{"id":"https://openalex.org/W3174469560","doi":"https://doi.org/10.1109/aicas51828.2021.9458410","title":"HPPU: An Energy-Efficient Sparse DNN Training Processor with Hybrid Weight Pruning","display_name":"HPPU: An Energy-Efficient Sparse DNN Training Processor with Hybrid Weight Pruning","publication_year":2021,"publication_date":"2021-06-06","ids":{"openalex":"https://openalex.org/W3174469560","doi":"https://doi.org/10.1109/aicas51828.2021.9458410","mag":"3174469560"},"language":"en","primary_location":{"id":"doi:10.1109/aicas51828.2021.9458410","is_oa":false,"landing_page_url":"https://doi.org/10.1109/aicas51828.2021.9458410","pdf_url":null,"source":null,"license":null,"license_id":null,"version":"publishedVersion","is_accepted":true,"is_published":true,"raw_source_name":"2021 IEEE 3rd International Conference on Artificial Intelligence Circuits and Systems (AICAS)","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/A5100322887","display_name":"Yang Wang","orcid":"https://orcid.org/0000-0003-1029-9280"},"institutions":[{"id":"https://openalex.org/I4210119392","display_name":"Institute of Microelectronics","ror":"https://ror.org/02s6gs133","country_code":"CN","type":"facility","lineage":["https://openalex.org/I19820366","https://openalex.org/I4210119392"]},{"id":"https://openalex.org/I99065089","display_name":"Tsinghua University","ror":"https://ror.org/03cve4549","country_code":"CN","type":"education","lineage":["https://openalex.org/I99065089"]}],"countries":["CN"],"is_corresponding":true,"raw_author_name":"Yang Wang","raw_affiliation_strings":["Institute of Microelectronics, Tsinghua University, Beijing, China"],"affiliations":[{"raw_affiliation_string":"Institute of Microelectronics, Tsinghua University, Beijing, China","institution_ids":["https://openalex.org/I99065089","https://openalex.org/I4210119392"]}]},{"author_position":"middle","author":{"id":"https://openalex.org/A5071432313","display_name":"Yubin Qin","orcid":"https://orcid.org/0000-0001-5530-5416"},"institutions":[{"id":"https://openalex.org/I4210119392","display_name":"Institute of Microelectronics","ror":"https://ror.org/02s6gs133","country_code":"CN","type":"facility","lineage":["https://openalex.org/I19820366","https://openalex.org/I4210119392"]},{"id":"https://openalex.org/I99065089","display_name":"Tsinghua University","ror":"https://ror.org/03cve4549","country_code":"CN","type":"education","lineage":["https://openalex.org/I99065089"]}],"countries":["CN"],"is_corresponding":false,"raw_author_name":"Yubin Qin","raw_affiliation_strings":["Institute of Microelectronics, Tsinghua University, Beijing, China"],"affiliations":[{"raw_affiliation_string":"Institute of Microelectronics, Tsinghua University, Beijing, China","institution_ids":["https://openalex.org/I99065089","https://openalex.org/I4210119392"]}]},{"author_position":"middle","author":{"id":"https://openalex.org/A5100358856","display_name":"Leibo Liu","orcid":"https://orcid.org/0000-0001-7548-4116"},"institutions":[{"id":"https://openalex.org/I99065089","display_name":"Tsinghua University","ror":"https://ror.org/03cve4549","country_code":"CN","type":"education","lineage":["https://openalex.org/I99065089"]},{"id":"https://openalex.org/I4210119392","display_name":"Institute of Microelectronics","ror":"https://ror.org/02s6gs133","country_code":"CN","type":"facility","lineage":["https://openalex.org/I19820366","https://openalex.org/I4210119392"]}],"countries":["CN"],"is_corresponding":false,"raw_author_name":"Leibo Liu","raw_affiliation_strings":["Institute of