{"id":"https://openalex.org/W4414603156","doi":"https://doi.org/10.1109/tcsi.2025.3610210","title":"TEA-SPS: A Tiny and Efficient Architecture for Softmax With Parallelism and Sparsity Adaptability","display_name":"TEA-SPS: A Tiny and Efficient Architecture for Softmax With Parallelism and Sparsity Adaptability","publication_year":2025,"publication_date":"2025-09-29","ids":{"openalex":"https://openalex.org/W4414603156","doi":"https://doi.org/10.1109/tcsi.2025.3610210"},"language":"en","primary_location":{"id":"doi:10.1109/tcsi.2025.3610210","is_oa":false,"landing_page_url":"https://doi.org/10.1109/tcsi.2025.3610210","pdf_url":null,"source":{"id":"https://openalex.org/S116977442","display_name":"IEEE Transactions on Circuits and Systems I Regular Papers","issn_l":"1549-8328","issn":["1549-8328","1558-0806"],"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 Circuits and Systems I: Regular Papers","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/A5119761187","display_name":"Zhanhao Cui","orcid":null},"institutions":[{"id":"https://openalex.org/I881766915","display_name":"Nanjing University","ror":"https://ror.org/01rxvg760","country_code":"CN","type":"education","lineage":["https://openalex.org/I881766915"]}],"countries":["CN"],"is_corresponding":false,"raw_author_name":"Zhanhao Cui","raw_affiliation_strings":["Nanjing University, Nanjing, China"],"raw_orcid":"https://orcid.org/0009-0007-4143-0575","affiliations":[{"raw_affiliation_string":"Nanjing University, Nanjing, China","institution_ids":["https://openalex.org/I881766915"]}]},{"author_position":"middle","author":{"id":"https://openalex.org/A5101447106","display_name":"Yuhao Ma","orcid":"https://orcid.org/0000-0002-3578-7822"},"institutions":[{"id":"https://openalex.org/I881766915","display_name":"Nanjing University","ror":"https://ror.org/01rxvg760","country_code":"CN","type":"education","lineage":["https://openalex.org/I881766915"]}],"countries":["CN"],"is_corresponding":false,"raw_author_name":"Yuhao Ma","raw_affiliation_strings":["Nanjing University, Nanjing, China"],"raw_orcid":null,"affiliations":[{"raw_affiliation_string":"Nanjing University, Nanjing, China","institution_ids":["https://openalex.org/I881766915"]}]},{"author_position":"middle","author":{"id":"https://openalex.org/A5112749177","display_name":"Yuan Cheng","orcid":"https://orcid.org/0009-0001-7591-0418"},"institutions":[{"id":"https://openalex.org/I881766915","display_name":"Nanjing University","ror":"https://ror.org/01rxvg760","country_code":"CN","type":"education","lineage":["https://openalex.org/I881766915"]}],"countries":["CN"],"is_corresponding":false,"raw_author_name":"Yuan Cheng","raw_affiliation_strings":["Nanjing University, Nanjing, China"],"raw_orcid":null,"affiliations":[{"raw_affiliation_string":"Nanjing University, Nanjing, China","institution_ids":["https://openalex.org/I881766915"]}]},{"author_position":"middle","author":{"id":"https://openalex.org/A5020727498","display_name":"Zhengyu Mei","orcid":"https://orcid.org/0000-0003-0964-3507"},"institutions":[{"id":"https://openalex.org/I4210153300","display_name":"Shanghai Research Center for Wireless Communications","ror":"https://ror.org/04xfd0826","country_code":"CN","type":"facility","lineage":["https://openalex.org/I4210153300"]}],"countries":["CN"],"is_corresponding":false,"raw_author_name":"Zhengyu Mei","raw_affiliation_strings":["Wireless Terminal Architecture and Design Department, HS, Shanghai, China"],"raw_orcid":"https://orcid.org/0000-0003-0964-3507","affiliations":[{"raw_affiliation_string":"Wireless Terminal Architecture and Design Department, HS, Shanghai, China","institution_ids":["https://openalex.org/I4210153300"]}]},{"author_position":"middle","author":{"id":"https://openalex.org/A5010044441","display_name":"Feng Yan","orcid":"https://orcid.org/0000-0001-7385-6334"},"institutions":[{"id":"https://openalex.org/I881766915","display_name":"Nanjing University","ror":"https://ror.org/01rxvg760","country_code":"CN","type":"education","lineage":["https://openalex.org/I881766915"]}],"countries":["CN"],"is_corresponding":false,"raw_author_name":"Feng