7162","display_name":"Genetics","level":1,"score":0.11865547299385071},{"id":"https://openalex.org/C86803240","wikidata":"https://www.wikidata.org/wiki/Q420","display_name":"Biology","level":0,"score":0.0981931984424591},{"id":"https://openalex.org/C141231307","wikidata":"https://www.wikidata.org/wiki/Q7020","display_name":"Genome","level":3,"score":0.0841352641582489},{"id":"https://openalex.org/C76155785","wikidata":"https://www.wikidata.org/wiki/Q418","display_name":"Telecommunications","level":1,"score":0.07786226272583008},{"id":"https://openalex.org/C104317684","wikidata":"https://www.wikidata.org/wiki/Q7187","display_name":"Gene","level":2,"score":0.0}],"mesh":[],"locations_count":1,"locations":[{"id":"doi:10.1109/hpca61900.2025.00075","is_oa":false,"landing_page_url":"https://doi.org/10.1109/hpca61900.2025.00075","pdf_url":null,"source":null,"license":null,"license_id":null,"version":"publishedVersion","is_accepted":true,"is_published":true,"raw_source_name":"2025 IEEE International Symposium on High Performance Computer Architecture (HPCA)","raw_type":"proceedings-article"}],"best_oa_location":null,"sustainable_development_goals":[],"awards":[],"funders":[{"id":"https://openalex.org/F4320314786","display_name":"Xilinx","ror":"https://ror.org/01rb7bk56"},{"id":"https://openalex.org/F4320321001","display_name":"National Natural Science Foundation of China","ror":"https://ror.org/01h0zpd94"},{"id":"https://openalex.org/F4320327803","display_name":"Shanghai Rising-Star Program","ror":null}],"has_content":{"grobid_xml":false,"pdf":false},"content_urls":null,"referenced_works_count":69,"referenced_works":["https://openalex.org/W2034861439","https://openalex.org/W2116341502","https://openalex.org/W2126004407","https://openalex.org/W2194775991","https://openalex.org/W2251939518","https://openalex.org/W2588191434","https://openalex.org/W2604319603","https://openalex.org/W2914631005","https://openalex.org/W2915106038","https://openalex.org/W2923014074","https://openalex.org/W2946609015","https://openalex.org/W2980186997","https://openalex.org/W2994881943","https://openalex.org/W2998617917","https://openalex.org/W3016542674","https://openalex.org/W3038012435","https://openalex.org/W3103168911","https://openalex.org/W3130554079","https://openalex.org/W3133253223","https://openalex.org/W3159727696","https://openalex.org/W3189877953","https://openalex.org/W3194676777","https://openalex.org/W3198975860","https://openalex.org/W3206453033","https://openalex.org/W4214686755","https://openalex.org/W4281660701","https://openalex.org/W4293023340","https://openalex.org/W4308083753","https://openalex.org/W4360831803","https://openalex.org/W4360831840","https://openalex.org/W4360831975","https://openalex.org/W4366341968","https://openalex.org/W4380881077","https://openalex.org/W4389166750","https://openalex.org/W4393145114","https://openalex.org/W4393578753","https://openalex.org/W4395106460","https://openalex.org/W6605323724","https://openalex.org/W6677103964","https://openalex.org/W6677580257","https://openalex.org/W6724998850","https://openalex.org/W6739901393","https://openalex.org/W6745148473","https://openalex.org/W6749838110","https://openalex.org/W6755166560","https://openalex.org/W6755207826","https://openalex.org/W6771579010","https://openalex.org/W6787972765","https://openalex.org/W6789946416","https://openalex.org/W6790943123","https://openalex.org/W6794457956","https://openalex.org/W6803815625","https://openalex.org/W6811340617","https://openalex.org/W6839441226","https://openalex.org/W6839700459","https://openalex.org/W6848451824","https://openalex.org/W6849185048","https://openalex.org/W6850625674","https://openalex.org/W6850769804","https://openalex.org/W6852785088","https://openalex.org/W6854094408","https://openalex.org/W6857923988","https://openalex.org/W6858266299","https://openalex.org/W6859083497","https://openalex.org/W6861303737","https://openalex.org/W6862