{"id":"https://openalex.org/W4401408947","doi":"https://doi.org/10.1145/3673038.3673045","title":"VitBit: Enhancing Embedded GPU Performance for AI Workloads through Register Operand Packing","display_name":"VitBit: Enhancing Embedded GPU Performance for AI Workloads through Register Operand Packing","publication_year":2024,"publication_date":"2024-08-08","ids":{"openalex":"https://openalex.org/W4401408947","doi":"https://doi.org/10.1145/3673038.3673045"},"language":"en","primary_location":{"id":"doi:10.1145/3673038.3673045","is_oa":true,"landing_page_url":"https://doi.org/10.1145/3673038.3673045","pdf_url":null,"source":null,"license":"cc-by","license_id":"https://openalex.org/licenses/cc-by","version":"publishedVersion","is_accepted":true,"is_published":true,"raw_source_name":"Proceedings of the 53rd International Conference on Parallel Processing","raw_type":"proceedings-article"},"type":"conference-paper","indexed_in":["crossref"],"open_access":{"is_oa":true,"oa_status":"gold","oa_url":"https://doi.org/10.1145/3673038.3673045","any_repository_has_fulltext":null},"authorships":[{"author_position":"first","author":{"id":"https://openalex.org/A5102603746","display_name":"Jaebeom Jeon","orcid":"https://orcid.org/0009-0009-2947-043X"},"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":"Jaebeom Jeon","raw_affiliation_strings":["Korea University, South Korea"],"raw_orcid":"https://orcid.org/0009-0009-2947-043X","affiliations":[{"raw_affiliation_string":"Korea University, South Korea","institution_ids":["https://openalex.org/I197347611"]}]},{"author_position":"middle","author":{"id":"https://openalex.org/A5106355596","display_name":"Minseong Gil","orcid":null},"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":"Minseong Gil","raw_affiliation_strings":["Korea University, South Korea"],"raw_orcid":"https://orcid.org/0009-0005-8664-9840","affiliations":[{"raw_affiliation_string":"Korea University, South Korea","institution_ids":["https://openalex.org/I197347611"]}]},{"author_position":"middle","author":{"id":"https://openalex.org/A5076935570","display_name":"J. H. Kim","orcid":null},"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":"Junsu Kim","raw_affiliation_strings":["Korea University, South Korea"],"raw_orcid":"https://orcid.org/0009-0008-1582-5212","affiliations":[{"raw_affiliation_string":"Korea University, South Korea","institution_ids":["https://openalex.org/I197347611"]}]},{"author_position":"middle","author":{"id":null,"display_name":"Jaeyong Park","orcid":"https://orcid.org/0009-0001-4373-3963"},"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":"Jaeyong Park","raw_affiliation_strings":["Korea University, South Korea"],"raw_orcid":"https://orcid.org/0009-0001-4373-3963","affiliations":[{"raw_affiliation_string":"Korea University, South Korea","institution_ids":["https://openalex.org/I197347611"]}]},{"author_position":"middle","author":{"id":"https://openalex.org/A5055061586","display_name":"Gunjae Koo","orcid":"https://orcid.org/0000-0003-1706-6850"},"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":"Gunjae Koo","raw_affiliation_strings":["Korea University, South Korea"],"raw_orcid":"https://orcid.org/0000-0003-1706-6850","affiliations":[{"raw_affiliation_string":"Korea University, South Korea","institution_ids":["https://openalex.org/I197347611"]}]},{"author_position":"middle","author":{"id":"https://openalex.org/A5040253395","display_name":"Myung Kuk Yoon","orcid":"https://orcid.org/0000-0002-9332-0251"},"institutions":[{"id":"https://openalex.org/I138925566","display_name":"Ewha Womans University","ror":"https://ror.org/053fp5c05","country_code":"KR","type":"education","lineage":["https://openalex.org/I138925566"]}],"countries":["KR"],"is_corresponding":false,"raw_author_name":"Myung Kuk Yoon","raw_affiliation_strings":["Ewha Womans University, South Korea"],"raw_orcid":"https://orcid.org/0000-0002-9332-0251","affiliations":[{"raw_affiliation_string":"Ewha Womans University, South Korea","institution_ids":["https://openalex.org/I138925566"]}]},{"author_position":"last","author":{"id":"https://openalex.org/A5015526090","display_name":"Yunho Oh","orcid":"https://orcid.org/0000-0001-6442-3705"},"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":"Yunho Oh","raw_affiliation_strings":["Korea University, South