{"id":"https://openalex.org/W4387870780","doi":"https://doi.org/10.1109/icc45041.2023.10278567","title":"g-BERT: Enabling Green BERT Deployment on FPGA via Hardware-Aware Hybrid Pruning","display_name":"g-BERT: Enabling Green BERT Deployment on FPGA via Hardware-Aware Hybrid Pruning","publication_year":2023,"publication_date":"2023-05-28","ids":{"openalex":"https://openalex.org/W4387870780","doi":"https://doi.org/10.1109/icc45041.2023.10278567"},"language":"en","primary_location":{"id":"doi:10.1109/icc45041.2023.10278567","is_oa":false,"landing_page_url":"https://doi.org/10.1109/icc45041.2023.10278567","pdf_url":null,"source":null,"license":null,"license_id":null,"version":"publishedVersion","is_accepted":true,"is_published":true,"raw_source_name":"ICC 2023 - IEEE International Conference on Communications","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/A5100580160","display_name":"Yueyin Bai","orcid":"https://orcid.org/0009-0006-2938-5153"},"institutions":[{"id":"https://openalex.org/I24943067","display_name":"Fudan University","ror":"https://ror.org/013q1eq08","country_code":"CN","type":"education","lineage":["https://openalex.org/I24943067"]}],"countries":["CN"],"is_corresponding":false,"raw_author_name":"Yueyin Bai","raw_affiliation_strings":["State Key Lab of ASIC &#x0026; System, Fudan University,Shanghai,China"],"raw_orcid":null,"affiliations":[{"raw_affiliation_string":"State Key Lab of ASIC &#x0026; System, Fudan University,Shanghai,China","institution_ids":["https://openalex.org/I24943067"]}]},{"author_position":"middle","author":{"id":"https://openalex.org/A5102009223","display_name":"Hao Zhou","orcid":"https://orcid.org/0000-0003-0642-881X"},"institutions":[{"id":"https://openalex.org/I24943067","display_name":"Fudan University","ror":"https://ror.org/013q1eq08","country_code":"CN","type":"education","lineage":["https://openalex.org/I24943067"]}],"countries":["CN"],"is_corresponding":false,"raw_author_name":"Hao Zhou","raw_affiliation_strings":["State Key Lab of ASIC &#x0026; System, Fudan University,Shanghai,China"],"raw_orcid":null,"affiliations":[{"raw_affiliation_string":"State Key Lab of ASIC &#x0026; System, Fudan University,Shanghai,China","institution_ids":["https://openalex.org/I24943067"]}]},{"author_position":"middle","author":{"id":"https://openalex.org/A5100761386","display_name":"Ruiqi Chen","orcid":"https://orcid.org/0000-0001-6837-5675"},"institutions":[{"id":"https://openalex.org/I24943067","display_name":"Fudan University","ror":"https://ror.org/013q1eq08","country_code":"CN","type":"education","lineage":["https://openalex.org/I24943067"]}],"countries":["CN"],"is_corresponding":false,"raw_author_name":"Ruiqi Chen","raw_affiliation_strings":["State Key Lab of ASIC &#x0026; System, Fudan University,Shanghai,China"],"raw_orcid":null,"affiliations":[{"raw_affiliation_string":"State Key Lab of ASIC &#x0026; System, Fudan University,Shanghai,China","institution_ids":["https://openalex.org/I24943067"]}]},{"author_position":"middle","author":{"id":"https://openalex.org/A5113029559","display_name":"Kuangjie Zou","orcid":null},"institutions":[{"id":"https://openalex.org/I24943067","display_name":"Fudan University","ror":"https://ror.org/013q1eq08","country_code":"CN","type":"education","lineage":["https://openalex.org/I24943067"]}],"countries":["CN"],"is_corresponding":false,"raw_author_name":"Kuangjie Zou","raw_affiliation_strings":["State Key Lab of ASIC &#x0026; System, Fudan University,Shanghai,China"],"raw_orcid":null,"affiliations":[{"raw_affiliation_string":"State Key Lab of ASIC &#x0026; System, Fudan