{"id":"https://openalex.org/W4386474816","doi":"https://doi.org/10.1109/isvlsi59464.2023.10238524","title":"tubGEMM: Energy-Efficient and Sparsity-Effective Temporal-Unary-Binary Based Matrix Multiply Unit","display_name":"tubGEMM: Energy-Efficient and Sparsity-Effective Temporal-Unary-Binary Based Matrix Multiply Unit","publication_year":2023,"publication_date":"2023-06-20","ids":{"openalex":"https://openalex.org/W4386474816","doi":"https://doi.org/10.1109/isvlsi59464.2023.10238524"},"language":"en","primary_location":{"id":"doi:10.1109/isvlsi59464.2023.10238524","is_oa":false,"landing_page_url":"https://doi.org/10.1109/isvlsi59464.2023.10238524","pdf_url":null,"source":null,"license":null,"license_id":null,"version":"publishedVersion","is_accepted":true,"is_published":true,"raw_source_name":"2023 IEEE Computer Society Annual Symposium on VLSI (ISVLSI)","raw_type":"proceedings-article"},"type":"preprint","indexed_in":["arxiv","crossref"],"open_access":{"is_oa":true,"oa_status":"green","oa_url":"https://arxiv.org/pdf/2412.17955","any_repository_has_fulltext":true},"authorships":[{"author_position":"first","author":{"id":"https://openalex.org/A5021629244","display_name":"Prabhu Vellaisamy","orcid":"https://orcid.org/0009-0007-7750-8725"},"institutions":[{"id":"https://openalex.org/I74973139","display_name":"Carnegie Mellon University","ror":"https://ror.org/05x2bcf33","country_code":"US","type":"education","lineage":["https://openalex.org/I74973139"]}],"countries":["US"],"is_corresponding":true,"raw_author_name":"Prabhu Vellaisamy","raw_affiliation_strings":["ECE, Carnegie Mellon University","MediaTek USA Inc"],"affiliations":[{"raw_affiliation_string":"ECE, Carnegie Mellon University","institution_ids":["https://openalex.org/I74973139"]},{"raw_affiliation_string":"MediaTek USA Inc","institution_ids":[]}]},{"author_position":"middle","author":{"id":"https://openalex.org/A5011022699","display_name":"Harideep Nair","orcid":"https://orcid.org/0009-0008-8015-4339"},"institutions":[{"id":"https://openalex.org/I74973139","display_name":"Carnegie Mellon University","ror":"https://ror.org/05x2bcf33","country_code":"US","type":"education","lineage":["https://openalex.org/I74973139"]}],"countries":["US"],"is_corresponding":false,"raw_author_name":"Harideep Nair","raw_affiliation_strings":["ECE, Carnegie Mellon University","MediaTek USA Inc"],"affiliations":[{"raw_affiliation_string":"ECE, Carnegie Mellon University","institution_ids":["https://openalex.org/I74973139"]},{"raw_affiliation_string":"MediaTek USA Inc","institution_ids":[]}]},{"author_position":"middle","author":{"id":"https://openalex.org/A5047079251","display_name":"J. Paul Finn","orcid":null},"institutions":[{"id":"https://openalex.org/I74973139","display_name":"Carnegie Mellon University","ror":"https://ror.org/05x2bcf33","country_code":"US","type":"education","lineage":["https://openalex.org/I74973139"]}],"countries":["US"],"is_corresponding":false,"raw_author_name":"Joseph Finn","raw_affiliation_strings":["ECE, Carnegie Mellon University"],"affiliations":[{"raw_affiliation_string":"ECE, Carnegie Mellon University","institution_ids":["https://openalex.org/I74973139"]}]},{"author_position":"middle","author":{"id":"https://openalex.org/A5106977572","display_name":"Manav Trivedi","orcid":null},"institutions":[{"id":"https://openalex.org/I74973139","display_name":"Carnegie Mellon University","ror":"https://ror.org/05x2bcf33","country_code":"US","type":"education","lineage":["https://openalex.org/I74973139"]}],"countries":["US"],"is_corresponding":false,"raw_author_name":"Manav