Microelectronics, Tsinghua University, Beijing, China"],"affiliations":[{"raw_affiliation_string":"Institute of Microelectronics, Tsinghua University, Beijing, China","institution_ids":["https://openalex.org/I99065089","https://openalex.org/I4210119392"]}]},{"author_position":"middle","author":{"id":"https://openalex.org/A5036023084","display_name":"Shaojun Wei","orcid":"https://orcid.org/0000-0001-5117-7920"},"institutions":[{"id":"https://openalex.org/I99065089","display_name":"Tsinghua University","ror":"https://ror.org/03cve4549","country_code":"CN","type":"education","lineage":["https://openalex.org/I99065089"]},{"id":"https://openalex.org/I4210119392","display_name":"Institute of Microelectronics","ror":"https://ror.org/02s6gs133","country_code":"CN","type":"facility","lineage":["https://openalex.org/I19820366","https://openalex.org/I4210119392"]}],"countries":["CN"],"is_corresponding":false,"raw_author_name":"Shaojun Wei","raw_affiliation_strings":["Institute of Microelectronics, Tsinghua University, Beijing, China"],"affiliations":[{"raw_affiliation_string":"Institute of Microelectronics, Tsinghua University, Beijing, China","institution_ids":["https://openalex.org/I99065089","https://openalex.org/I4210119392"]}]},{"author_position":"last","author":{"id":"https://openalex.org/A5054524841","display_name":"Shouyi Yin","orcid":"https://orcid.org/0000-0003-2309-572X"},"institutions":[{"id":"https://openalex.org/I4210119392","display_name":"Institute of Microelectronics","ror":"https://ror.org/02s6gs133","country_code":"CN","type":"facility","lineage":["https://openalex.org/I19820366","https://openalex.org/I4210119392"]},{"id":"https://openalex.org/I99065089","display_name":"Tsinghua University","ror":"https://ror.org/03cve4549","country_code":"CN","type":"education","lineage":["https://openalex.org/I99065089"]}],"countries":["CN"],"is_corresponding":false,"raw_author_name":"Shouyi Yin","raw_affiliation_strings":["Institute of Microelectronics, Tsinghua University, Beijing, China"],"affiliations":[{"raw_affiliation_string":"Institute of Microelectronics, Tsinghua University, Beijing, China","institution_ids":["https://openalex.org/I99065089","https://openalex.org/I4210119392"]}]}],"institutions":[],"countries_distinct_count":1,"institutions_distinct_count":5,"corresponding_author_ids":["https://openalex.org/A5100322887"],"corresponding_institution_ids":["https://openalex.org/I4210119392","https://openalex.org/I99065089"],"apc_list":null,"apc_paid":null,"fwci":0.6725,"has_fulltext":false,"cited_by_count":9,"citation_normalized_percentile":{"value":0.70795752,"is_in_top_1_percent":false,"is_in_top_10_percent":false},"cited_by_percentile_year":{"min":89,"max":97},"biblio":{"volume":null,"issue":null,"first_page":"1","last_page":"4"},"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/T12676","display_name":"Machine Learning and ELM","score":0.9937999844551086,"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/T11307","display_name":"Domain Adaptation and Few-Shot Learning","score":0.9919999837875366,"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/pruning","display_name":"Pruning","score":0.8773013353347778},{"id":"https://openalex.org/keywords/computer-science","display_name":"Computer science","score":0.7415682077407837},{"id":"https://openalex.org/keywords/speedup","display_name":"Speedup","score":0.5862807035446167},{"id":"https://openalex.org/keywords/parallel-computing","display_name":"Parallel