{"id":"https://openalex.org/W7163512084","doi":"https://doi.org/10.23919/date69613.2026.11539457","title":"SSR: Sparse Segment Reduction for Ternary GEMM Acceleration","display_name":"SSR: Sparse Segment Reduction for Ternary GEMM Acceleration","publication_year":2026,"publication_date":"2026-04-20","ids":{"openalex":"https://openalex.org/W7163512084","doi":"https://doi.org/10.23919/date69613.2026.11539457"},"language":null,"primary_location":{"id":"doi:10.23919/date69613.2026.11539457","is_oa":false,"landing_page_url":"https://doi.org/10.23919/date69613.2026.11539457","pdf_url":null,"source":null,"license":null,"license_id":null,"version":"publishedVersion","is_accepted":true,"is_published":true,"raw_source_name":"2026 Design, Automation & Test in Europe Conference (DATE)","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/A5137865474","display_name":"Adeline Pittet","orcid":null},"institutions":[{"id":"https://openalex.org/I35440088","display_name":"ETH Zurich","ror":"https://ror.org/05a28rw58","country_code":"CH","type":"education","lineage":["https://openalex.org/I2799323385","https://openalex.org/I35440088"]},{"id":"https://openalex.org/I4210123994","display_name":"SystemsX.ch","ror":"https://ror.org/02s09gf16","country_code":"CH","type":"facility","lineage":["https://openalex.org/I4210123994"]}],"countries":["CH"],"is_corresponding":false,"raw_author_name":"Adeline Pittet","raw_affiliation_strings":["ETH Zurich,Systems Group,Switzerland"],"raw_orcid":null,"affiliations":[{"raw_affiliation_string":"ETH Zurich,Systems Group,Switzerland","institution_ids":["https://openalex.org/I35440088","https://openalex.org/I4210123994"]}]},{"author_position":"middle","author":{"id":"https://openalex.org/A5137814465","display_name":"Shien Zhu","orcid":null},"institutions":[{"id":"https://openalex.org/I35440088","display_name":"ETH Zurich","ror":"https://ror.org/05a28rw58","country_code":"CH","type":"education","lineage":["https://openalex.org/I2799323385","https://openalex.org/I35440088"]},{"id":"https://openalex.org/I4210123994","display_name":"SystemsX.ch","ror":"https://ror.org/02s09gf16","country_code":"CH","type":"facility","lineage":["https://openalex.org/I4210123994"]}],"countries":["CH"],"is_corresponding":false,"raw_author_name":"Shien Zhu","raw_affiliation_strings":["ETH Zurich,Systems Group,Switzerland"],"raw_orcid":null,"affiliations":[{"raw_affiliation_string":"ETH Zurich,Systems Group,Switzerland","institution_ids":["https://openalex.org/I35440088","https://openalex.org/I4210123994"]}]},{"author_position":"middle","author":{"id":"https://openalex.org/A5137862820","display_name":"Val\u00e9rie Verdan","orcid":null},"institutions":[{"id":"https://openalex.org/I35440088","display_name":"ETH Zurich","ror":"https://ror.org/05a28rw58","country_code":"CH","type":"education","lineage":["https://openalex.org/I2799323385","https://openalex.org/I35440088"]},{"id":"https://openalex.org/I4210123994","display_name":"SystemsX.ch","ror":"https://ror.org/02s09gf16","country_code":"CH","type":"facility","lineage":["https://openalex.org/I4210123994"]}],"countries":["CH"],"is_corresponding":false,"raw_author_name":"Val\u00e9rie Verdan","raw_affiliation_strings":["ETH Zurich,Systems Group,Switzerland"],"raw_orcid":null,"affiliations":[{"raw_affiliation_string":"ETH Zurich,Systems Group,Switzerland","institution_ids":["https://openalex.org/I35440088","https://openalex.org/I4210123994"]}]},{"author_position":"last","author":{"id":"https://openalex.org/A5056034917","display_name":"Gustavo Alonso","orcid":"https://orcid.org/0000-0002-1220-5586"},"institutions":[{"id":"https://openalex.org/I35440088","display_name":"ETH Zurich","ror":"https://ror.org/05a28rw58","country_code":"CH","type":"education","lineage":["https://openalex.org/I2799323385","https://openalex.org/I35440088"]},{"id":"https://openalex.org/I4210123994","display_name":"SystemsX.ch","ror":"https://ror.org/02s09gf16","country_code":"CH","type":"facility","lineage":["https://openalex.org/I4210123994"]}],"countries":["CH"],"is_corresponding":false,"raw_author_name":"Gustavo Alonso","raw_affiliation_strings":["ETH Zurich,Systems Group,Switzerland"],"raw_orcid":null,"affiliations":[{"raw_affiliation_string":"ETH