{"id":"https://openalex.org/W4386764384","doi":"https://doi.org/10.1109/dac56929.2023.10248013","title":"DBPS: Dynamic Block Size and Precision Scaling for Efficient DNN Training Supported by RISC-V ISA Extensions","display_name":"DBPS: Dynamic Block Size and Precision Scaling for Efficient DNN Training Supported by RISC-V ISA Extensions","publication_year":2023,"publication_date":"2023-07-09","ids":{"openalex":"https://openalex.org/W4386764384","doi":"https://doi.org/10.1109/dac56929.2023.10248013"},"language":"en","primary_location":{"id":"doi:10.1109/dac56929.2023.10248013","is_oa":false,"landing_page_url":"https://doi.org/10.1109/dac56929.2023.10248013","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":"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/A5100426789","display_name":"Seunghyun Lee","orcid":"https://orcid.org/0000-0002-4701-2856"},"institutions":[{"id":"https://openalex.org/I197347611","display_name":"Korea University","ror":"https://ror.org/047dqcg40","country_code":"KR","type":"education","lineage":["https://openalex.org/I197347611"]}],"countries":["KR"],"is_corresponding":false,"raw_author_name":"Seunghyun Lee","raw_affiliation_strings":["Korea University,School of Electrical Engineering,South Korea","School of Electrical Engineering, Korea University, South Korea"],"raw_orcid":null,"affiliations":[{"raw_affiliation_string":"Korea University,School of Electrical Engineering,South Korea","institution_ids":["https://openalex.org/I197347611"]},{"raw_affiliation_string":"School of Electrical Engineering, Korea University, South Korea","institution_ids":["https://openalex.org/I197347611"]}]},{"author_position":"middle","author":{"id":"https://openalex.org/A5082063936","display_name":"Jeik Choi","orcid":"https://orcid.org/0000-0002-6674-450X"},"institutions":[{"id":"https://openalex.org/I193352282","display_name":"Daegu Gyeongbuk Institute of Science and Technology","ror":"https://ror.org/03frjya69","country_code":"KR","type":"education","lineage":["https://openalex.org/I193352282"]}],"countries":["KR"],"is_corresponding":false,"raw_author_name":"Jeik Choi","raw_affiliation_strings":["DGIST,Department of EECS,South Korea","Department of EECS, DGIST, South Korea"],"raw_orcid":null,"affiliations":[{"raw_affiliation_string":"DGIST,Department of EECS,South Korea","institution_ids":["https://openalex.org/I193352282"]},{"raw_affiliation_string":"Department of EECS, DGIST, South Korea","institution_ids":["https://openalex.org/I193352282"]}]},{"author_position":"middle","author":{"id":"https://openalex.org/A5046532535","display_name":"Seock-Hwan Noh","orcid":"https://orcid.org/0000-0002-9464-2949"},"institutions":[{"id":"https://openalex.org/I193352282","display_name":"Daegu Gyeongbuk Institute of Science and Technology","ror":"https://ror.org/03frjya69","country_code":"KR","type":"education","lineage":["https://openalex.org/I193352282"]}],"countries":["KR"],"is_corresponding":false,"raw_author_name":"Seockhwan Noh","raw_affiliation_strings":["DGIST,Department of EECS,South Korea","Department of EECS, DGIST, South Korea"],"raw_orcid":null,"affiliations":[{"raw_affiliation_string":"DGIST,Department of EECS,South Korea","institution_ids":["https://openalex.org/I193352282"]},{"raw_affiliation_string":"Department of EECS, DGIST, South Korea","institution_ids":["https://openalex.org/I193352282"]}]},{"author_position":"middle","author":{"id":"https://openalex.org/A5014191562","display_name":"Jahyun Koo","orcid":"https://orcid.org/0000-0002-2951-9522"},"institutions":[{"id":"https://openalex.org/I193352282","display_name":"Daegu Gyeongbuk Institute of Science and Technology","ror":"https://ror.org/03frjya69","country_code":"KR","type":"education","lineage":["https://openalex.org/I193352282"]}],"countries":["KR"],"is_corresponding":false,"raw_author_name":"Jahyun Koo","raw_affiliation_strings":["DGIST,Department of EECS,South Korea","Department of EECS, DGIST, South Korea"],"raw_orcid":null,"affiliations":[{"raw_affiliation_string":"DGIST,Department of EECS,South Korea","institution_ids":["https://openalex.org/I193352282"]},{"raw_affiliation_string":"Department of EECS, DGIST, South Korea","institution_ids":["https://openalex.org/I193352282"]}]},{"author_position":"last","author":{"id":"https://openalex.org/A5079872127","display_name":"Jaeha Kung","orcid":"https://orcid.org/0000-0001-6151-8602"},"institutions":[{"id":"https://openalex.org/I197347611","display_name":"Korea University","ror":"https://ror.org/047dqcg40","country_code":"KR","type":"education","lineage":["https://openalex.org/I197347611"]}],"countries":["KR"],"is_corresponding":false,"raw_author_name":"Jaeha