Trivedi","raw_affiliation_strings":["ECE, Carnegie Mellon University"],"affiliations":[{"raw_affiliation_string":"ECE, Carnegie Mellon University","institution_ids":["https://openalex.org/I74973139"]}]},{"author_position":"middle","author":{"id":"https://openalex.org/A5100428823","display_name":"Albert Chen","orcid":"https://orcid.org/0000-0003-3708-3332"},"institutions":[{"id":"https://openalex.org/I74973139","display_name":"Carnegie Mellon University","ror":"https://ror.org/05x2bcf33","country_code":"US","type":"education","lineage":["https://openalex.org/I74973139"]}],"countries":["US"],"is_corresponding":false,"raw_author_name":"Albert Chen","raw_affiliation_strings":["ECE, Carnegie Mellon University"],"affiliations":[{"raw_affiliation_string":"ECE, Carnegie Mellon University","institution_ids":["https://openalex.org/I74973139"]}]},{"author_position":"middle","author":{"id":"https://openalex.org/A5033362933","display_name":"Anna Li","orcid":"https://orcid.org/0000-0002-5753-1430"},"institutions":[{"id":"https://openalex.org/I74973139","display_name":"Carnegie Mellon University","ror":"https://ror.org/05x2bcf33","country_code":"US","type":"education","lineage":["https://openalex.org/I74973139"]}],"countries":["US"],"is_corresponding":false,"raw_author_name":"Anna Li","raw_affiliation_strings":["ECE, Carnegie Mellon University"],"affiliations":[{"raw_affiliation_string":"ECE, Carnegie Mellon University","institution_ids":["https://openalex.org/I74973139"]}]},{"author_position":"middle","author":{"id":"https://openalex.org/A5101781727","display_name":"Tsung-Han Lin","orcid":"https://orcid.org/0000-0003-0149-7621"},"institutions":[],"countries":[],"is_corresponding":false,"raw_author_name":"Tsung-Han Lin","raw_affiliation_strings":["MediaTek USA Inc"],"affiliations":[{"raw_affiliation_string":"MediaTek USA Inc","institution_ids":[]}]},{"author_position":"middle","author":{"id":"https://openalex.org/A5007621907","display_name":"Peiran Wang","orcid":"https://orcid.org/0009-0009-3731-1156"},"institutions":[],"countries":[],"is_corresponding":false,"raw_author_name":"Perry Wang","raw_affiliation_strings":["MediaTek USA Inc"],"affiliations":[{"raw_affiliation_string":"MediaTek USA Inc","institution_ids":[]}]},{"author_position":"middle","author":{"id":"https://openalex.org/A5039038157","display_name":"R. D. Blanton","orcid":"https://orcid.org/0000-0001-6108-2925"},"institutions":[{"id":"https://openalex.org/I74973139","display_name":"Carnegie Mellon University","ror":"https://ror.org/05x2bcf33","country_code":"US","type":"education","lineage":["https://openalex.org/I74973139"]}],"countries":["US"],"is_corresponding":false,"raw_author_name":"Shawn Blanton","raw_affiliation_strings":["ECE, Carnegie Mellon University"],"affiliations":[{"raw_affiliation_string":"ECE, Carnegie Mellon University","institution_ids":["https://openalex.org/I74973139"]}]},{"author_position":"last","author":{"id":"https://openalex.org/A5102751221","display_name":"John Paul Shen","orcid":"https://orcid.org/0000-0002-7225-0629"},"institutions":[{"id":"https://openalex.org/I74973139","display_name":"Carnegie Mellon University","ror":"https://ror.org/05x2bcf33","country_code":"US","type":"education","lineage":["https://openalex.org/I74973139"]}],"countries":["US"],"is_corresponding":false,"raw_author_name":"John Paul Shen","raw_affiliation_strings":["ECE, Carnegie Mellon University"],"affiliations":[{"raw_affiliation_string":"ECE, Carnegie Mellon