{"id":"https://openalex.org/W4415883106","doi":"https://doi.org/10.1109/tetc.2025.3626723","title":"MaxiMoff: Designing Matrix Multiplication Accelerator for Effective Multiply-Add Operations Offloading","display_name":"MaxiMoff: Designing Matrix Multiplication Accelerator for Effective Multiply-Add Operations Offloading","publication_year":2025,"publication_date":"2025-11-04","ids":{"openalex":"https://openalex.org/W4415883106","doi":"https://doi.org/10.1109/tetc.2025.3626723"},"language":null,"primary_location":{"id":"doi:10.1109/tetc.2025.3626723","is_oa":false,"landing_page_url":"https://doi.org/10.1109/tetc.2025.3626723","pdf_url":null,"source":{"id":"https://openalex.org/S2496326734","display_name":"IEEE Transactions on Emerging Topics in Computing","issn_l":"2168-6750","issn":["2168-6750","2376-4562"],"is_oa":false,"is_in_doaj":false,"is_core":true,"host_organization":"https://openalex.org/P4310319808","host_organization_name":"Institute of Electrical and Electronics Engineers","host_organization_lineage":["https://openalex.org/P4310319808"],"host_organization_lineage_names":["Institute of Electrical and Electronics Engineers"],"type":"journal"},"license":null,"license_id":null,"version":"publishedVersion","is_accepted":true,"is_published":true,"raw_source_name":"IEEE Transactions on Emerging Topics in Computing","raw_type":"journal-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/A5064083568","display_name":"S. Kim","orcid":"https://orcid.org/0009-0009-0106-6590"},"institutions":[{"id":"https://openalex.org/I193775966","display_name":"Yonsei University","ror":"https://ror.org/01wjejq96","country_code":"KR","type":"education","lineage":["https://openalex.org/I193775966"]}],"countries":["KR"],"is_corresponding":true,"raw_author_name":"Sungwoo Kim","raw_affiliation_strings":["School of Electrical and Electronic Engineering, Yonsei University, Seoul, South Korea"],"raw_orcid":"https://orcid.org/0009-0009-0106-6590","affiliations":[{"raw_affiliation_string":"School of Electrical and Electronic Engineering, Yonsei University, Seoul, South Korea","institution_ids":["https://openalex.org/I193775966"]}]},{"author_position":"middle","author":{"id":"https://openalex.org/A5103111147","display_name":"Dongho Ha","orcid":"https://orcid.org/0009-0005-4090-4025"},"institutions":[{"id":"https://openalex.org/I4210092558","display_name":"BC Platforms (Finland)","ror":"https://ror.org/002fen565","country_code":"FI","type":"company","lineage":["https://openalex.org/I4210092558"]},{"id":"https://openalex.org/I4210122681","display_name":"Aqua Metrology Systems (United States)","ror":"https://ror.org/02mvdhz93","country_code":"US","type":"company","lineage":["https://openalex.org/I4210122681"]}],"countries":["FI","US"],"is_corresponding":false,"raw_author_name":"Dongho Ha","raw_affiliation_strings":["Meta Platforms Inc., Sunnyvale, CA, USA"],"raw_orcid":"https://orcid.org/0009-0005-4090-4025","affiliations":[{"raw_affiliation_string":"Meta Platforms Inc., Sunnyvale, CA, USA","institution_ids":["https://openalex.org/I4210092558","https://openalex.org/I4210122681"]}]},{"author_position":"middle","author":{"id":"https://openalex.org/A5102922751","display_name":"Seunghwan Sung","orcid":"https://orcid.org/0009-0008-2523-7043"},"institutions":[{"id":"https://openalex.org/I193775966","display_name":"Yonsei University","ror":"https://ror.org/01wjejq96","country_code":"KR","type":"education","lineage":["https://openalex.org/I193775966"]}],"countries":["KR"],"is_corresponding":false,"raw_author_name":"Seunghwan Sung","raw_affiliation_strings":["School of Electrical and Electronic Engineering, Yonsei University, Seoul, South Korea"],"raw_orcid":"https://orcid.org/0009-0008-2523-7043","affiliations":[{"raw_affiliation_string":"School of Electrical and Electronic Engineering, Yonsei University, Seoul, South Korea","institution_ids":["https://openalex.org/I193775966"]}]},{"author_position":"last","author":{"id":"https://openalex.org/A5017913155","display_name":"Won Woo Ro","orcid":"https://orcid.org/0000-0001-5390-6445"},"institutions":[{"id":"https://openalex.org/I193775966","display_name":"Yonsei University","ror":"https://ror.org/01wjejq96","country_code":"KR","type":"education","lineage":["https://openalex.org/I193775966"]}],"countries":["KR"],"is_corresponding":false,"raw_author_name":"Won