Korea"],"raw_orcid":"https://orcid.org/0000-0001-6442-3705","affiliations":[{"raw_affiliation_string":"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":0.4177,"has_fulltext":false,"cited_by_count":1,"citation_normalized_percentile":{"value":0.5013876,"is_in_top_1_percent":false,"is_in_top_10_percent":false},"cited_by_percentile_year":{"min":91,"max":95},"biblio":{"volume":null,"issue":null,"first_page":"1012","last_page":"1021"},"is_retracted":false,"is_paratext":false,"is_xpac":false,"primary_topic":{"id":"https://openalex.org/T10054","display_name":"Parallel Computing and Optimization Techniques","score":0.9998999834060669,"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":0.9998999834060669,"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/T10036","display_name":"Advanced Neural Network Applications","score":0.9991999864578247,"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/T11181","display_name":"Advanced Data Storage Technologies","score":0.9944000244140625,"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/computer-science","display_name":"Computer science","score":0.8506423234939575},{"id":"https://openalex.org/keywords/benchmark","display_name":"Benchmark (surveying)","score":0.6212842464447021},{"id":"https://openalex.org/keywords/scalability","display_name":"Scalability","score":0.5798923373222351},{"id":"https://openalex.org/keywords/throughput","display_name":"Throughput","score":0.579200267791748},{"id":"https://openalex.org/keywords/parallel-computing","display_name":"Parallel computing","score":0.5780872106552124},{"id":"https://openalex.org/keywords/cuda","display_name":"CUDA","score":0.5610499382019043},{"id":"https://openalex.org/keywords/floating-point","display_name":"Floating point","score":0.5517320036888123},{"id":"https://openalex.org/keywords/operand","display_name":"Operand","score":0.4992814064025879},{"id":"https://openalex.org/keywords/graphics-processing-unit","display_name":"Graphics processing unit","score":0.47124892473220825},{"id":"https://openalex.org/keywords/general-purpose-computing-on-graphics-processing-units","display_name":"General-purpose computing on graphics processing units","score":0.47043320536613464},{"id":"https://openalex.org/keywords/computer-architecture","display_name":"Computer architecture","score":0.4517783224582672},{"id":"https://openalex.org/keywords/instruction-set","display_name":"Instruction set","score":0.45061177015304565},{"id":"https://openalex.org/keywords/supercomputer","display_name":"Supercomputer","score":0.42055124044418335},{"id":"https://openalex.org/keywords/reduction","display_name":"Reduction (mathematics)","score":0.4192717373371124},{"id":"https://openalex.org/keywords/computation","display_name":"Computation","score":0.41897594928741455},{"id":"https://openalex.org/keywords/graphics","display_name":"Graphics","score":0.413382351398468},{"id":"https://openalex.org/keywords/speedup","display_name":"Speedup","score":0.4104156494140625},{"id":"https://openalex.org/keywords/computer-engineering","display_name":"Computer engineering","score":0.3942588269710541},{"id":"https://openalex.org/keywords/computer-hardware","display_name":"Computer hardware","score":0.25660014152526855},{"id":"https://openalex.org/keywords/algorithm","display_name":"Algorithm","score":0.13014450669288635},{"id":"https://openalex.org/keywords/operating-system","display_name":"Operating system","score":0.11262762546539307}],"concepts":[{"id":"https://openalex.org/C41008148","wikidata":"https://www.wikidata.org/wiki/Q21198","display_name":"Computer science","level":0,"score":0.8506423234939575},{"id":"https://openalex.org/C185798385","wikidata":"https://www.wikidata.org/wiki/Q1161707","display_name":"Benchmark (surveying)","level":2,"score":0.6212842464447021},{"id":"https://openalex.org/C48044578","wikidata":"https://www.wikidata.org/wiki/Q727490","display_name":"Scalability","level":2,"score":0.5798923373222351},{"id":"https://openalex.org/C157764524","wikidata":"https://www.wikidata.org/wiki/Q1383412","display_name":"Throughput","level":3,"score":0.579200267791748},{"id":"https://openalex.org/C173608175","wikidata":"https://www.wikidata.org/wiki/Q232661","display_name":"Parallel