Yan","raw_affiliation_strings":["Nanjing University, Nanjing, China"],"raw_orcid":null,"affiliations":[{"raw_affiliation_string":"Nanjing University, Nanjing, China","institution_ids":["https://openalex.org/I881766915"]}]},{"author_position":"middle","author":{"id":"https://openalex.org/A5027952470","display_name":"Hongbing Pan","orcid":"https://orcid.org/0000-0002-7181-8278"},"institutions":[{"id":"https://openalex.org/I881766915","display_name":"Nanjing University","ror":"https://ror.org/01rxvg760","country_code":"CN","type":"education","lineage":["https://openalex.org/I881766915"]}],"countries":["CN"],"is_corresponding":false,"raw_author_name":"Hongbing Pan","raw_affiliation_strings":["Nanjing University, Nanjing, China"],"raw_orcid":"https://orcid.org/0000-0002-7181-8278","affiliations":[{"raw_affiliation_string":"Nanjing University, Nanjing, China","institution_ids":["https://openalex.org/I881766915"]}]},{"author_position":"last","author":{"id":"https://openalex.org/A5031435494","display_name":"Yuxuan Wang","orcid":"https://orcid.org/0000-0001-9981-0588"},"institutions":[{"id":"https://openalex.org/I881766915","display_name":"Nanjing University","ror":"https://ror.org/01rxvg760","country_code":"CN","type":"education","lineage":["https://openalex.org/I881766915"]}],"countries":["CN"],"is_corresponding":false,"raw_author_name":"Yuxuan Wang","raw_affiliation_strings":["Nanjing University, Nanjing, China"],"raw_orcid":"https://orcid.org/0000-0001-9981-0588","affiliations":[{"raw_affiliation_string":"Nanjing University, Nanjing, China","institution_ids":["https://openalex.org/I881766915"]}]}],"institutions":[],"countries_distinct_count":1,"institutions_distinct_count":2,"corresponding_author_ids":[],"corresponding_institution_ids":[],"apc_list":null,"apc_paid":null,"fwci":1.377,"has_fulltext":false,"cited_by_count":1,"citation_normalized_percentile":{"value":0.82350357,"is_in_top_1_percent":false,"is_in_top_10_percent":false},"cited_by_percentile_year":{"min":94,"max":97},"biblio":{"volume":"73","issue":"2","first_page":"1245","last_page":"1258"},"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/T10715","display_name":"Distributed and Parallel Computing Systems","score":0.996999979019165,"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/T10829","display_name":"Interconnection Networks and Systems","score":0.9847999811172485,"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/softmax-function","display_name":"Softmax function","score":0.852400004863739},{"id":"https://openalex.org/keywords/data-parallelism","display_name":"Data parallelism","score":0.5109999775886536},{"id":"https://openalex.org/keywords/throughput","display_name":"Throughput","score":0.5059000253677368},{"id":"https://openalex.org/keywords/adaptability","display_name":"Adaptability","score":0.42149999737739563},{"id":"https://openalex.org/keywords/piecewise","display_name":"Piecewise","score":0.4106999933719635},{"id":"https://openalex.org/keywords/masking","display_name":"Masking (illustration)","score":0.3644999861717224},{"id":"https://openalex.org/keywords/parallelism","display_name":"Parallelism (grammar)","score":0.3544999957084656},{"id":"https://openalex.org/keywords/design-space-exploration","display_name":"Design space exploration","score":0.3319000005722046}],"concepts":[{"id":"https://openalex.org/C188441871","wikidata":"https://www.wikidata.org/wiki/Q7554146","display_name":"Softmax function","level":3,"score":0.852400004863739},{"id":"https://openalex.org/C41008148","wikidata":"https://www.wikidata.org/wiki/Q21198","display_name":"Computer science","level":0,"score":0.7516999840736389},{"id":"https://openalex.org/C61483411","wikidata":"https://www.wikidata.org/wiki/Q3124522","display_name":"Data parallelism","level":3,"score":0.5109999775886536},{"id":"https://openalex.org/C157764524","wikidata":"https://www.wikidata.org/wiki/Q1383412","display_name":"Throughput","level":3,"score":0.5059000253677368},{"id":"https://openalex.org/C173608175","wikidata":"https://www.wikidata.org/wiki/Q232661","display_name":"Parallel