Woo Ro","raw_affiliation_strings":["School of Electrical and Electronic Engineering, Yonsei University, Seoul, South Korea"],"raw_orcid":"https://orcid.org/0000-0001-5390-6445","affiliations":[{"raw_affiliation_string":"School of Electrical and Electronic Engineering, Yonsei University, Seoul, South Korea","institution_ids":["https://openalex.org/I193775966"]}]}],"institutions":[],"countries_distinct_count":3,"institutions_distinct_count":4,"corresponding_author_ids":["https://openalex.org/A5064083568"],"corresponding_institution_ids":["https://openalex.org/I193775966"],"apc_list":null,"apc_paid":null,"fwci":0.0,"has_fulltext":false,"cited_by_count":0,"citation_normalized_percentile":{"value":0.34303342,"is_in_top_1_percent":false,"is_in_top_10_percent":false},"cited_by_percentile_year":null,"biblio":{"volume":"14","issue":"1","first_page":"3","last_page":"16"},"is_retracted":false,"is_paratext":false,"is_xpac":false,"primary_topic":{"id":"https://openalex.org/T10054","display_name":"Parallel Computing and Optimization Techniques","score":0.47440001368522644,"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":0.47440001368522644,"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/T10363","display_name":"Low-power high-performance VLSI design","score":0.15549999475479126,"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/T12808","display_name":"Ferroelectric and Negative Capacitance Devices","score":0.14560000598430634,"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"}}],"keywords":[{"id":"https://openalex.org/keywords/matrix-multiplication","display_name":"Matrix multiplication","score":0.7462000250816345},{"id":"https://openalex.org/keywords/speedup","display_name":"Speedup","score":0.6686999797821045},{"id":"https://openalex.org/keywords/leverage","display_name":"Leverage (statistics)","score":0.6140000224113464},{"id":"https://openalex.org/keywords/energy-consumption","display_name":"Energy consumption","score":0.5170000195503235},{"id":"https://openalex.org/keywords/adaptability","display_name":"Adaptability","score":0.4422999918460846},{"id":"https://openalex.org/keywords/process","display_name":"Process (computing)","score":0.4399999976158142},{"id":"https://openalex.org/keywords/multiplication","display_name":"Multiplication (music)","score":0.43059998750686646},{"id":"https://openalex.org/keywords/limiting","display_name":"Limiting","score":0.4223000109195709},{"id":"https://openalex.org/keywords/hardware-acceleration","display_name":"Hardware acceleration","score":0.4205000102519989}],"concepts":[{"id":"https://openalex.org/C41008148","wikidata":"https://www.wikidata.org/wiki/Q21198","display_name":"Computer science","level":0,"score":0.8734999895095825},{"id":"https://openalex.org/C17349429","wikidata":"https://www.wikidata.org/wiki/Q1049914","display_name":"Matrix multiplication","level":3,"score":0.7462000250816345},{"id":"https://openalex.org/C68339613","wikidata":"https://www.wikidata.org/wiki/Q1549489","display_name":"Speedup","level":2,"score":0.6686999797821045},{"id":"https://openalex.org/C153083717","wikidata":"https://www.wikidata.org/wiki/Q6535263","display_name":"Leverage (statistics)","level":2,"score":0.6140000224113464},{"id":"https://openalex.org/C173608175","wikidata":"https://www.wikidata.org/wiki/Q232661","display_name":"Parallel computing","level":1,"score":0.5961999893188477},{"id":"https://openalex.org/C2780165032","wikidata":"https://www.wikidata.org/wiki/Q16869822","display_name":"Energy consumption","level":2,"score":0.5170000195503235},{"id":"https://openalex.org/C177606310","wikidata":"https://www.wikidata.org/wiki/Q5674297","display_name":"Adaptability","level":2,"score":0.4422999918460846},{"id":"https://openalex.org/C98045186","wikidata":"https://www.wikidata.org/wiki/Q205663","display_name":"Process (computing)","level":2,"score":0.4399999976158142},{"id":"https://openalex.org/C2780595030","wikidata":"https://www.wikidata.org/wiki/Q3860309","display_name":"Multiplication (music)","level":2,"score":0.43059998750686646},{"id":"https://openalex.org/C188198153","wikidata":"https://www.wikidata.org/wiki/Q1613840","display_name":"Limiting","level":2,"score":0.4223000109195709},{"id":"https://openalex.org/C13164978","wikidata":"https://www.wikidata.org/wiki/Q600158","display_name":"Hardware