{"id":"https://openalex.org/W4399487278","doi":"https://doi.org/10.1145/3649476.3658709","title":"HSCONN: Hardware-Software Co-Optimization of Self-Attention Neural Networks for Large Language Models","display_name":"HSCONN: Hardware-Software Co-Optimization of Self-Attention Neural Networks for Large Language Models","publication_year":2024,"publication_date":"2024-06-10","ids":{"openalex":"https://openalex.org/W4399487278","doi":"https://doi.org/10.1145/3649476.3658709"},"language":"en","primary_location":{"id":"doi:10.1145/3649476.3658709","is_oa":false,"landing_page_url":"https://doi.org/10.1145/3649476.3658709","pdf_url":null,"source":null,"license":null,"license_id":null,"version":"publishedVersion","is_accepted":true,"is_published":true,"raw_source_name":"Proceedings of the Great Lakes Symposium on VLSI 2024","raw_type":"proceedings-article"},"type":"conference-paper","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/A5072756161","display_name":"Siqin Liu","orcid":"https://orcid.org/0000-0002-0565-5845"},"institutions":[{"id":"https://openalex.org/I4210106879","display_name":"Ohio University","ror":"https://ror.org/01jr3y717","country_code":"US","type":"education","lineage":["https://openalex.org/I4210106879"]}],"countries":["US"],"is_corresponding":false,"raw_author_name":"Siqin Liu","raw_affiliation_strings":["Electrical Engineering and Computer Science, Ohio University, USA"],"raw_orcid":"https://orcid.org/0000-0002-0565-5845","affiliations":[{"raw_affiliation_string":"Electrical Engineering and Computer Science, Ohio University, USA","institution_ids":["https://openalex.org/I4210106879"]}]},{"author_position":"middle","author":{"id":null,"display_name":"Prakash Chand Kuve","orcid":"https://orcid.org/0000-0001-6827-3169"},"institutions":[],"countries":[],"is_corresponding":false,"raw_author_name":"Prakash Chand Kuve","raw_affiliation_strings":["Microchip Technology Corporation, India"],"raw_orcid":"https://orcid.org/0000-0001-6827-3169","affiliations":[{"raw_affiliation_string":"Microchip Technology Corporation, India","institution_ids":[]}]},{"author_position":"last","author":{"id":"https://openalex.org/A5014895124","display_name":"Avinash Karanth","orcid":"https://orcid.org/0000-0002-9472-4637"},"institutions":[{"id":"https://openalex.org/I4210106879","display_name":"Ohio University","ror":"https://ror.org/01jr3y717","country_code":"US","type":"education","lineage":["https://openalex.org/I4210106879"]}],"countries":["US"],"is_corresponding":false,"raw_author_name":"Avinash Karanth","raw_affiliation_strings":["Ohio University, USA"],"raw_orcid":"https://orcid.org/0000-0002-9472-4637","affiliations":[{"raw_affiliation_string":"Ohio University, USA","institution_ids":["https://openalex.org/I4210106879"]}]}],"institutions":[],"countries_distinct_count":1,"institutions_distinct_count":1,"corresponding_author_ids":[],"corresponding_institution_ids":[],"apc_list":null,"apc_paid":null,"fwci":0.9755,"has_fulltext":false,"cited_by_count":3,"citation_normalized_percentile":{"value":0.75289215,"is_in_top_1_percent":false,"is_in_top_10_percent":false},"cited_by_percentile_year":{"min":91,"max":98},"biblio":{"volume":null,"issue":null,"first_page":"736","last_page":"741"},"is_retracted":false,"is_paratext":false,"is_xpac":false,"primary_topic":{"id":"https://openalex.org/T10028","display_name":"Topic Modeling","score":0.9979000091552734,"subfield":{"id":"https://openalex.org/subfields/1702","display_name":"Artificial Intelligence"},"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/T10028","display_name":"Topic Modeling","score":0.9979000091552734,"subfield":{"id":"https://openalex.org/subfields/1702","display_name":"Artificial Intelligence"},"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/T10036","display_name":"Advanced Neural Network Applications","score":0.9922000169754028,"subfield":{"id":"https://openalex.org/subfields/1707","display_name":"Computer Vision and Pattern Recognition"},"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/T11948","display_name":"Machine Learning in Materials Science","score":0.9864000082015991,"subfield":{"id":"https://openalex.org/subfields/2505","display_name":"Materials Chemistry"},"field":{"id":"https://openalex.org/fields/25","display_name":"Materials Science"},"domain":{"id":"https://openalex.org/domains/3","display_name":"Physical Sciences"}}],"keywords":[{"id":"https://openalex.org/keywords/computer-science","display_name":"Computer science","score":0.8106253743171692},{"id":"https://openalex.org/keywords/artificial-neural-network","display_name":"Artificial neural network","score":0.5498181581497192},{"id":"https://openalex.org/keywords/software","display_name":"Software","score":0.5460209846496582},{"id":"https://openalex.org/keywords/computer-architecture","display_name":"Computer architecture","score":0.3609035909175873},{"id":"https://openalex.org/keywords/embedded-system","display_name":"Embedded system","score":0.3245961666107178},{"id":"https://openalex.org/keywords/artificial-intelligence","display_name":"Artificial intelligence","score":0.26365959644317627},{"id":"https://openalex.org/keywords/programming-language","display_name":"Programming