University,Shanghai,China","institution_ids":["https://openalex.org/I24943067"]}]},{"author_position":"middle","author":{"id":"https://openalex.org/A5072580709","display_name":"Jialin Cao","orcid":"https://orcid.org/0009-0001-9054-5500"},"institutions":[{"id":"https://openalex.org/I24943067","display_name":"Fudan University","ror":"https://ror.org/013q1eq08","country_code":"CN","type":"education","lineage":["https://openalex.org/I24943067"]}],"countries":["CN"],"is_corresponding":false,"raw_author_name":"Jialin Cao","raw_affiliation_strings":["State Key Lab of ASIC &#x0026; System, Fudan University,Shanghai,China"],"raw_orcid":null,"affiliations":[{"raw_affiliation_string":"State Key Lab of ASIC &#x0026; System, Fudan University,Shanghai,China","institution_ids":["https://openalex.org/I24943067"]}]},{"author_position":"middle","author":{"id":"https://openalex.org/A5056063611","display_name":"Haoyang Zhang","orcid":"https://orcid.org/0000-0003-4496-4752"},"institutions":[{"id":"https://openalex.org/I24943067","display_name":"Fudan University","ror":"https://ror.org/013q1eq08","country_code":"CN","type":"education","lineage":["https://openalex.org/I24943067"]}],"countries":["CN"],"is_corresponding":false,"raw_author_name":"Haoyang Zhang","raw_affiliation_strings":["State Key Lab of ASIC &#x0026; System, Fudan University,Shanghai,China"],"raw_orcid":null,"affiliations":[{"raw_affiliation_string":"State Key Lab of ASIC &#x0026; System, Fudan University,Shanghai,China","institution_ids":["https://openalex.org/I24943067"]}]},{"author_position":"middle","author":{"id":"https://openalex.org/A5101854580","display_name":"Jianli Chen","orcid":"https://orcid.org/0000-0002-1391-2696"},"institutions":[{"id":"https://openalex.org/I24943067","display_name":"Fudan University","ror":"https://ror.org/013q1eq08","country_code":"CN","type":"education","lineage":["https://openalex.org/I24943067"]}],"countries":["CN"],"is_corresponding":false,"raw_author_name":"Jianli Chen","raw_affiliation_strings":["State Key Lab of ASIC &#x0026; System, Fudan University,Shanghai,China"],"raw_orcid":null,"affiliations":[{"raw_affiliation_string":"State Key Lab of ASIC &#x0026; System, Fudan University,Shanghai,China","institution_ids":["https://openalex.org/I24943067"]}]},{"author_position":"middle","author":{"id":"https://openalex.org/A5112536909","display_name":"Jun Yu","orcid":"https://orcid.org/0009-0006-3920-9321"},"institutions":[{"id":"https://openalex.org/I24943067","display_name":"Fudan University","ror":"https://ror.org/013q1eq08","country_code":"CN","type":"education","lineage":["https://openalex.org/I24943067"]}],"countries":["CN"],"is_corresponding":false,"raw_author_name":"Jun Yu","raw_affiliation_strings":["State Key Lab of ASIC &#x0026; System, Fudan University,Shanghai,China"],"raw_orcid":null,"affiliations":[{"raw_affiliation_string":"State Key Lab of ASIC &#x0026; System, Fudan University,Shanghai,China","institution_ids":["https://openalex.org/I24943067"]}]},{"author_position":"last","author":{"id":"https://openalex.org/A5100366737","display_name":"Kun Wang","orcid":"https://orcid.org/0000-0003-4180-6150"},"institutions":[{"id":"https://openalex.org/I24943067","display_name":"Fudan University","ror":"https://ror.org/013q1eq08","country_code":"CN","type":"education","lineage":["https://openalex.org/I24943067"]}],"countries":["CN"],"is_corresponding":false,"raw_author_name":"Kun Wang","raw_affiliation_strings":["State Key Lab of ASIC &#x0026; System, Fudan University,Shanghai,China"],"raw_orcid":null,"affiliations":[{"raw_affiliation_string":"State Key Lab of ASIC &#x0026; System, Fudan