Kung","raw_affiliation_strings":["Korea University,School of Electrical Engineering,South Korea","School of Electrical Engineering, Korea University, South Korea"],"raw_orcid":null,"affiliations":[{"raw_affiliation_string":"Korea University,School of Electrical Engineering,South Korea","institution_ids":["https://openalex.org/I197347611"]},{"raw_affiliation_string":"School of Electrical Engineering, Korea University, South Korea","institution_ids":["https://openalex.org/I197347611"]}]}],"institutions":[],"countries_distinct_count":1,"institutions_distinct_count":2,"corresponding_author_ids":[],"corresponding_institution_ids":[],"apc_list":null,"apc_paid":null,"fwci":null,"has_fulltext":false,"cited_by_count":9,"citation_normalized_percentile":null,"cited_by_percentile_year":null,"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":0.9994000196456909,"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":0.9994000196456909,"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":0.9940000176429749,"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/T10320","display_name":"Neural Networks and Applications","score":0.9847999811172485,"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"}}],"keywords":[{"id":"https://openalex.org/keywords/computer-science","display_name":"Computer science","score":0.8131531476974487},{"id":"https://openalex.org/keywords/block","display_name":"Block (permutation group theory)","score":0.7176996469497681},{"id":"https://openalex.org/keywords/block-size","display_name":"Block size","score":0.5900739431381226},{"id":"https://openalex.org/keywords/floating-point","display_name":"Floating point","score":0.5181504487991333},{"id":"https://openalex.org/keywords/scaling","display_name":"Scaling","score":0.49259573221206665},{"id":"https://openalex.org/keywords/reduced-instruction-set-computing","display_name":"Reduced instruction set computing","score":0.4922899305820465},{"id":"https://openalex.org/keywords/energy-consumption","display_name":"Energy consumption","score":0.4645763337612152},{"id":"https://openalex.org/keywords/point","display_name":"Point (geometry)","score":0.4243376851081848},{"id":"https://openalex.org/keywords/computer-hardware","display_name":"Computer hardware","score":0.36186957359313965},{"id":"https://openalex.org/keywords/instruction-set","display_name":"Instruction set","score":0.3270832300186157},{"id":"https://openalex.org/keywords/parallel-computing","display_name":"Parallel computing","score":0.32189056277275085},{"id":"https://openalex.org/keywords/algorithm","display_name":"Algorithm","score":0.22551435232162476},{"id":"https://openalex.org/keywords/operating-system","display_name":"Operating system","score":0.0814456045627594},{"id":"https://openalex.org/keywords/key","display_name":"Key (lock)","score":0.08111515641212463}],"concepts":[{"id":"https://openalex.org/C41008148","wikidata":"https://www.wikidata.org/wiki/Q21198","display_name":"Computer science","level":0,"score":0.8131531476974487},{"id":"https://openalex.org/C2777210771","wikidata":"https://www.wikidata.org/wiki/Q4927124","display_name":"Block (permutation group theory)","level":2,"score":0.7176996469497681},{"id":"https://openalex.org/C41431624","wikidata":"https://www.wikidata.org/wiki/Q1053357","display_name":"Block size","level":3,"score":0.5900739431381226},{"id":"https://openalex.org/C84211073","wikidata":"https://www.wikidata.org/wiki/Q117879","display_name":"Floating point","level":2,"score":0.5181504487991333},{"id":"https://openalex.org/C99844830","wikidata":"https://www.wikidata.org/wiki/Q102441924","display_name":"Scaling","level":2,"score":0.49259573221206665},{"id":"https://openalex.org/C126298526","wikidata":"https://www.wikidata.org/wiki/Q189376","display_name":"Reduced instruction set computing","level":3,"score":0.4922899305820465},{"id":"https://openalex.org/C2780165032","wikidata":"https://www.wikidata.org/wiki/Q16869822","display_name":"Energy consumption","level":2,"score":0.4645763337612152},{"id":"https://openalex.org/C28719098","wikidata":"https://www.wikidata.org/wiki/Q44946","display_name":"Point (geometry)","level":2,"score":0.4243376851081848},{"id":"https://openalex.org/C9390403","wikidata":"https://www.wikidata.org/wiki/Q3966","display_name":"Computer hardware","level":1,"score":0.36186957359313965},{"id":"https://openalex.org/C202491316","wikidata":"https://www.wikidata.org/wiki/Q272683","display_name":"Instruction set","level":2,"score":0.3270832300186157},{"id":"https://openalex.org/C173608175","wikidata":"https://www.wikidata.org/wiki/Q232661","display_name":"Parallel