Zurich,Systems Group,Switzerland","institution_ids":["https://openalex.org/I35440088","https://openalex.org/I4210123994"]}]}],"institutions":[],"countries_distinct_count":1,"institutions_distinct_count":4,"corresponding_author_ids":[],"corresponding_institution_ids":[],"apc_list":null,"apc_paid":null,"fwci":0.0,"has_fulltext":false,"cited_by_count":0,"citation_normalized_percentile":{"value":0.89557036,"is_in_top_1_percent":false,"is_in_top_10_percent":false},"cited_by_percentile_year":null,"biblio":{"volume":null,"issue":null,"first_page":"1","last_page":"7"},"is_retracted":false,"is_paratext":false,"is_xpac":false,"primary_topic":{"id":"https://openalex.org/T10048","display_name":"Particle physics theoretical and experimental studies","score":0.35269999504089355,"subfield":{"id":"https://openalex.org/subfields/3106","display_name":"Nuclear and High Energy Physics"},"field":{"id":"https://openalex.org/fields/31","display_name":"Physics and Astronomy"},"domain":{"id":"https://openalex.org/domains/3","display_name":"Physical Sciences"}},"topics":[{"id":"https://openalex.org/T10048","display_name":"Particle physics theoretical and experimental studies","score":0.35269999504089355,"subfield":{"id":"https://openalex.org/subfields/3106","display_name":"Nuclear and High Energy Physics"},"field":{"id":"https://openalex.org/fields/31","display_name":"Physics and Astronomy"},"domain":{"id":"https://openalex.org/domains/3","display_name":"Physical Sciences"}},{"id":"https://openalex.org/T10224","display_name":"Quantum Chromodynamics and Particle Interactions","score":0.0689999982714653,"subfield":{"id":"https://openalex.org/subfields/3106","display_name":"Nuclear and High Energy Physics"},"field":{"id":"https://openalex.org/fields/31","display_name":"Physics and Astronomy"},"domain":{"id":"https://openalex.org/domains/3","display_name":"Physical Sciences"}},{"id":"https://openalex.org/T11044","display_name":"Particle Detector Development and Performance","score":0.06780000030994415,"subfield":{"id":"https://openalex.org/subfields/3106","display_name":"Nuclear and High Energy Physics"},"field":{"id":"https://openalex.org/fields/31","display_name":"Physics and Astronomy"},"domain":{"id":"https://openalex.org/domains/3","display_name":"Physical Sciences"}}],"keywords":[{"id":"https://openalex.org/keywords/acceleration","display_name":"Acceleration","score":0.6331999897956848},{"id":"https://openalex.org/keywords/reduction","display_name":"Reduction (mathematics)","score":0.49939998984336853},{"id":"https://openalex.org/keywords/ternary-operation","display_name":"Ternary operation","score":0.44020000100135803},{"id":"https://openalex.org/keywords/feature","display_name":"Feature (linguistics)","score":0.33660000562667847},{"id":"https://openalex.org/keywords/minification","display_name":"Minification","score":0.29030001163482666}],"concepts":[{"id":"https://openalex.org/C117896860","wikidata":"https://www.wikidata.org/wiki/Q11376","display_name":"Acceleration","level":2,"score":0.6331999897956848},{"id":"https://openalex.org/C11413529","wikidata":"https://www.wikidata.org/wiki/Q8366","display_name":"Algorithm","level":1,"score":0.5950000286102295},{"id":"https://openalex.org/C111335779","wikidata":"https://www.wikidata.org/wiki/Q3454686","display_name":"Reduction (mathematics)","level":2,"score":0.49939998984336853},{"id":"https://openalex.org/C33923547","wikidata":"https://www.wikidata.org/wiki/Q395","display_name":"Mathematics","level":0,"score":0.4666000008583069},{"id":"https://openalex.org/C64452783","wikidata":"https://www.wikidata.org/wiki/Q1524945","display_name":"Ternary operation","level":2,"score":0.44020000100135803},{"id":"https://openalex.org/C41008148","wikidata":"https://www.wikidata.org/wiki/Q21198","display_name":"Computer science","level":0,"score":0.42309999465942383},{"id":"https://openalex.org/C28826006","wikidata":"https://www.wikidata.org/wiki/Q33521","display_name":"Applied mathematics","level":1,"score":0.3547999858856201},{"id":"https://openalex.org/C2776401178","wikidata":"https://www.wikidata.org/wiki/Q12050496","display_name":"Feature (linguistics)","level":2,"score":0.33660000562667847},{"id":"https://openalex.org/C147764199","wikidata":"https://www.wikidata.org/wiki/Q6865248","display_name":"Minification","level":2,"score":0.29030001163482666},{"id":"https://openalex.org/C2777735758","wikidata":"https://www.wikidata.org/wiki/Q817765","display_name":"Path (computing)","level":2,"score":0.2709999978542328},{"id":"https://openalex.org/C116834253","wikidata":"https://www.wikidata.org/wiki/Q2039217","display_name":"Identification (biology)","level":2,"score":0.25690001249313354},{"id":"https://openalex.org/C112972136","wikidata":"https://www.wikidata.org/wiki/Q7595718","display_name":"Stability (learning theory)","level":2,"score":0.251800000667572}],"mesh":[],"locations_count":1,"locations":[{"id":"doi:10.23919/date69613.2026.11539457","is_oa":false,"landing_page_url":"https://doi.org/10.23919/date69613.2026.11539457","pdf_url":null,"source":null,"license":null,"license_id":null,"version":"publishedVersion","is_accepted":true,"is_published":true,"raw_source_name":"2026 Design, Automation & Test in Europe Conference (DATE)","raw_type":"proceedings-article"}],"best_oa_location":null,"sustainable_development_goals":[],"awards":[],"funders":[],"has_content":{"grobid_xml":false,"pdf":false},"content_urls":null,"referenced_works_count":14,"referenced_works":["https://openalex.org/W2551895583","https://openalex.org/W2767737961","https://openalex.org/W2777372517","https://openalex.org/W2904725623","https://openalex.org/W2905006747","https://openalex.org/W2911109435","https://openalex.org/W2962676243","https://openalex.org/W2964132039","https://openalex.org/W3090912412","https://openalex.org/W3175903058","https://openalex.org/W4246554365","https://openalex.org/W4312934552","https://openalex.org/W4405935039","https://openalex.org/W4412945527"],"related_works":[],"abstract_inverted_index":{"Large":[0],"Language":[1],"Models":[2],"(LLMs)":[3],"require":[4],"substantial":[5],"computational":[6],"resources,":[7],"limiting":[8],"their":[9],"deployment":[10],"on":[11,157,176],"resource-constrained":[12],"hardware.":[13],"Ternary":[14,95],"LLMs":[15,92],"mitigate":[16],"these":[17],"demands":[18],"through":[19,115],"weight":[20],"quantization":[21],"via":[22],"ternary":[23,42,66,81,91,104,158],"values,":[24],"achieving":[25],"significant":[26],"compression":[27],"often":[28],"with":[29,120,127,160],"50":[30],"\u2212":[31],"90%":[32],"sparsity.":[33,122,162],"However,":[34],"existing":[35],"approaches":[36],"have":[37],"limitations:":[38],"methods":[39],"optimized":[40,103],"for":[41,133],"weights,":[43],"such":[44],"as":[45],"BitNet,":[46],"redundant":[47],"segment":[48],"reduction":[49],"(RSR),":[50],"and":[51,93,107,169],"its":[52],"improved":[53],"version":[54],"RSR++,":[55],"do":[56],"not":[57],"exploit":[58],"sparsity":[59,113,134,145],"structures,":[60],"while":[61,137],"conventional":[62],"sparse":[63],"formats":[64],"neglect":[65],"characteristics,":[67],"foregoing":[68],"dual":[69],"optimization":[70],"opportunities.In":[71],"this":[72],"paper,":[73],"we":[74],"introduce":[75],"Sparse":[76],"Segment":[77],"Reduction":[78],"(SSR),":[79],"a":[80,101],"matrix":[82],"multiplication":[83],"method":[84],"designed":[85],"to":[86],"accelerate":[87],"the":[88,121,177],"inference":[89,130],"of":[90,171],"general":[94],"Weight":[96],"Networks":[97],"(TWNs).":[98],"SSR":[99,123,151,164],"has":[100],"dedicated":[102],"data":[105],"format":[106],"an":[108],"algorithm":[109],"that":[110,118,150],"systematically":[111],"exploits":[112],"patterns":[114],"computation":[116],"trees":[117],"scale":[119],"provides":[124],"theoretical":[125],"gains":[126],"asymptotically":[128],"faster":[129],"than":[131],"RSR++":[132,156,175],"above":[135],"50%,":[136],"practical":[138],"evaluations":[139],"reveal":[140],"performance":[141],"improvements":[142],"across":[143],"all":[144],"levels.":[146],"Evaluation":[147],"results":[148],"show":[149],"achieves":[152,165],"2.1-11.3\u00d7":[153],"speedup":[154,168],"over":[155,174],"GEMM":[159],"45-95%":[161],"Furthermore,":[163],"3.5-6.3\u00d7":[166],"end-to-end":[167],"4.9%":[170],"memory":[172],"saving":[173],"Llama-3":[178],"1B":[179],"model":[180],"inference.":[181]},"counts_by_year":[],"updated_date":"2026-06-11T09:08:48.828518","created_date":"2026-06-05T00:00:00"}