computing","level":1,"score":0.5780872106552124},{"id":"https://openalex.org/C2778119891","wikidata":"https://www.wikidata.org/wiki/Q477690","display_name":"CUDA","level":2,"score":0.5610499382019043},{"id":"https://openalex.org/C84211073","wikidata":"https://www.wikidata.org/wiki/Q117879","display_name":"Floating point","level":2,"score":0.5517320036888123},{"id":"https://openalex.org/C55526617","wikidata":"https://www.wikidata.org/wiki/Q719375","display_name":"Operand","level":2,"score":0.4992814064025879},{"id":"https://openalex.org/C2779851693","wikidata":"https://www.wikidata.org/wiki/Q183484","display_name":"Graphics processing unit","level":2,"score":0.47124892473220825},{"id":"https://openalex.org/C50630238","wikidata":"https://www.wikidata.org/wiki/Q971505","display_name":"General-purpose computing on graphics processing units","level":3,"score":0.47043320536613464},{"id":"https://openalex.org/C118524514","wikidata":"https://www.wikidata.org/wiki/Q173212","display_name":"Computer architecture","level":1,"score":0.4517783224582672},{"id":"https://openalex.org/C202491316","wikidata":"https://www.wikidata.org/wiki/Q272683","display_name":"Instruction set","level":2,"score":0.45061177015304565},{"id":"https://openalex.org/C83283714","wikidata":"https://www.wikidata.org/wiki/Q121117","display_name":"Supercomputer","level":2,"score":0.42055124044418335},{"id":"https://openalex.org/C111335779","wikidata":"https://www.wikidata.org/wiki/Q3454686","display_name":"Reduction (mathematics)","level":2,"score":0.4192717373371124},{"id":"https://openalex.org/C45374587","wikidata":"https://www.wikidata.org/wiki/Q12525525","display_name":"Computation","level":2,"score":0.41897594928741455},{"id":"https://openalex.org/C21442007","wikidata":"https://www.wikidata.org/wiki/Q1027879","display_name":"Graphics","level":2,"score":0.413382351398468},{"id":"https://openalex.org/C68339613","wikidata":"https://www.wikidata.org/wiki/Q1549489","display_name":"Speedup","level":2,"score":0.4104156494140625},{"id":"https://openalex.org/C113775141","wikidata":"https://www.wikidata.org/wiki/Q428691","display_name":"Computer engineering","level":1,"score":0.3942588269710541},{"id":"https://openalex.org/C9390403","wikidata":"https://www.wikidata.org/wiki/Q3966","display_name":"Computer hardware","level":1,"score":0.25660014152526855},{"id":"https://openalex.org/C11413529","wikidata":"https://www.wikidata.org/wiki/Q8366","display_name":"Algorithm","level":1,"score":0.13014450669288635},{"id":"https://openalex.org/C111919701","wikidata":"https://www.wikidata.org/wiki/Q9135","display_name":"Operating system","level":1,"score":0.11262762546539307},{"id":"https://openalex.org/C2524010","wikidata":"https://www.wikidata.org/wiki/Q8087","display_name":"Geometry","level":1,"score":0.0},{"id":"https://openalex.org/C205649164","wikidata":"https://www.wikidata.org/wiki/Q1071","display_name":"Geography","level":0,"score":0.0},{"id":"https://openalex.org/C13280743","wikidata":"https://www.wikidata.org/wiki/Q131089","display_name":"Geodesy","level":1,"score":0.0},{"id":"https://openalex.org/C33923547","wikidata":"https://www.wikidata.org/wiki/Q395","display_name":"Mathematics","level":0,"score":0.0},{"id":"https://openalex.org/C555944384","wikidata":"https://www.wikidata.org/wiki/Q249","display_name":"Wireless","level":2,"score":0.0}],"mesh":[],"locations_count":1,"locations":[{"id":"doi:10.1145/3673038.3673045","is_oa":true,"landing_page_url":"https://doi.org/10.1145/3673038.3673045","pdf_url":null,"source":null,"license":"cc-by","license_id":"https://openalex.org/licenses/cc-by","version":"publishedVersion","is_accepted":true,"is_published":true,"raw_source_name":"Proceedings of the 53rd International Conference on Parallel Processing","raw_type":"proceedings-article"}],"best_oa_location":{"id":"doi:10.1145/3673038.3673045","is_oa":true,"landing_page_url":"https://doi.org/10.1145/3673038.3673045","pdf_url":null,"source":null,"license":"cc-by","license_id":"https://openalex.org/licenses/cc-by","version":"publishedVersion","is_accepted":true,"is_published":true,"raw_source_name":"Proceedings of the 53rd International Conference on Parallel Processing","raw_type":"proceedings-article"},"sustainable_development_goals":[{"score":0.49000000953674316,"display_name":"Affordable and clean