University","institution_ids":["https://openalex.org/I74973139"]}]}],"institutions":[],"countries_distinct_count":1,"institutions_distinct_count":10,"corresponding_author_ids":["https://openalex.org/A5021629244"],"corresponding_institution_ids":["https://openalex.org/I74973139"],"apc_list":null,"apc_paid":null,"fwci":0.9159,"has_fulltext":true,"cited_by_count":3,"citation_normalized_percentile":{"value":0.71759117,"is_in_top_1_percent":false,"is_in_top_10_percent":false},"cited_by_percentile_year":{"min":90,"max":96},"biblio":{"volume":null,"issue":null,"first_page":"1","last_page":"6"},"is_retracted":false,"is_paratext":false,"is_xpac":false,"primary_topic":{"id":"https://openalex.org/T10054","display_name":"Parallel Computing and Optimization Techniques","score":0.9994000196456909,"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":0.9994000196456909,"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/T10502","display_name":"Advanced Memory and Neural Computing","score":0.9988999962806702,"subfield":{"id":"https://openalex.org/subfields/2208","display_name":"Electrical and Electronic Engineering"},"field":{"id":"https://openalex.org/fields/22","display_name":"Engineering"},"domain":{"id":"https://openalex.org/domains/3","display_name":"Physical Sciences"}},{"id":"https://openalex.org/T11321","display_name":"Error Correcting Code Techniques","score":0.9987999796867371,"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/unary-operation","display_name":"Unary operation","score":0.9510474801063538},{"id":"https://openalex.org/keywords/binary-number","display_name":"Binary number","score":0.6173069477081299},{"id":"https://openalex.org/keywords/computer-science","display_name":"Computer science","score":0.6129881739616394},{"id":"https://openalex.org/keywords/algorithm","display_name":"Algorithm","score":0.5756546258926392},{"id":"https://openalex.org/keywords/energy","display_name":"Energy (signal processing)","score":0.5738694071769714},{"id":"https://openalex.org/keywords/matrix","display_name":"Matrix (chemical analysis)","score":0.49310705065727234},{"id":"https://openalex.org/keywords/encoding","display_name":"Encoding (memory)","score":0.45759516954421997},{"id":"https://openalex.org/keywords/parallel-computing","display_name":"Parallel computing","score":0.4444776475429535},{"id":"https://openalex.org/keywords/power","display_name":"Power (physics)","score":0.43946096301078796},{"id":"https://openalex.org/keywords/matrix-multiplication","display_name":"Matrix multiplication","score":0.4138805866241455},{"id":"https://openalex.org/keywords/mathematics","display_name":"Mathematics","score":0.20695790648460388},{"id":"https://openalex.org/keywords/arithmetic","display_name":"Arithmetic","score":0.20440036058425903},{"id":"https://openalex.org/keywords/discrete-mathematics","display_name":"Discrete mathematics","score":0.17055478692054749},{"id":"https://openalex.org/keywords/artificial-intelligence","display_name":"Artificial intelligence","score":0.10431823134422302},{"id":"https://openalex.org/keywords/physics","display_name":"Physics","score":0.09225666522979736}],"concepts":[{"id":"https://openalex.org/C78023250","wikidata":"https://www.wikidata.org/wiki/Q657596","display_name":"Unary operation","level":2,"score":0.9510474801063538},{"id":"https://openalex.org/C48372109","wikidata":"https://www.wikidata.org/wiki/Q3913","display_name":"Binary number","level":2,"score":0.6173069477081299},{"id":"https://openalex.org/C41008148","wikidata":"https://www.wikidata.org/wiki/Q21198","display_name":"Computer