computing","score":0.5282340049743652},{"id":"https://openalex.org/keywords/computation","display_name":"Computation","score":0.47461003065109253},{"id":"https://openalex.org/keywords/convolutional-neural-network","display_name":"Convolutional neural network","score":0.45539751648902893},{"id":"https://openalex.org/keywords/artificial-neural-network","display_name":"Artificial neural network","score":0.4428616762161255},{"id":"https://openalex.org/keywords/computer-engineering","display_name":"Computer engineering","score":0.41034039855003357},{"id":"https://openalex.org/keywords/artificial-intelligence","display_name":"Artificial intelligence","score":0.397370308637619},{"id":"https://openalex.org/keywords/algorithm","display_name":"Algorithm","score":0.35852324962615967}],"concepts":[{"id":"https://openalex.org/C108010975","wikidata":"https://www.wikidata.org/wiki/Q500094","display_name":"Pruning","level":2,"score":0.8773013353347778},{"id":"https://openalex.org/C41008148","wikidata":"https://www.wikidata.org/wiki/Q21198","display_name":"Computer science","level":0,"score":0.7415682077407837},{"id":"https://openalex.org/C68339613","wikidata":"https://www.wikidata.org/wiki/Q1549489","display_name":"Speedup","level":2,"score":0.5862807035446167},{"id":"https://openalex.org/C173608175","wikidata":"https://www.wikidata.org/wiki/Q232661","display_name":"Parallel computing","level":1,"score":0.5282340049743652},{"id":"https://openalex.org/C45374587","wikidata":"https://www.wikidata.org/wiki/Q12525525","display_name":"Computation","level":2,"score":0.47461003065109253},{"id":"https://openalex.org/C81363708","wikidata":"https://www.wikidata.org/wiki/Q17084460","display_name":"Convolutional neural network","level":2,"score":0.45539751648902893},{"id":"https://openalex.org/C50644808","wikidata":"https://www.wikidata.org/wiki/Q192776","display_name":"Artificial neural network","level":2,"score":0.4428616762161255},{"id":"https://openalex.org/C113775141","wikidata":"https://www.wikidata.org/wiki/Q428691","display_name":"Computer engineering","level":1,"score":0.41034039855003357},{"id":"https://openalex.org/C154945302","wikidata":"https://www.wikidata.org/wiki/Q11660","display_name":"Artificial intelligence","level":1,"score":0.397370308637619},{"id":"https://openalex.org/C11413529","wikidata":"https://www.wikidata.org/wiki/Q8366","display_name":"Algorithm","level":1,"score":0.35852324962615967},{"id":"https://openalex.org/C6557445","wikidata":"https://www.wikidata.org/wiki/Q173113","display_name":"Agronomy","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/aicas51828.2021.9458410","is_oa":false,"landing_page_url":"https://doi.org/10.1109/aicas51828.2021.9458410","pdf_url":null,"source":null,"license":null,"license_id":null,"version":"publishedVersion","is_accepted":true,"is_published":true,"raw_source_name":"2021 IEEE 3rd International Conference on Artificial Intelligence Circuits and Systems (AICAS)","raw_type":"proceedings-article"}],"best_oa_location":null,"sustainable_development_goals":[{"display_name":"Affordable and clean energy","score":0.8999999761581421,"id":"https://metadata.un.org/sdg/7"}],"awards":[],"funders":[{"id":"https://openalex.org/F4320327539","display_name":"Beijing Innovation Center for Future