language","score":0.26087334752082825}],"concepts":[{"id":"https://openalex.org/C41008148","wikidata":"https://www.wikidata.org/wiki/Q21198","display_name":"Computer science","level":0,"score":0.8106253743171692},{"id":"https://openalex.org/C50644808","wikidata":"https://www.wikidata.org/wiki/Q192776","display_name":"Artificial neural network","level":2,"score":0.5498181581497192},{"id":"https://openalex.org/C2777904410","wikidata":"https://www.wikidata.org/wiki/Q7397","display_name":"Software","level":2,"score":0.5460209846496582},{"id":"https://openalex.org/C118524514","wikidata":"https://www.wikidata.org/wiki/Q173212","display_name":"Computer architecture","level":1,"score":0.3609035909175873},{"id":"https://openalex.org/C149635348","wikidata":"https://www.wikidata.org/wiki/Q193040","display_name":"Embedded system","level":1,"score":0.3245961666107178},{"id":"https://openalex.org/C154945302","wikidata":"https://www.wikidata.org/wiki/Q11660","display_name":"Artificial intelligence","level":1,"score":0.26365959644317627},{"id":"https://openalex.org/C199360897","wikidata":"https://www.wikidata.org/wiki/Q9143","display_name":"Programming language","level":1,"score":0.26087334752082825}],"mesh":[],"locations_count":1,"locations":[{"id":"doi:10.1145/3649476.3658709","is_oa":false,"landing_page_url":"https://doi.org/10.1145/3649476.3658709","pdf_url":null,"source":null,"license":null,"license_id":null,"version":"publishedVersion","is_accepted":true,"is_published":true,"raw_source_name":"Proceedings of the Great Lakes Symposium on VLSI 2024","raw_type":"proceedings-article"}],"best_oa_location":null,"sustainable_development_goals":[],"awards":[{"id":"https://openalex.org/G2971260893","display_name":null,"funder_award_id":"CCF- 1703013, CCF-1901192, CCF-1936794, CCF-2324645 and CCF-2311544","funder_id":"https://openalex.org/F4320323817","funder_display_name":"Universitas Brawijaya"}],"funders":[{"id":"https://openalex.org/F4320323817","display_name":"Universitas Brawijaya","ror":"https://ror.org/01wk3d929"}],"has_content":{"pdf":false,"grobid_xml":false},"content_urls":null,"referenced_works_count":11,"referenced_works":["https://openalex.org/W2973657294","https://openalex.org/W2995607652","https://openalex.org/W2998617917","https://openalex.org/W3017024317","https://openalex.org/W3037110041","https://openalex.org/W3097660259","https://openalex.org/W3131922516","https://openalex.org/W3159727696","https://openalex.org/W3206453033","https://openalex.org/W4206943884","https://openalex.org/W6778883912"],"related_works":["https://openalex.org/W4391375266","https://openalex.org/W2748952813","https://openalex.org/W2390279801","https://openalex.org/W2358668433","https://openalex.org/W4396701345","https://openalex.org/W2376932109","https://openalex.org/W2001405890","https://openalex.org/W4396696052","https://openalex.org/W2382290278","https://openalex.org/W4395014643"],"abstract_inverted_index":{"Self-attention":[0],"models":[1],"excel":[2],"in":[3,80,113],"natural":[4,114],"language":[5,115],"processing":[6,99,116],"and":[7,26,64,71,94,108,125,136,143,155],"computer":[8],"vision":[9],"by":[10],"capturing":[11],"contextual":[12],"information":[13],"but":[14],"encounter":[15],"several":[16],"challenges":[17],"such":[18],"as":[19,34],"efficient":[20],"data":[21,46],"movement,":[22],"quadratic":[23],"computational":[24,63],"complexity,":[25],"excessive":[27],"memory":[28,65,106],"accesses.":[29],"Sparse":[30],"attention":[31],"techniques":[32],"emerge":[33],"a":[35,90],"solution,":[36],"however,":[37],"their":[38,49],"irregular":[39],"or":[40],"regular":[41],"patterns,":[42],"coupled":[43],"with":[44,76],"costly":[45],"pre-processing,":[47],"diminish":[48],"hardware":[50,58],"efficiency.":[51,86],"This":[52],"paper":[53],"introduces":[54],"HSCONN,":[55],"an":[56],"energy-efficient":[57],"accelerator":[59],"for":[60,98],"self-attention,":[61],"mitigating":[62],"overheads.":[66],"HSCONN":[67,145],"employs":[68],"dynamic":[69,78],"voltage":[70],"frequency":[72],"scaling":[73],"(DVFS)":[74],"along":[75],"exploiting":[77,101],"sparsity":[79,103],"matrix":[81],"multiplication,":[82],"thereby":[83],"optimizing":[84],"energy":[85,110,126,157],"The":[87],"approach":[88],"includes":[89],"row-wise":[91],"pruning":[92],"algorithm":[93],"independent":[95],"voltage/frequency":[96],"islands":[97],"elements,":[100],"additional":[102],"to":[104,129,140],"reduce":[105],"access":[107],"overall":[109],"consumption.":[111],"Experiments":[112],"showcase":[117],"HSCONN\u2019s":[118],"remarkable":[119],"speedups":[120],"(1952":[121],"\u00d7,":[122,131,150,152,160,162],"615":[123],"\u00d7)":[124,133,154],"reductions":[127],"(up":[128],"820":[130],"113":[132],"over":[134],"CPU":[135],"GPU":[137],"architectures.":[138],"Compared":[139],"A3,":[141],"SpAtten,":[142],"Sanger,":[144],"demonstrates":[146],"superior":[147],"speedup":[148],"(1.71":[149],"1.25":[151],"1.47":[153],"higher":[156],"efficiency":[158],"(1.5":[159],"1.7":[161],"1.4":[163],"\u00d7).":[164]},"counts_by_year":[{"year":2026,"cited_by_count":2},{"year":2025,"cited_by_count":1}],"updated_date":"2026-08-21T09:56:20.448147","created_date":"2025-10-10T00:00:00"}