science","level":0,"score":0.6129881739616394},{"id":"https://openalex.org/C11413529","wikidata":"https://www.wikidata.org/wiki/Q8366","display_name":"Algorithm","level":1,"score":0.5756546258926392},{"id":"https://openalex.org/C186370098","wikidata":"https://www.wikidata.org/wiki/Q442787","display_name":"Energy (signal processing)","level":2,"score":0.5738694071769714},{"id":"https://openalex.org/C106487976","wikidata":"https://www.wikidata.org/wiki/Q685816","display_name":"Matrix (chemical analysis)","level":2,"score":0.49310705065727234},{"id":"https://openalex.org/C125411270","wikidata":"https://www.wikidata.org/wiki/Q18653","display_name":"Encoding (memory)","level":2,"score":0.45759516954421997},{"id":"https://openalex.org/C173608175","wikidata":"https://www.wikidata.org/wiki/Q232661","display_name":"Parallel computing","level":1,"score":0.4444776475429535},{"id":"https://openalex.org/C163258240","wikidata":"https://www.wikidata.org/wiki/Q25342","display_name":"Power (physics)","level":2,"score":0.43946096301078796},{"id":"https://openalex.org/C17349429","wikidata":"https://www.wikidata.org/wiki/Q1049914","display_name":"Matrix multiplication","level":3,"score":0.4138805866241455},{"id":"https://openalex.org/C33923547","wikidata":"https://www.wikidata.org/wiki/Q395","display_name":"Mathematics","level":0,"score":0.20695790648460388},{"id":"https://openalex.org/C94375191","wikidata":"https://www.wikidata.org/wiki/Q11205","display_name":"Arithmetic","level":1,"score":0.20440036058425903},{"id":"https://openalex.org/C118615104","wikidata":"https://www.wikidata.org/wiki/Q121416","display_name":"Discrete mathematics","level":1,"score":0.17055478692054749},{"id":"https://openalex.org/C154945302","wikidata":"https://www.wikidata.org/wiki/Q11660","display_name":"Artificial intelligence","level":1,"score":0.10431823134422302},{"id":"https://openalex.org/C121332964","wikidata":"https://www.wikidata.org/wiki/Q413","display_name":"Physics","level":0,"score":0.09225666522979736},{"id":"https://openalex.org/C159985019","wikidata":"https://www.wikidata.org/wiki/Q181790","display_name":"Composite material","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/C62520636","wikidata":"https://www.wikidata.org/wiki/Q944","display_name":"Quantum mechanics","level":1,"score":0.0},{"id":"https://openalex.org/C105795698","wikidata":"https://www.wikidata.org/wiki/Q12483","display_name":"Statistics","level":1,"score":0.0},{"id":"https://openalex.org/C192562407","wikidata":"https://www.wikidata.org/wiki/Q228736","display_name":"Materials science","level":0,"score":0.0}],"mesh":[],"locations_count":2,"locations":[{"id":"doi:10.1109/isvlsi59464.2023.10238524","is_oa":false,"landing_page_url":"https://doi.org/10.1109/isvlsi59464.2023.10238524","pdf_url":null,"source":null,"license":null,"license_id":null,"version":"publishedVersion","is_accepted":true,"is_published":true,"raw_source_name":"2023 IEEE Computer Society Annual Symposium on VLSI (ISVLSI)","raw_type":"proceedings-article"},{"id":"pmh:oai:arXiv.org:2412.17955","is_oa":true,"landing_page_url":"http://arxiv.org/abs/2412.17955","pdf_url":"https://arxiv.org/pdf/2412.17955","source":{"id":"https://openalex.org/S4306400194","display_name":"arXiv (Cornell University)","issn_l":null,"issn":null,"is_oa":true,"is_in_doaj":false,"is_core":false,"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"}],"best_oa_location":{"id":"pmh:oai:arXiv.org:2412.17955","is_oa":true,"landing_page_url":"http://arxiv.org/abs/2412.17955","pdf_url":"https://arxiv.org/pdf/2412.17955","source":{"id":"https://openalex.org/S4306400194","display_name":"arXiv (Cornell