{"id":"https://openalex.org/W4386764151","doi":"https://doi.org/10.1109/dac56929.2023.10247855","title":"Flex-SFU: Accelerating DNN Activation Functions by Non-Uniform Piecewise Approximation","display_name":"Flex-SFU: Accelerating DNN Activation Functions by Non-Uniform Piecewise Approximation","publication_year":2023,"publication_date":"2023-07-09","ids":{"openalex":"https://openalex.org/W4386764151","doi":"https://doi.org/10.1109/dac56929.2023.10247855"},"language":"en","primary_location":{"id":"doi:10.1109/dac56929.2023.10247855","is_oa":false,"landing_page_url":"https://doi.org/10.1109/dac56929.2023.10247855","pdf_url":null,"source":null,"license":null,"license_id":null,"version":"publishedVersion","is_accepted":true,"is_published":true,"raw_source_name":"2023 60th ACM/IEEE Design Automation Conference (DAC)","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/A5064165125","display_name":"Enrico Reggiani","orcid":"https://orcid.org/0000-0003-1385-7962"},"institutions":[{"id":"https://openalex.org/I2250955327","display_name":"Huawei Technologies (China)","ror":"https://ror.org/00cmhce21","country_code":"CN","type":"company","lineage":["https://openalex.org/I2250955327"]},{"id":"https://openalex.org/I2799803557","display_name":"Barcelona Supercomputing Center","ror":"https://ror.org/05sd8tv96","country_code":"ES","type":"facility","lineage":["https://openalex.org/I2799803557","https://openalex.org/I9617848"]},{"id":"https://openalex.org/I9617848","display_name":"Universitat Polit\u00e8cnica de Catalunya","ror":"https://ror.org/03mb6wj31","country_code":"ES","type":"education","lineage":["https://openalex.org/I9617848"]}],"countries":["CN","ES"],"is_corresponding":false,"raw_author_name":"Enrico Reggiani","raw_affiliation_strings":["Huawei Zurich Research Center,Computing Systems Lab,Switzerland","Computing Systems Lab, Huawei Zurich Research Center, Switzerland","Barcelona Supercomputing Center, Spain"],"raw_orcid":null,"affiliations":[{"raw_affiliation_string":"Huawei Zurich Research Center,Computing Systems Lab,Switzerland","institution_ids":["https://openalex.org/I2250955327"]},{"raw_affiliation_string":"Computing Systems Lab, Huawei Zurich Research Center, Switzerland","institution_ids":[]},{"raw_affiliation_string":"Barcelona Supercomputing Center, Spain","institution_ids":["https://openalex.org/I9617848","https://openalex.org/I2799803557"]}]},{"author_position":"middle","author":{"id":"https://openalex.org/A5050957615","display_name":"Renzo Andri","orcid":"https://orcid.org/0000-0002-8776-5158"},"institutions":[{"id":"https://openalex.org/I2250955327","display_name":"Huawei Technologies (China)","ror":"https://ror.org/00cmhce21","country_code":"CN","type":"company","lineage":["https://openalex.org/I2250955327"]}],"countries":["CN"],"is_corresponding":false,"raw_author_name":"Renzo Andri","raw_affiliation_strings":["Huawei Zurich Research Center,Computing Systems Lab,Switzerland","Computing Systems Lab, Huawei Zurich Research Center, Switzerland"],"raw_orcid":null,"affiliations":[{"raw_affiliation_string":"Huawei Zurich Research Center,Computing Systems Lab,Switzerland","institution_ids":["https://openalex.org/I2250955327"]},{"raw_affiliation_string":"Computing Systems Lab, Huawei Zurich Research Center, Switzerland","institution_ids":[]}]},{"author_position":"last","author":{"id":"https://openalex.org/A5025399641","display_name":"Lukas Cavigelli","orcid":"https://orcid.org/0000-0003-1767-7715"},"institutions":[{"id":"https://openalex.org/I2250955327","display_name":"Huawei Technologies (China)","ror":"https://ror.org/00cmhce21","country_code":"CN","type":"company","lineage":["https://openalex.org/I2250955327"]}],"countries":["CN"],"is_corresponding":false,"raw_author_name":"Lukas Cavigelli","raw_affiliation_strings":["Huawei Zurich Research Center,Computing Systems Lab,Switzerland","Computing Systems Lab, Huawei Zurich Research Center, Switzerland"],"raw_orcid":null,"affiliations":[{"raw_affiliation_string":"Huawei Zurich Research Center,Computing Systems Lab,Switzerland","institution_ids":["https://openalex.org/I2250955327"]},{"raw_affiliation_string":"Computing Systems Lab, Huawei Zurich Research Center, Switzerland","institution_ids":[]}]}],"institutions":[],"countries_distinct_count":2,"institutions_distinct_count":3,"corresponding_author_ids":[],"corresponding_institution_ids":[],"apc_list":null,"apc_paid":null,"fwci":1.7964,"has_fulltext":false,"cited_by_count":16,"citation_normalized_percentile":{"value":0.87521662,"is_in_top_1_percent":false,"is_in_top_10_percent":false},"cited_by_percentile_year":{"min":94,"max":99},"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/T10036","display_name":"Advanced