computing","level":1,"score":0.4763000011444092},{"id":"https://openalex.org/C177606310","wikidata":"https://www.wikidata.org/wiki/Q5674297","display_name":"Adaptability","level":2,"score":0.42149999737739563},{"id":"https://openalex.org/C9390403","wikidata":"https://www.wikidata.org/wiki/Q3966","display_name":"Computer hardware","level":1,"score":0.4185999929904938},{"id":"https://openalex.org/C164660894","wikidata":"https://www.wikidata.org/wiki/Q2037833","display_name":"Piecewise","level":2,"score":0.4106999933719635},{"id":"https://openalex.org/C2777402240","wikidata":"https://www.wikidata.org/wiki/Q6783436","display_name":"Masking (illustration)","level":2,"score":0.3644999861717224},{"id":"https://openalex.org/C2781172179","wikidata":"https://www.wikidata.org/wiki/Q853109","display_name":"Parallelism (grammar)","level":2,"score":0.3544999957084656},{"id":"https://openalex.org/C113775141","wikidata":"https://www.wikidata.org/wiki/Q428691","display_name":"Computer engineering","level":1,"score":0.349700003862381},{"id":"https://openalex.org/C118524514","wikidata":"https://www.wikidata.org/wiki/Q173212","display_name":"Computer architecture","level":1,"score":0.34209999442100525},{"id":"https://openalex.org/C2776221188","wikidata":"https://www.wikidata.org/wiki/Q21072556","display_name":"Design space exploration","level":2,"score":0.3319000005722046},{"id":"https://openalex.org/C65232700","wikidata":"https://www.wikidata.org/wiki/Q5656403","display_name":"Hardware architecture","level":3,"score":0.3246000111103058},{"id":"https://openalex.org/C11413529","wikidata":"https://www.wikidata.org/wiki/Q8366","display_name":"Algorithm","level":1,"score":0.31049999594688416},{"id":"https://openalex.org/C163258240","wikidata":"https://www.wikidata.org/wiki/Q25342","display_name":"Power (physics)","level":2,"score":0.3082999885082245},{"id":"https://openalex.org/C106515295","wikidata":"https://www.wikidata.org/wiki/Q26806595","display_name":"Parallel processing","level":2,"score":0.3012000024318695},{"id":"https://openalex.org/C177264268","wikidata":"https://www.wikidata.org/wiki/Q1514741","display_name":"Set (abstract data type)","level":2,"score":0.2838999927043915},{"id":"https://openalex.org/C37139622","wikidata":"https://www.wikidata.org/wiki/Q386953","display_name":"Cellular architecture","level":5,"score":0.28049999475479126},{"id":"https://openalex.org/C123657996","wikidata":"https://www.wikidata.org/wiki/Q12271","display_name":"Architecture","level":2,"score":0.2791999876499176},{"id":"https://openalex.org/C45347329","wikidata":"https://www.wikidata.org/wiki/Q5166604","display_name":"Convolution (computer science)","level":3,"score":0.2565999925136566},{"id":"https://openalex.org/C104250799","wikidata":"https://www.wikidata.org/wiki/Q1780765","display_name":"Companding","level":4,"score":0.25279998779296875},{"id":"https://openalex.org/C177148314","wikidata":"https://www.wikidata.org/wiki/Q170084","display_name":"Generalization","level":2,"score":0.2524000108242035},{"id":"https://openalex.org/C24858836","wikidata":"https://www.wikidata.org/wiki/Q844718","display_name":"Theory of computation","level":2,"score":0.25220000743865967}],"mesh":[],"locations_count":1,"locations":[{"id":"doi:10.1109/tcsi.2025.3610210","is_oa":false,"landing_page_url":"https://doi.org/10.1109/tcsi.2025.3610210","pdf_url":null,"source":{"id":"https://openalex.org/S116977442","display_name":"IEEE Transactions on Circuits and Systems I Regular Papers","issn_l":"1549-8328","issn":["1549-8328","1558-0806"],"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 Circuits and Systems I: Regular Papers","raw_type":"journal-article"}],"best_oa_location":null,"sustainable_development_goals":[],"awards":[{"id":"https://openalex.org/G6122486374","display_name":null,"funder_award_id":"BE2023019-1","funder_id":"https://openalex.org/F4320327780","funder_display_name":"Key Research and Development Program of Jiangxi Province"}],"funders":[{"id":"https://openalex.org/F4320327780","display_name":"Key Research and Development Program of Jiangxi