University,Shanghai,China","institution_ids":["https://openalex.org/I24943067"]}]}],"institutions":[],"countries_distinct_count":1,"institutions_distinct_count":1,"corresponding_author_ids":[],"corresponding_institution_ids":["https://openalex.org/I24943067"],"apc_list":null,"apc_paid":null,"fwci":null,"has_fulltext":false,"cited_by_count":2,"citation_normalized_percentile":null,"cited_by_percentile_year":null,"biblio":{"volume":null,"issue":null,"first_page":"1706","last_page":"1711"},"is_retracted":false,"is_paratext":false,"is_xpac":false,"primary_topic":{"id":"https://openalex.org/T10036","display_name":"Advanced Neural Network Applications","score":0.9998000264167786,"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":0.9998000264167786,"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":0.9965999722480774,"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/T10502","display_name":"Advanced Memory and Neural Computing","score":0.9962000250816345,"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/computer-science","display_name":"Computer science","score":0.8043822050094604},{"id":"https://openalex.org/keywords/field-programmable-gate-array","display_name":"Field-programmable gate array","score":0.7217928171157837},{"id":"https://openalex.org/keywords/flops","display_name":"FLOPS","score":0.709568977355957},{"id":"https://openalex.org/keywords/pruning","display_name":"Pruning","score":0.6670110821723938},{"id":"https://openalex.org/keywords/latency","display_name":"Latency (audio)","score":0.4637734293937683},{"id":"https://openalex.org/keywords/inference","display_name":"Inference","score":0.4515704810619354},{"id":"https://openalex.org/keywords/embedded-system","display_name":"Embedded system","score":0.45023828744888306},{"id":"https://openalex.org/keywords/hardware-architecture","display_name":"Hardware architecture","score":0.43075618147850037},{"id":"https://openalex.org/keywords/software-deployment","display_name":"Software deployment","score":0.41030973196029663},{"id":"https://openalex.org/keywords/computer-hardware","display_name":"Computer hardware","score":0.39248600602149963},{"id":"https://openalex.org/keywords/parallel-computing","display_name":"Parallel computing","score":0.38133227825164795},{"id":"https://openalex.org/keywords/computer-engineering","display_name":"Computer engineering","score":0.36089348793029785},{"id":"https://openalex.org/keywords/artificial-intelligence","display_name":"Artificial intelligence","score":0.3462945222854614},{"id":"https://openalex.org/keywords/operating-system","display_name":"Operating system","score":0.09825468063354492},{"id":"https://openalex.org/keywords/software","display_name":"Software","score":0.0952589213848114}],"concepts":[{"id":"https://openalex.org/C41008148","wikidata":"https://www.wikidata.org/wiki/Q21198","display_name":"Computer science","level":0,"score":0.8043822050094604},{"id":"https://openalex.org/C42935608","wikidata":"https://www.wikidata.org/wiki/Q190411","display_name":"Field-programmable gate array","level":2,"score":0.7217928171157837},{"id":"https://openalex.org/C3826847","wikidata":"https://www.wikidata.org/wiki/Q188768","display_name":"FLOPS","level":2,"score":0.709568977355957},{"id":"https://openalex.org/C108010975","wikidata":"https://www.wikidata.org/wiki/Q500094","display_name":"Pruning","level":2,"score":0.6670110821723938},{"id":"https://openalex.org/C82876162","wikidata":"https://www.wikidata.org/wiki/Q17096504","display_name":"Latency (audio)","level":2,"score":0.4637734293937683},{"id":"https://openalex.org/C2776214188","wikidata":"https://www.wikidata.org/wiki/Q408386","display_name":"Inference","level":2,"score":0.4515704810619354},{"id":"https://openalex.org/C149635348","wikidata":"https://www.wikidata.org/wiki/Q193040","display_name":"Embedded