Neural Network Applications","score":0.9998999834060669,"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"}},"topics":[{"id":"https://openalex.org/T10036","display_name":"Advanced Neural Network Applications","score":0.9998999834060669,"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/T10054","display_name":"Parallel Computing and Optimization Techniques","score":0.9993000030517578,"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/T12808","display_name":"Ferroelectric and Negative Capacitance Devices","score":0.9980000257492065,"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/interpolation","display_name":"Interpolation (computer graphics)","score":0.7200688719749451},{"id":"https://openalex.org/keywords/computer-science","display_name":"Computer science","score":0.6460756063461304},{"id":"https://openalex.org/keywords/piecewise","display_name":"Piecewise","score":0.6268621683120728},{"id":"https://openalex.org/keywords/floating-point","display_name":"Floating point","score":0.48415452241897583},{"id":"https://openalex.org/keywords/algorithm","display_name":"Algorithm","score":0.4816835820674896},{"id":"https://openalex.org/keywords/approximation-error","display_name":"Approximation error","score":0.4574735164642334},{"id":"https://openalex.org/keywords/heuristics","display_name":"Heuristics","score":0.4476003646850586},{"id":"https://openalex.org/keywords/speedup","display_name":"Speedup","score":0.4471323490142822},{"id":"https://openalex.org/keywords/overhead","display_name":"Overhead (engineering)","score":0.4420841336250305},{"id":"https://openalex.org/keywords/parallel-computing","display_name":"Parallel computing","score":0.23930037021636963},{"id":"https://openalex.org/keywords/mathematics","display_name":"Mathematics","score":0.22390460968017578},{"id":"https://openalex.org/keywords/artificial-intelligence","display_name":"Artificial intelligence","score":0.1758694350719452}],"concepts":[{"id":"https://openalex.org/C137800194","wikidata":"https://www.wikidata.org/wiki/Q11713455","display_name":"Interpolation (computer graphics)","level":3,"score":0.7200688719749451},{"id":"https://openalex.org/C41008148","wikidata":"https://www.wikidata.org/wiki/Q21198","display_name":"Computer science","level":0,"score":0.6460756063461304},{"id":"https://openalex.org/C164660894","wikidata":"https://www.wikidata.org/wiki/Q2037833","display_name":"Piecewise","level":2,"score":0.6268621683120728},{"id":"https://openalex.org/C84211073","wikidata":"https://www.wikidata.org/wiki/Q117879","display_name":"Floating point","level":2,"score":0.48415452241897583},{"id":"https://openalex.org/C11413529","wikidata":"https://www.wikidata.org/wiki/Q8366","display_name":"Algorithm","level":1,"score":0.4816835820674896},{"id":"https://openalex.org/C122383733","wikidata":"https://www.wikidata.org/wiki/Q865920","display_name":"Approximation error","level":2,"score":0.4574735164642334},{"id":"https://openalex.org/C127705205","wikidata":"https://www.wikidata.org/wiki/Q5748245","display_name":"Heuristics","level":2,"score":0.4476003646850586},{"id":"https://openalex.org/C68339613","wikidata":"https://www.wikidata.org/wiki/Q1549489","display_name":"Speedup","level":2,"score":0.4471323490142822},{"id":"https://openalex.org/C2779960059","wikidata":"https://www.wikidata.org/wiki/Q7113681","display_name":"Overhead (engineering)","level":2,"score":0.4420841336250305},{"id":"https://openalex.org/C173608175","wikidata":"https://www.wikidata.org/wiki/Q232661","display_name":"Parallel computing","level":1,"score":0.23930037021636963},{"id":"https://openalex.org/C33923547","wikidata":"https://www.wikidata.org/wiki/Q395","display_name":"Mathematics","level":0,"score":0.22390460968017578},{"id":"https://openalex.org/C154945302","wikidata":"https://www.wikidata.org/wiki/Q11660","display_name":"Artificial intelligence","level":1,"score":0.1758694350719452},{"id":"https://openalex.org/C111919701","wikidata":"https://www.wikidata.org/wiki/Q9135","display_name":"Operating system","level":1,"score":0.0},{"id":"https://openalex.org/C134306372","wikidata":"https://www.wikidata.org/wiki/Q7754","display_name":"Mathematical