University)","issn_l":null,"issn":null,"is_oa":true,"is_in_doaj":false,"is_core":false,"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"},"sustainable_development_goals":[{"id":"https://metadata.un.org/sdg/7","display_name":"Affordable and clean energy","score":0.9100000262260437}],"awards":[],"funders":[],"has_content":{"pdf":true,"grobid_xml":false},"content_urls":{"pdf":"https://content.openalex.org/works/W4386474816.pdf"},"referenced_works_count":13,"referenced_works":["https://openalex.org/W1988425770","https://openalex.org/W2003056114","https://openalex.org/W2296101864","https://openalex.org/W2606722458","https://openalex.org/W2625226188","https://openalex.org/W2883012749","https://openalex.org/W2896608207","https://openalex.org/W2943267175","https://openalex.org/W2973538245","https://openalex.org/W3043216033","https://openalex.org/W3101100041","https://openalex.org/W4280635517","https://openalex.org/W6785662278"],"related_works":["https://openalex.org/W2888918612","https://openalex.org/W2166285859","https://openalex.org/W3148597776","https://openalex.org/W1538206641","https://openalex.org/W2025370853","https://openalex.org/W3211656382","https://openalex.org/W2060412688","https://openalex.org/W2035521462","https://openalex.org/W1925544630","https://openalex.org/W2004686618"],"abstract_inverted_index":{"General":[0],"Matrix":[1],"Multiplication":[2],"(GEMM)":[3],"is":[4],"a":[5,56],"ubiquitous":[6],"compute":[7],"kernel":[8],"in":[9,116,141],"deep":[10],"learning":[11],"(DL).":[12],"To":[13],"support":[14],"energy-efficient":[15],"edge-native":[16],"processing,":[17],"new":[18],"GEMM":[19],"hardware":[20],"units":[21],"have":[22],"been":[23],"proposed":[24],"that":[25,61],"operate":[26],"on":[27,41,113],"unary":[28,36,43,90],"encoded":[29],"bitstreams":[30],"using":[31],"much":[32],"simpler":[33],"hardware.":[34],"Most":[35],"approaches":[37],"thus":[38],"far":[39],"focus":[40],"rate-based":[42],"encoding":[44,68],"of":[45],"values":[46],"and":[47,65,69,98,103,132,152,157,165],"perform":[48],"stochastic":[49],"approximate":[50],"computation.":[51],"This":[52],"work":[53],"presents":[54],"tubGEMM,":[55],"novel":[57],"matrix-multiply":[58],"unit":[59],"design":[60,91,108],"employs":[62],"hybrid":[63],"temporal-unary":[64],"binary":[66],"(tub)":[67],"performs":[70],"exact":[71],"(not":[72],"approximate)":[73],"GEMM.":[74],"It":[75],"intrinsically":[76],"exploits":[77],"dynamic":[78],"value":[79],"sparsity":[80,140],"to":[81,86,155],"improve":[82],"energy":[83,99,147],"efficiency.":[84],"Compared":[85],"the":[87],"current":[88],"best":[89],"uGEMM,":[92],"tubGEMM":[93,107],"significantly":[94],"reduces":[95,146,161],"area,":[96,128],"power,":[97,131],"by":[100,148,163],"89%,":[101],"87%,":[102],"50%":[104],"respectively.":[105,167],"A":[106],"performing":[109],"128x128":[110],"matrix":[111],"multiply":[112],"8-bit":[114],"integers,":[115],"commercial":[117],"TSMC":[118],"N5":[119],"(5nm)":[120],"process":[121],"node,":[122],"consumes":[123],"just":[124],"0.22":[125],"m$\\mathrm{m}^{2}$":[126],"die":[127],"417.72":[129],"mW":[130],"8.86":[133],"$\\mu$J":[134],"energy,":[135],"assuming":[136],"no":[137],"sparsity.":[138],"Typical":[139],"DL":[142],"workloads":[143],"(MobileNetv2,":[144],"ResNet50)":[145],"more":[149],"than":[150],"3x,":[151],"lowering":[153],"precision":[154],"4":[156],"2":[158],"bits":[159],"further":[160],"it":[162],"24x":[164],"104x":[166]},"counts_by_year":[{"year":2025,"cited_by_count":2},{"year":2024,"cited_by_count":1}],"updated_date":"2026-03-20T23:20:44.827607","created_date":"2025-10-10T00:00:00"}