acceleration","level":3,"score":0.4205000102519989},{"id":"https://openalex.org/C106487976","wikidata":"https://www.wikidata.org/wiki/Q685816","display_name":"Matrix (chemical analysis)","level":2,"score":0.41370001435279846},{"id":"https://openalex.org/C42935608","wikidata":"https://www.wikidata.org/wiki/Q190411","display_name":"Field-programmable gate array","level":2,"score":0.3634999990463257},{"id":"https://openalex.org/C39096654","wikidata":"https://www.wikidata.org/wiki/Q728507","display_name":"Strassen algorithm","level":4,"score":0.35749998688697815},{"id":"https://openalex.org/C56372850","wikidata":"https://www.wikidata.org/wiki/Q1050404","display_name":"Sparse matrix","level":3,"score":0.35589998960494995},{"id":"https://openalex.org/C9390403","wikidata":"https://www.wikidata.org/wiki/Q3966","display_name":"Computer hardware","level":1,"score":0.3472000062465668},{"id":"https://openalex.org/C149635348","wikidata":"https://www.wikidata.org/wiki/Q193040","display_name":"Embedded system","level":1,"score":0.3398999869823456},{"id":"https://openalex.org/C2742236","wikidata":"https://www.wikidata.org/wiki/Q924713","display_name":"Efficient energy use","level":2,"score":0.33889999985694885},{"id":"https://openalex.org/C123657996","wikidata":"https://www.wikidata.org/wiki/Q12271","display_name":"Architecture","level":2,"score":0.3386000096797943},{"id":"https://openalex.org/C118524514","wikidata":"https://www.wikidata.org/wiki/Q173212","display_name":"Computer architecture","level":1,"score":0.32339999079704285},{"id":"https://openalex.org/C186370098","wikidata":"https://www.wikidata.org/wiki/Q442787","display_name":"Energy (signal processing)","level":2,"score":0.3019999861717224},{"id":"https://openalex.org/C459310","wikidata":"https://www.wikidata.org/wiki/Q117801","display_name":"Computational science","level":1,"score":0.29649999737739563},{"id":"https://openalex.org/C120314980","wikidata":"https://www.wikidata.org/wiki/Q180634","display_name":"Distributed computing","level":1,"score":0.2955999970436096},{"id":"https://openalex.org/C63479239","wikidata":"https://www.wikidata.org/wiki/Q7353546","display_name":"Robustness (evolution)","level":3,"score":0.28360000252723694},{"id":"https://openalex.org/C147297375","wikidata":"https://www.wikidata.org/wiki/Q6674930","display_name":"Look-ahead","level":2,"score":0.2736000120639801},{"id":"https://openalex.org/C172430144","wikidata":"https://www.wikidata.org/wiki/Q17111997","display_name":"Symmetric multiprocessor system","level":2,"score":0.26589998602867126},{"id":"https://openalex.org/C50630238","wikidata":"https://www.wikidata.org/wiki/Q971505","display_name":"General-purpose computing on graphics processing units","level":3,"score":0.2615000009536743},{"id":"https://openalex.org/C42355184","wikidata":"https://www.wikidata.org/wiki/Q1361088","display_name":"Matrix decomposition","level":3,"score":0.2612999975681305},{"id":"https://openalex.org/C83283714","wikidata":"https://www.wikidata.org/wiki/Q121117","display_name":"Supercomputer","level":2,"score":0.26100000739097595},{"id":"https://openalex.org/C2778119891","wikidata":"https://www.wikidata.org/wiki/Q477690","display_name":"CUDA","level":2,"score":0.2578999996185303},{"id":"https://openalex.org/C106516650","wikidata":"https://www.wikidata.org/wiki/Q8366","display_name":"Algorithm design","level":2,"score":0.25679999589920044},{"id":"https://openalex.org/C2778456923","wikidata":"https://www.wikidata.org/wiki/Q5337692","display_name":"Edge computing","level":3,"score":0.25619998574256897},{"id":"https://openalex.org/C142962650","wikidata":"https://www.wikidata.org/wiki/Q240838","display_name":"Reconfigurable computing","level":3,"score":0.25609999895095825}],"mesh":[],"locations_count":1,"locations":[{"id":"doi:10.1109/tetc.2025.3626723","is_oa":false,"landing_page_url":"https://doi.org/10.1109/tetc.2025.3626723","pdf_url":null,"source":{"id":"https://openalex.org/S2496326734","display_name":"IEEE Transactions on Emerging Topics in Computing","issn_l":"2168-6750","issn":["2168-6750","2376-4562"],"is_oa":false,"is_in_doaj":false,"is_core":true,"host_organization":"https://openalex.org/P4310319808","host_organization_name":"Institute