analysis","level":1,"score":0.0},{"id":"https://openalex.org/C104114177","wikidata":"https://www.wikidata.org/wiki/Q79782","display_name":"Motion (physics)","level":2,"score":0.0}],"mesh":[],"locations_count":1,"locations":[{"id":"doi:10.1109/dac56929.2023.10247855","is_oa":false,"landing_page_url":"https://doi.org/10.1109/dac56929.2023.10247855","pdf_url":null,"source":null,"license":null,"license_id":null,"version":"publishedVersion","is_accepted":true,"is_published":true,"raw_source_name":"2023 60th ACM/IEEE Design Automation Conference (DAC)","raw_type":"proceedings-article"}],"best_oa_location":null,"sustainable_development_goals":[],"awards":[],"funders":[],"has_content":{"pdf":false,"grobid_xml":false},"content_urls":null,"referenced_works_count":29,"referenced_works":["https://openalex.org/W1522301498","https://openalex.org/W1686810756","https://openalex.org/W1890351280","https://openalex.org/W2000956074","https://openalex.org/W2042519306","https://openalex.org/W2073256350","https://openalex.org/W2120052858","https://openalex.org/W2167425979","https://openalex.org/W2169621283","https://openalex.org/W2194775991","https://openalex.org/W2606722458","https://openalex.org/W2618530766","https://openalex.org/W2896478499","https://openalex.org/W2922301133","https://openalex.org/W2982219368","https://openalex.org/W3023212902","https://openalex.org/W3036878841","https://openalex.org/W3043675702","https://openalex.org/W3054316493","https://openalex.org/W3092319144","https://openalex.org/W3160664745","https://openalex.org/W4200126970","https://openalex.org/W4283451911","https://openalex.org/W4283791586","https://openalex.org/W4295312788","https://openalex.org/W4307079523","https://openalex.org/W6631190155","https://openalex.org/W6637373629","https://openalex.org/W6766978945"],"related_works":["https://openalex.org/W2058965144","https://openalex.org/W2164382479","https://openalex.org/W2146343568","https://openalex.org/W98480971","https://openalex.org/W2150291671","https://openalex.org/W2013643406","https://openalex.org/W2027972911","https://openalex.org/W2157978810","https://openalex.org/W4399871426","https://openalex.org/W2032140442"],"abstract_inverted_index":{"Modern":[0],"DNN":[1],"workloads":[2],"increasingly":[3],"rely":[4],"on":[5,89,120],"activation":[6,35],"functions":[7,36],"consisting":[8],"of":[9,126,157],"computationally":[10],"complex":[11],"operations.":[12],"This":[13,26],"poses":[14],"a":[15,30,54],"challenge":[16],"to":[17,70,81,97,135,162],"current":[18],"accelerators":[19,130],"optimized":[20],"for":[21,34],"convolutions":[22],"and":[23,47,84,111,149,154,159],"matrix-matrix":[24],"multiplications.":[25],"work":[27],"presents":[28],"Flex-SFU,":[29],"lightweight":[31],"hardware":[32,129],"accelerator":[33],"implementing":[37,53],"non-uniform":[38,82],"piecewise":[39,99],"interpolation":[40,73,83,101],"supporting":[41],"multiple":[42],"data":[43],"formats.":[44],"Non-Uniform":[45],"segments":[46],"floating-point":[48,85],"numbers":[49],"are":[50],"enabled":[51],"by":[52,131],"binary-tree":[55],"comparison":[56],"within":[57],"the":[58,72,76,123,142,163],"address":[59],"decoding":[60],"unit.":[61,167],"An":[62],"SGD-based":[63],"optimization":[64],"algorithm":[65],"with":[66,105,138],"heuristics":[67],"is":[68],"proposed":[69],"find":[71],"function":[74],"reducing":[75],"mean":[77,93],"squared":[78,94],"error.":[79],"Thanks":[80],"support,":[86],"Flex-SFU":[87,118],"achieves":[88],"average":[90],"22.3x":[91],"better":[92],"error":[95],"compared":[96],"previous":[98],"linear":[100],"approaches.":[102],"The":[103],"evaluation":[104],"more":[106],"than":[107],"700":[108],"computer":[109],"vision":[110],"natural":[112],"language":[113],"processing":[114,166],"models":[115],"shows":[116],"that":[117],"can,":[119],"average,":[121],"improve":[122],"end-to-end":[124],"performance":[125],"state-of-the-art":[127],"AI":[128],"35.7%,":[132],"achieving":[133],"up":[134],"3.3x":[136],"speedup":[137],"negligible":[139],"impact":[140],"in":[141],"models\u2019":[143],"accuracy":[144],"when":[145],"using":[146],"32":[147],"segments,":[148],"only":[150],"introducing":[151],"an":[152],"area":[153],"power":[155],"overhead":[156],"5.9%":[158],"0.8%":[160],"relative":[161],"baseline":[164],"vector":[165]},"counts_by_year":[{"year":2026,"cited_by_count":3},{"year":2025,"cited_by_count":7},{"year":2024,"cited_by_count":4},{"year":2023,"cited_by_count":2}],"updated_date":"2026-06-11T09:08:48.828518","created_date":"2025-10-10T00:00:00"}