energy","id":"https://metadata.un.org/sdg/7"}],"awards":[{"id":"https://openalex.org/G2221399033","display_name":null,"funder_award_id":"2022R1C1C1011021","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":{"grobid_xml":false,"pdf":false},"content_urls":null,"referenced_works_count":21,"referenced_works":["https://openalex.org/W95608104","https://openalex.org/W2153190325","https://openalex.org/W2734941459","https://openalex.org/W2924943819","https://openalex.org/W2933147490","https://openalex.org/W2971428146","https://openalex.org/W2979429416","https://openalex.org/W3035232708","https://openalex.org/W3092106071","https://openalex.org/W3098871454","https://openalex.org/W3130554079","https://openalex.org/W3130673459","https://openalex.org/W3183895983","https://openalex.org/W4210338699","https://openalex.org/W4226335630","https://openalex.org/W4280633999","https://openalex.org/W4312309118","https://openalex.org/W4313069943","https://openalex.org/W4380881150","https://openalex.org/W4387723591","https://openalex.org/W4390873361"],"related_works":["https://openalex.org/W2151046618","https://openalex.org/W1972148443","https://openalex.org/W1969233021","https://openalex.org/W2167646277","https://openalex.org/W2063573318","https://openalex.org/W4389476319","https://openalex.org/W4232549588","https://openalex.org/W2037018251","https://openalex.org/W4241438378","https://openalex.org/W2388314963"],"abstract_inverted_index":{"The":[0],"rapid":[1],"advancement":[2],"of":[3,14,49,59,94,119,130,147,162,172,218],"Artificial":[4],"Intelligence":[5],"(AI)":[6],"necessitates":[7],"significant":[8],"enhancements":[9],"in":[10,54,109,182,194,230],"the":[11,56,117,127,144,208,215],"energy":[12],"efficiency":[13],"Graphics":[15],"Processing":[16],"Units":[17],"(GPUs)":[18],"for":[19,31,67,123,197,226],"Deep":[20],"Neural":[21],"Network":[22],"(DNN)":[23],"workloads.":[24,111],"Such":[25],"a":[26,46,81,191,221],"challenge":[27],"is":[28],"particularly":[29],"critical":[30],"embedded":[32],"GPUs,":[33,149],"which":[34,105,158],"operate":[35],"within":[36],"stringent":[37],"power":[38],"constraints.":[39],"Traditional":[40],"GPU":[41,228],"architectures,":[42],"designed":[43,85],"to":[44,71,86,190],"support":[45,66],"limited":[47],"set":[48],"numeric":[50,69],"formats,":[51],"face":[52],"challenges":[53],"meeting":[55],"diverse":[57],"requirements":[58],"modern":[60,148],"AI":[61,110,199,231],"applications.":[62,232],"These":[63],"applications":[64],"demand":[65],"various":[68],"formats":[70,122],"optimize":[72],"computational":[73,216],"speed":[74],"and":[75,126,134,137,164,185,214,223],"efficiency.":[76],"This":[77,141],"paper":[78],"proposes":[79],"VitBit,":[80],"novel":[82],"software":[83],"technique":[84],"overcome":[87],"these":[88],"limitations":[89],"by":[90],"enabling":[91],"efficient":[92],"processing":[93],"arbitrary":[95,120],"integer":[96,121],"format":[97],"values,":[98],"especially":[99],"those":[100,152],"8":[101],"bits":[102],"or":[103],"fewer,":[104],"are":[106],"increasingly":[107],"prevalent":[108],"VitBit":[112,173,205],"introduces":[113],"two":[114],"key":[115],"innovations:":[116],"packing":[118],"parallel":[124],"computation":[125],"simultaneous":[128],"execution":[129,195],"Tensor":[131],"cores,":[132],"INT":[133],"FP":[135],"(Integer":[136],"Floating-Point)":[138],"CUDA":[139],"cores.":[140],"approach":[142],"leverages":[143],"architectural":[145],"features":[146],"such":[150],"as":[151],"based":[153],"on":[154,174],"NVIDIA":[155,175],"Ampere":[156],"architecture,":[157],"allows":[159],"concurrent":[160],"operation":[161],"FP32":[163],"INT32":[165],"cores":[166],"at":[167],"full":[168],"throughput.":[169],"Our":[170],"evaluation":[171],"Jetson":[176],"AGX":[177],"Orin":[178],"demonstrates":[179],"substantial":[180],"improvements":[181],"arithmetic":[183],"density":[184],"peak":[186],"throughput,":[187],"achieving":[188],"up":[189],"22%":[192],"reduction":[193],"time":[196],"benchmark":[198],"workloads":[200],"without":[201],"compromising":[202],"inference":[203],"accuracy.":[204],"effectively":[206],"bridges":[207],"gap":[209],"between":[210],"current":[211],"hardware":[212],"capabilities":[213],"demands":[217],"AI,":[219],"offering":[220],"scalable":[222],"cost-effective":[224],"method":[225],"enhancing":[227],"performance":[229]},"counts_by_year":[{"year":2025,"cited_by_count":1}],"updated_date":"2026-08-18T07:49:30.821534","created_date":"2025-10-10T00:00:00"}