of Electrical and Electronics Engineers","host_organization_lineage":["https://openalex.org/P4310319808"],"host_organization_lineage_names":["Institute of Electrical and Electronics Engineers"],"type":"journal"},"license":null,"license_id":null,"version":"publishedVersion","is_accepted":true,"is_published":true,"raw_source_name":"IEEE Transactions on Emerging Topics in Computing","raw_type":"journal-article"}],"best_oa_location":null,"sustainable_development_goals":[],"awards":[{"id":"https://openalex.org/G1729495879","display_name":null,"funder_award_id":"IO250214-11969-01","funder_id":"https://openalex.org/F4320332195","funder_display_name":"Samsung"}],"funders":[{"id":"https://openalex.org/F4320322120","display_name":"National Research Foundation of Korea","ror":"https://ror.org/013aysd81"},{"id":"https://openalex.org/F4320332195","display_name":"Samsung","ror":"https://ror.org/04w3jy968"}],"has_content":{"grobid_xml":false,"pdf":false},"content_urls":null,"referenced_works_count":35,"referenced_works":["https://openalex.org/W1584707396","https://openalex.org/W1965512225","https://openalex.org/W1971868176","https://openalex.org/W1976168687","https://openalex.org/W1979527452","https://openalex.org/W2028960610","https://openalex.org/W2044001169","https://openalex.org/W2080592089","https://openalex.org/W2106191211","https://openalex.org/W2149234156","https://openalex.org/W2155503253","https://openalex.org/W2791673912","https://openalex.org/W2895305554","https://openalex.org/W2901549770","https://openalex.org/W2963989532","https://openalex.org/W2980186997","https://openalex.org/W2995872593","https://openalex.org/W3042610241","https://openalex.org/W3044913359","https://openalex.org/W3046751266","https://openalex.org/W3102510044","https://openalex.org/W3132532188","https://openalex.org/W3187908937","https://openalex.org/W3201073812","https://openalex.org/W3205226039","https://openalex.org/W3205883294","https://openalex.org/W3206857466","https://openalex.org/W3207730444","https://openalex.org/W4213195788","https://openalex.org/W4221160294","https://openalex.org/W4225869830","https://openalex.org/W4234988573","https://openalex.org/W4283025350","https://openalex.org/W4360831824","https://openalex.org/W4405756040"],"related_works":[],"abstract_inverted_index":{"Contemporary":[0],"GPU":[1,100],"architectures":[2],"integrate":[3],"specialized":[4],"computing":[5,33,109],"units":[6,13],"for":[7],"matrix":[8,11,27,41],"multiplication,":[9],"named":[10],"multiplication":[12,74],"(MXUs),":[14],"to":[15,26,40,47,102],"effectively":[16],"process":[17],"neural":[18],"network":[19],"applications.":[20],"However,":[21],"since":[22],"MXUs":[23,49,107],"are":[24,38,54],"limited":[25],"multiplications,":[28],"GPUs":[29],"show":[30],"inefficiencies":[31],"in":[32,50],"resource":[34],"utilization":[35,64],"while":[36,108],"applications":[37],"unrelated":[39],"multiplications.":[42],"Furthermore,":[43],"despite":[44],"prior":[45],"work":[46],"leverage":[48],"general-purpose":[51,104],"computing,":[52],"they":[53],"constrained":[55],"by":[56,129],"static":[57],"analysis,":[58],"limiting":[59],"their":[60],"adaptability":[61],"and":[62,106,124],"hardware":[63],"efficiency.":[65],"This":[66],"study":[67],"observes":[68],"that":[69],"the":[70],"techniques":[71],"emulating":[72],"high-bitwidth":[73,81],"with":[75],"low-bitwidth":[76,87],"ones":[77],"transform":[78],"a":[79,86,98],"single":[80],"Multiply-and-ADd":[82],"(MAD)":[83],"operation":[84],"into":[85],"dot-product":[88],"operation.":[89],"Leveraging":[90],"this":[91,114],"observation,":[92],"we":[93],"propose":[94],"<italic":[95],"xmlns:mml=\"http://www.w3.org/1998/Math/MathML\"":[96],"xmlns:xlink=\"http://www.w3.org/1999/xlink\">MaxiMoff</i>,":[97],"novel":[99],"architecture":[101],"utilize":[103],"cores":[105],"MAD":[110],"instructions":[111],"dynamically.":[112],"With":[113],"extended":[115],"design,":[116],"MaxiMoff":[117],"achieves":[118],"an":[119],"average":[120],"speedup":[121],"of":[122],"1.39\u00d7":[123],"reduces":[125],"total":[126],"energy":[127],"consumption":[128],"17.3%.":[130]},"counts_by_year":[],"updated_date":"2026-03-14T06:41:57.775601","created_date":"2025-11-04T00:00:00"}