system","level":1,"score":0.45023828744888306},{"id":"https://openalex.org/C65232700","wikidata":"https://www.wikidata.org/wiki/Q5656403","display_name":"Hardware architecture","level":3,"score":0.43075618147850037},{"id":"https://openalex.org/C105339364","wikidata":"https://www.wikidata.org/wiki/Q2297740","display_name":"Software deployment","level":2,"score":0.41030973196029663},{"id":"https://openalex.org/C9390403","wikidata":"https://www.wikidata.org/wiki/Q3966","display_name":"Computer hardware","level":1,"score":0.39248600602149963},{"id":"https://openalex.org/C173608175","wikidata":"https://www.wikidata.org/wiki/Q232661","display_name":"Parallel computing","level":1,"score":0.38133227825164795},{"id":"https://openalex.org/C113775141","wikidata":"https://www.wikidata.org/wiki/Q428691","display_name":"Computer engineering","level":1,"score":0.36089348793029785},{"id":"https://openalex.org/C154945302","wikidata":"https://www.wikidata.org/wiki/Q11660","display_name":"Artificial intelligence","level":1,"score":0.3462945222854614},{"id":"https://openalex.org/C111919701","wikidata":"https://www.wikidata.org/wiki/Q9135","display_name":"Operating system","level":1,"score":0.09825468063354492},{"id":"https://openalex.org/C2777904410","wikidata":"https://www.wikidata.org/wiki/Q7397","display_name":"Software","level":2,"score":0.0952589213848114},{"id":"https://openalex.org/C86803240","wikidata":"https://www.wikidata.org/wiki/Q420","display_name":"Biology","level":0,"score":0.0},{"id":"https://openalex.org/C76155785","wikidata":"https://www.wikidata.org/wiki/Q418","display_name":"Telecommunications","level":1,"score":0.0},{"id":"https://openalex.org/C6557445","wikidata":"https://www.wikidata.org/wiki/Q173113","display_name":"Agronomy","level":1,"score":0.0}],"mesh":[],"locations_count":1,"locations":[{"id":"doi:10.1109/icc45041.2023.10278567","is_oa":false,"landing_page_url":"https://doi.org/10.1109/icc45041.2023.10278567","pdf_url":null,"source":null,"license":null,"license_id":null,"version":"publishedVersion","is_accepted":true,"is_published":true,"raw_source_name":"ICC 2023 - IEEE International Conference on Communications","raw_type":"proceedings-article"}],"best_oa_location":null,"sustainable_development_goals":[{"display_name":"Affordable and clean energy","id":"https://metadata.un.org/sdg/7","score":0.8500000238418579}],"awards":[{"id":"https://openalex.org/G6507269736","display_name":null,"funder_award_id":"2021YFA1003602","funder_id":"https://openalex.org/F4320335777","funder_display_name":"National Key Research and Development Program of China"}],"funders":[{"id":"https://openalex.org/F4320335777","display_name":"National Key Research and Development Program of China","ror":null}],"has_content":{"grobid_xml":false,"pdf":false},"content_urls":null,"referenced_works_count":28,"referenced_works":["https://openalex.org/W2468983453","https://openalex.org/W2767785892","https://openalex.org/W2885311373","https://openalex.org/W2896457183","https://openalex.org/W2951104886","https://openalex.org/W2963918968","https://openalex.org/W2965373594","https://openalex.org/W2965658867","https://openalex.org/W2975381464","https://openalex.org/W2976833415","https://openalex.org/W2979455536","https://openalex.org/W2982479999","https://openalex.org/W2991232789","https://openalex.org/W3008408165","https://openalex.org/W3034100311","https://openalex.org/W3034560159","https://openalex.org/W3107893198","https://openalex.org/W3206749822","https