Chip","ror":null}],"has_content":{"grobid_xml":false,"pdf":false},"content_urls":null,"referenced_works_count":10,"referenced_works":["https://openalex.org/W2899481200","https://openalex.org/W2907886210","https://openalex.org/W2921918777","https://openalex.org/W2949413532","https://openalex.org/W2949866178","https://openalex.org/W2963363373","https://openalex.org/W2970500560","https://openalex.org/W3015729306","https://openalex.org/W3048842964","https://openalex.org/W6757914502"],"related_works":["https://openalex.org/W1967627035","https://openalex.org/W156843270","https://openalex.org/W2757143157","https://openalex.org/W4294734199","https://openalex.org/W2007449167","https://openalex.org/W4310614714","https://openalex.org/W2770717529","https://openalex.org/W3131107792","https://openalex.org/W4289086683","https://openalex.org/W4297817880"],"abstract_inverted_index":{"Enlightened":[0],"by":[1,174],"the":[2,27,63,84,90,94,97,127,144,165],"fact":[3],"that":[4],"deep-neural-networks":[5],"(DNNs)":[6],"are":[7,31],"typically":[8],"highly":[9],"over-parameterized,":[10],"weight-pruning-based":[11],"sparse":[12,69,102,156],"training":[13,21,211],"(ST)":[14],"becomes":[15],"a":[16,34,38,44,49,80,107,121,134,149,155,212],"practical":[17],"method":[18],"to":[19,43,88,100,142],"reduce":[20],"computation":[22],"and":[23,93,116,154,190,220,224],"compress":[24],"models.":[25],"However,":[26],"previous":[28],"pruning":[29,46,109,123,185,189],"algorithms":[30],"either":[32],"with":[33,177],"coarse-grained":[35,188],"pattern":[36],"or":[37,48,57],"fine-grained":[39,195],"pattern.":[40],"They":[41],"lead":[42],"limited":[45],"ratio":[47,124,186],"drastically":[50],"irregular":[51],"sparsity":[52,167],"distribution,":[53],"which":[54,159],"is":[55,86,202],"computation-intensive":[56],"logic-complex":[58],"for":[59],"hardware":[60,91,95,128,135,161],"implementation.":[61],"Meanwhile,":[62],"current":[64],"DNN":[65],"processors":[66],"focus":[67],"on":[68],"inference":[70],"but":[71],"cannot":[72],"support":[73],"emerging":[74],"ST":[75],"techniques.":[76],"This":[77],"paper":[78],"proposes":[79],"co-design":[81],"approach":[82],"where":[83],"algorithm":[85,98],"adapted":[87],"suit":[89],"constraints":[92],"exploit":[96],"property":[99],"accelerate":[101,143],"training.":[103,170],"We":[104,171],"first":[105],"present":[106],"novel":[108],"algorithm,":[110],"hybrid":[111,166],"weight":[112],"pruning,":[113],"including":[114],"channel-wise":[115],"line-wise":[117],"pruning.":[118,196],"It":[119,147],"reaches":[120],"considerable":[122],"while":[125],"maintaining":[126],"friendly":[129],"property.":[130],"Then":[131],"we":[132],"design":[133],"architecture,":[136],"Hybrid":[137],"Pruning":[138],"Processing":[139],"Unit,":[140],"HPPU,":[141],"proposed":[145],"algorithm.":[146],"develops":[148],"2-level":[150],"active":[151],"data":[152],"selector":[153],"convolution":[157],"engine,":[158],"maximize":[160],"utilization":[162],"when":[163],"handling":[164],"patterns":[168],"during":[169],"evaluate":[172],"HPPU":[173,181,201,215],"synthesizing":[175],"it":[176],"28nm":[178],"CMOS":[179],"technology.":[180],"achieves":[182],"50.1%":[183],"higher":[184,192],"than":[187,194],"1.53\u00d7":[191],"energy-efficiency":[193,199],"The":[197],"peak":[198],"of":[200],"126.04TFLOPs/W,":[203],"outperforming":[204],"state-of-the-art":[205],"trainable":[206],"processor":[207],"GANPU":[208],"1.67\u00d7.":[209],"When":[210],"ResNet18":[213],"model,":[214],"consumes":[216],"3.72\u00d7":[217],"less":[218],"energy":[219],"offers":[221],"4.69\u00d7":[222],"speedup,":[223],"maintains":[225],"unpruned":[226],"accuracy.":[227]},"counts_by_year":[{"year":2025,"cited_by_count":2},{"year":2024,"cited_by_count":1},{"year":2023,"cited_by_count":2},{"year":2022,"cited_by_count":3},{"year":2021,"cited_by_count":1}],"updated_date":"2025-11-06T03:46:38.306776","created_date":"2025-10-10T00:00:00"}