computing","level":1,"score":0.32189056277275085},{"id":"https://openalex.org/C11413529","wikidata":"https://www.wikidata.org/wiki/Q8366","display_name":"Algorithm","level":1,"score":0.22551435232162476},{"id":"https://openalex.org/C111919701","wikidata":"https://www.wikidata.org/wiki/Q9135","display_name":"Operating system","level":1,"score":0.0814456045627594},{"id":"https://openalex.org/C26517878","wikidata":"https://www.wikidata.org/wiki/Q228039","display_name":"Key (lock)","level":2,"score":0.08111515641212463},{"id":"https://openalex.org/C33923547","wikidata":"https://www.wikidata.org/wiki/Q395","display_name":"Mathematics","level":0,"score":0.0},{"id":"https://openalex.org/C18903297","wikidata":"https://www.wikidata.org/wiki/Q7150","display_name":"Ecology","level":1,"score":0.0},{"id":"https://openalex.org/C2524010","wikidata":"https://www.wikidata.org/wiki/Q8087","display_name":"Geometry","level":1,"score":0.0},{"id":"https://openalex.org/C86803240","wikidata":"https://www.wikidata.org/wiki/Q420","display_name":"Biology","level":0,"score":0.0}],"mesh":[],"locations_count":1,"locations":[{"id":"doi:10.1109/dac56929.2023.10248013","is_oa":false,"landing_page_url":"https://doi.org/10.1109/dac56929.2023.10248013","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":[{"id":"https://metadata.un.org/sdg/7","score":0.9100000262260437,"display_name":"Affordable and clean energy"}],"awards":[],"funders":[{"id":"https://openalex.org/F4320322202","display_name":"IC Design Education Center","ror":"https://ror.org/005v57z85"},{"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":11,"referenced_works":["https://openalex.org/W2269491305","https://openalex.org/W2750822049","https://openalex.org/W2946572707","https://openalex.org/W2949833565","https://openalex.org/W2963367920","https://openalex.org/W3106213055","https://openalex.org/W3208633927","https://openalex.org/W3213528054","https://openalex.org/W4287204036","https://openalex.org/W4292779060","https://openalex.org/W6751609549"],"related_works":["https://openalex.org/W2538644970","https://openalex.org/W4376881175","https://openalex.org/W4310584696","https://openalex.org/W4385730960","https://openalex.org/W4237840813","https://openalex.org/W4364295250","https://openalex.org/W2128502296","https://openalex.org/W2993622674","https://openalex.org/W1833044483","https://openalex.org/W4382053335"],"abstract_inverted_index":{"Over":[0],"the":[1,30,46,77,82,97,101,105,110,140,158,172],"past":[2],"decade,":[3],"it":[4],"has":[5],"been":[6],"found":[7],"that":[8,74,103],"deep":[9],"neural":[10],"networks":[11],"(DNNs)":[12],"perform":[13],"better":[14],"on":[15,91],"visual":[16],"perception":[17],"and":[18,36,81,100,120,161,167],"language":[19],"understanding":[20],"tasks":[21],"as":[22],"their":[23],"size":[24,78,99,119],"increases.":[25],"However,":[26],"this":[27,64],"comes":[28],"at":[29],"cost":[31],"of":[32,79,85],"high":[33],"energy":[34,162],"consumption":[35,84,163],"large":[37,43],"memory":[38],"requirement":[39],"to":[40],"train":[41],"such":[42],"models.":[44],"As":[45,155],"training":[47,94,106,112,159,173],"DNNs":[48],"necessitates":[49],"a":[50,68,132,152,156],"wide":[51],"dynamic":[52,117],"range":[53],"in":[54,151],"representing":[55],"tensors,":[56],"floating":[57,70],"point":[58,71],"formats":[59],"are":[60],"normally":[61],"used.":[62],"In":[63],"work,":[65],"we":[66,115],"utilize":[67],"block":[69,98,118],"(BFP)":[72],"format":[73],"significantly":[75],"reduces":[76],"tensors":[80],"power":[83],"arithmetic":[86,145],"units.":[87],"Unfortunately,":[88],"prior":[89],"work":[90],"BFP-based":[92,111],"DNN":[93,127],"empirically":[95],"selects":[96],"precision":[102,121],"maintain":[104],"accuracy.":[107,174],"To":[108],"make":[109],"more":[113],"feasible,":[114],"propose":[116],"scaling":[122],"(DBPS)":[123],"for":[124],"highly":[125],"efficient":[126],"training.":[128],"We":[129],"also":[130],"present":[131],"hardware":[133],"accelerator,":[134],"called":[135],"DBPS":[136,141],"core,":[137],"which":[138],"supports":[139],"control":[142],"by":[143,165],"configuring":[144],"units":[146],"with":[147],"custom":[148],"instructions":[149],"extended":[150],"RISC-V":[153],"processor.":[154],"result,":[157],"time":[160],"reduce":[164],"67.1%":[166],"72.0%,":[168],"respectively,":[169],"without":[170],"hurting":[171]},"counts_by_year":[{"year":2026,"cited_by_count":3},{"year":2025,"cited_by_count":3},{"year":2024,"cited_by_count":3}],"updated_date":"2026-07-14T23:27:15.235271","created_date":"2025-10-10T00:00:00"}