://openalex.org/W3212505503","https://openalex.org/W4205610676","https://openalex.org/W4205952419","https://openalex.org/W4226439885","https://openalex.org/W4288347855","https://openalex.org/W4320029498","https://openalex.org/W6752515464","https://openalex.org/W6753278433","https://openalex.org/W6762945437","https://openalex.org/W6774496999"],"related_works":["https://openalex.org/W4315697128","https://openalex.org/W3102845713","https://openalex.org/W2971502891","https://openalex.org/W3205506801","https://openalex.org/W4280599700","https://openalex.org/W3183570023","https://openalex.org/W4382323155","https://openalex.org/W2016508734","https://openalex.org/W4292794827","https://openalex.org/W4224939635"],"abstract_inverted_index":{"Transformer-based":[0],"models":[1,37,67,108],"suffer":[2],"from":[3],"large":[4],"num-ber":[5],"of":[6,77,83,151,167],"parameters":[7],"and":[8,59,128,148,169,208],"high":[9,25],"inference":[10,26,152],"latency,":[11],"whose":[12],"deployment":[13,122],"are":[14,52,94,137],"not":[15,88],"green":[16,120],"due":[17],"to":[18,34,54,64,105,109,144],"the":[19,56,66,107,110,114,146,176,194,221],"potential":[20],"environmental":[21],"damage":[22],"caused":[23],"by":[24,139],"energy":[27],"consumption.":[28],"In":[29],"addition,":[30],"it":[31],"is":[32,87],"difficult":[33],"deploy":[35],"such":[36,45,71],"on":[38,41,68,124,153,218],"devices,":[39],"especially":[40],"resource":[42,60],"constrained":[43],"devices":[44],"as":[46,63,80],"FPGA.":[47],"Various":[48],"model":[49,57,212],"pruning":[50,92,166],"methods":[51,72,93],"proposed":[53],"shrink":[55],"size":[58,213],"consumption,":[61],"so":[62],"fit":[65],"hardware.":[69],"However,":[70],"often":[73],"introduce":[74],"floating":[75],"point":[76],"operations":[78],"(FLOPs)":[79],"an":[81],"agent":[82],"hardware":[84],"performance,":[85],"which":[86,103,155],"accurate.":[89],"Furthermore,":[90],"structural":[91],"always":[95],"in":[96],"a":[97,119,187],"single":[98],"head-wise":[99],"or":[100],"layer-wise":[101],"pattern,":[102],"fails":[104],"compress":[106],"extreme.":[111],"To":[112,173],"resolve":[113],"above":[115],"issues,":[116],"we":[117,163],"propose":[118],"BERT":[121],"method":[123],"FPGA":[125],"via":[126],"hardware-aware":[127,135],"hybrid":[129],"pruning,":[130],"named":[131],"g-BERT.":[132],"Specifically,":[133],"two":[134],"metrics":[136],"introduced":[138],"High":[140],"Level":[141],"Synthesis":[142],"(HLS)":[143],"evaluate":[145],"latency":[147],"power":[149,205],"consumption":[150,206],"FPGA,":[154],"can":[156],"be":[157],"optimized":[158],"directly":[159],"while":[160],"pruning.":[161],"Moreover,":[162],"simultaneously":[164],"consider":[165],"heads":[168],"full":[170],"encoder":[171],"layers.":[172],"efficiently":[174],"find":[175],"optimal":[177],"structure,":[178],"g-BERT":[179,196],"applies":[180],"differentiable":[181],"neural":[182],"architecture":[183],"search":[184],"(NAS)":[185],"with":[186,193,215,220],"special":[188],"0\u20131":[189],"loss":[190],"function.":[191],"Compared":[192],"BERT-base,":[195],"achieves":[197],"<tex":[198,202,209],"xmlns:mml=\"http://www.w3.org/1998/Math/MathML\"":[199,203,210],"xmlns:xlink=\"http://www.w3.org/1999/xlink\">$2.1\\times$</tex>":[200],"speedup,":[201],"xmlns:xlink=\"http://www.w3.org/1999/xlink\">$1.9\\times$</tex>":[204],"reduction":[207,214],"xmlns:xlink=\"http://www.w3.org/1999/xlink\">$1.8\\times$</tex>":[211],"comparable":[216],"accuracy,":[217],"par":[219],"state-of-the-art":[222],"methods.":[223]},"counts_by_year":[{"year":2025,"cited_by_count":2}],"updated_date":"2026-07-14T23:27:15.235271","created_date":"2025-10-10T00:00:00"}