187481","https://openalex.org/W6872946721","https://openalex.org/W6873037751","https://openalex.org/W6967006125"],"related_works":["https://openalex.org/W4391375266","https://openalex.org/W2899084033","https://openalex.org/W2748952813","https://openalex.org/W4237591666","https://openalex.org/W2103833188","https://openalex.org/W2164152191","https://openalex.org/W1534300978","https://openalex.org/W2387589589","https://openalex.org/W68905254","https://openalex.org/W2026307780"],"abstract_inverted_index":{"The":[0,24],"computational":[1,241],"and":[2,29,39,54,89,114,205,229,260],"memory":[3],"demands":[4],"of":[5,111,153,257],"Deep":[6],"Learning":[7],"(DL)":[8],"models,":[9],"from":[10],"convolutional":[11],"neural":[12],"networks":[13],"to":[14,36,46,61,119,138,177,203],"Large":[15],"Language":[16],"Models":[17],"(LLMs),":[18],"are":[19,117],"experiencing":[20],"a":[21,33,48,131,196,223,239],"notable":[22],"surge.":[23],"sparsification":[25],"(e.g.,":[26,77],"weight":[27,109],"pruning":[28],"sparse":[30,168,269],"attention)":[31],"represents":[32],"significant":[34],"approach":[35],"reducing":[37],"latency":[38],"energy":[40],"consumption.":[41],"However,":[42],"it":[43],"is":[44,106,136],"non-trivial":[45],"identify":[47],"good":[49],"trade-off":[50],"between":[51],"model":[52],"accuracy":[53,73,81,174],"hardware":[55,64],"efficiency.":[56],"Existing":[57],"work":[58],"has":[59],"sought":[60],"mitigate":[62],"the":[63,71,92,108,112,139,158,185,190,208,250],"complexity":[65],"overhead":[66],"through":[67],"structured":[68,85,132,149,161,181],"sparsity,":[69],"yet":[70],"resulting":[72],"loss":[74],"remains":[75],"considerable":[76],"more":[78],"than":[79],"6%":[80],"drop":[82],"with":[83,164,222,238,245],"50%":[84],"sparsity":[86,133,162,182,187],"on":[87],"OPT-6.7B":[88],"Llama2-7B).To":[90],"address":[91],"above":[93],"challenges,":[94],"this":[95,126],"paper":[96],"proposes":[97],"Transposable":[98,197],"Block-wise":[99,198],"Structured":[100],"Sparsity":[101],"(TBS).":[102],"Our":[103],"key":[104],"insight":[105],"that":[107,135],"matrices":[110],"forward":[113],"backward":[115],"pass":[116],"transposed":[118],"each":[120],"other":[121,180],"during":[122],"DL":[123,270],"training.":[124],"Exploiting":[125],"transposition":[127],"property":[128],"facilitates":[129],"obtaining":[130],"pattern":[134,163],"closer":[137],"unstructured":[140],"sparsity.":[141,150],"In":[142,151],"contrast,":[143],"existing":[144,246],"studies":[145,183],"explore":[146],"only":[147],"one-dimensional":[148],"light":[152],"these":[154],"observations,":[155],"we":[156,193],"propose":[157,194],"transposable":[159],"block-wise":[160],"an":[165,213,231,255,262],"efficient":[166],"end-to-end":[167],"training":[169],"method.":[170],"This":[171],"method":[172],"improves":[173,249],"by":[175,254],"up":[176],"2.58%":[178],"over":[179],"under":[184],"same":[186],"degree.":[188],"At":[189],"micro-architecture":[191],"level,":[192],"TB-STC,":[195],"N:M":[199],"Sparse":[200],"Tensor":[201],"Core":[202],"efficiently":[204],"flexibly":[206],"facilitate":[207],"TBS":[209],"pattern.":[210],"TB-STC":[211,248],"introduces":[212],"adaptive":[214],"codec":[215],"architecture":[216,234],"for":[217,235],"on-the-fly":[218],"storage":[219],"format":[220],"conversion":[221],"higher":[224],"bandwidth":[225],"utilization":[226,242],"(1.47":[227],"\u00d7),":[228],"implements":[230],"I/O-aware":[232],"configurable":[233],"sparsity-aware":[236],"scheduling":[237],"better":[240],"(1.57\u00d7).":[243],"Compared":[244],"work,":[247],"Energy-Delay":[251],"Product":[252],"(EDP)":[253],"average":[256],"3.82":[258],"\u00d7":[259],"offers":[261],"enhanced":[263],"accuracy-EDP":[264],"Pareto":[265],"frontier":[266],"across":[267],"various":[268],"models.":[271]},"counts_by_year":[{"year":2026,"cited_by_count":2},{"year":2025,"cited_by_count":2}],"updated_date":"2026-03-06T13:50:29.536080","created_date":"2025-10-10T00:00:00"}