Province","ror":null}],"has_content":{"grobid_xml":false,"pdf":false},"content_urls":null,"referenced_works_count":27,"referenced_works":["https://openalex.org/W2608093348","https://openalex.org/W2906917284","https://openalex.org/W2910396952","https://openalex.org/W2946480111","https://openalex.org/W2970777192","https://openalex.org/W2979826702","https://openalex.org/W3036061420","https://openalex.org/W3043675702","https://openalex.org/W3159727696","https://openalex.org/W3172230466","https://openalex.org/W3183327882","https://openalex.org/W3207366647","https://openalex.org/W3211525823","https://openalex.org/W4206087890","https://openalex.org/W4286571858","https://openalex.org/W4293768835","https://openalex.org/W4313467238","https://openalex.org/W4364302069","https://openalex.org/W4384947491","https://openalex.org/W4385245566","https://openalex.org/W4387789586","https://openalex.org/W4388145652","https://openalex.org/W4394627484","https://openalex.org/W4402349404","https://openalex.org/W4404954358","https://openalex.org/W4411725376","https://openalex.org/W4412094103"],"related_works":[],"abstract_inverted_index":{"With":[0],"the":[1,34,48,86,102,106,118,123,144,158,167,191,210,233,245],"remarkable":[2],"performance":[3],"of":[4,36,38,88,112,126,136,148,160,170,193,213,224,242],"Transformer-based":[5],"networks":[6],"in":[7,29,128],"multiple":[8],"fields":[9],"and":[10,40,60,78,97,109,138,209],"increasing":[11],"demand":[12],"for":[13,23,46,51,221],"computational":[14,107],"resources":[15],"by":[16],"softmax":[17,28,66,72],"within":[18],"them,":[19,143],"it":[20],"is":[21,55,131,150],"inevitable":[22],"hardware":[24,49,61,139,146],"accelerators":[25],"to":[26,33,64,104,133,179,232],"support":[27,220],"Transformer.":[30],"However,":[31],"due":[32],"lack":[35],"co-design":[37],"algorithm":[39,59],"hardware,":[41],"there":[42],"still":[43],"remains":[44],"space":[45,125],"optimizing":[47],"architecture":[50,63,147],"softmax.":[52],"Therefore,":[53],"TEA-SPS":[54,149,171,218],"proposed":[56,132],"as":[57],"an":[58,238],"co-designed":[62],"improve":[65],"with":[67,73,190,227],"two":[68],"methods:":[69],"Configurable":[70],"Parallelism":[71],"Sparse":[74],"mask":[75],"Strategy":[76],"(CPSS)":[77],"Specific":[79],"Piecewise":[80],"Information":[81],"Extractor":[82],"(SPIE).":[83],"CPSS":[84],"has":[85],"advantage":[87],"supporting":[89],"different":[90],"throughput":[91],"requirements":[92],"through":[93],"configurable":[94],"data-level":[95],"parallelism":[96],"performing":[98],"sparse":[99,240],"masking":[100],"on":[101,142,244],"outputs":[103],"reduce":[105],"load":[108],"memory":[110],"access":[111],"subsequent":[113],"operations.":[114],"To":[115],"further":[116],"explore":[117],"optimal":[119],"solution":[120],"set":[121],"among":[122],"design":[124],"parameters":[127],"CPSS,":[129],"SPIE":[130],"achieve":[134],"co-optimization":[135],"accuracy":[137,229],"overhead.":[140],"Based":[141],"efficient":[145],"proposed.":[151],"The":[152],"implementation":[153],"results":[154],"show":[155],"that":[156],"at":[157],"frequency":[159],"0.5":[161],"GHz":[162],"under":[163],"TSMC":[164],"90-nm":[165],"technology,":[166],"peak":[168],"efficiency":[169],"processing":[172],"8-bit":[173],"quantized":[174,234],"data":[175],"can":[176],"reach":[177],"up":[178],"216.97":[180],"Gps/(mm<inline-formula":[181],"xmlns:mml=\"http://www.w3.org/1998/Math/MathML\"":[182,195,203],"xmlns:xlink=\"http://www.w3.org/1999/xlink\">":[183,196,204],"<tex-math":[184,197,205],"notation=\"LaTeX\">$\\boldsymbol":[185,206],"{^{2}\\cdot":[186],"}$</tex-math>":[187,200],"</inline-formula>":[188,201,208],"mW),":[189],"area":[192],"<inline-formula":[194],"notation=\"LaTeX\">$3290.21~\\boldsymbol":[198],"{\\mu":[199],"m<inline-formula":[202],"{^{2}}$</tex-math>":[207],"power":[211],"consumption":[212],"0.7004":[214],"mW.":[215],"In":[216],"addition,":[217],"provides":[219],"input":[222],"sequences":[223],"arbitrary":[225],"length":[226],"negligible":[228],"loss":[230],"compared":[231],"baseline,":[235],"while":[236],"achieving":[237],"average":[239],"rate":[241],"68.6%":[243],"GLUE":[246],"tasks.":[247]},"counts_by_year":[{"year":2026,"cited_by_count":1}],"updated_date":"2026-02-01T03:34:12.195049","created_date":"2025-10-10T00:00:00"}
