{"id":"https://openalex.org/W3198609354","doi":"https://doi.org/10.1109/lca.2021.3108505","title":"SmaQ: Smart Quantization for DNN Training by Exploiting Value Clustering","display_name":"SmaQ: Smart Quantization for DNN Training by Exploiting Value Clustering","publication_year":2021,"publication_date":"2021-07-01","ids":{"openalex":"https://openalex.org/W3198609354","doi":"https://doi.org/10.1109/lca.2021.3108505","mag":"3198609354"},"language":"en","primary_location":{"id":"doi:10.1109/lca.2021.3108505","is_oa":false,"landing_page_url":"https://doi.org/10.1109/lca.2021.3108505","pdf_url":null,"source":{"id":"https://openalex.org/S17643076","display_name":"IEEE Computer Architecture Letters","issn_l":"1556-6056","issn":["1556-6056","1556-6064","2473-2575"],"is_oa":false,"is_in_doaj":false,"is_core":true,"host_organization":"https://openalex.org/P4310319808","host_organization_name":"Institute of Electrical and Electronics Engineers","host_organization_lineage":["https://openalex.org/P4310319808"],"host_organization_lineage_names":["Institute of Electrical and Electronics Engineers"],"type":"journal"},"license":null,"license_id":null,"version":"publishedVersion","is_accepted":true,"is_published":true,"raw_source_name":"IEEE Computer Architecture Letters","raw_type":"journal-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/A5006200831","display_name":"Nima Shoghi","orcid":"https://orcid.org/0000-0002-6846-8721"},"institutions":[{"id":"https://openalex.org/I130701444","display_name":"Georgia Institute of Technology","ror":"https://ror.org/01zkghx44","country_code":"US","type":"education","lineage":["https://openalex.org/I130701444"]}],"countries":["US"],"is_corresponding":false,"raw_author_name":"Nima Shoghi","raw_affiliation_strings":["Georgia Institute of Technology, Atlanta, GA, USA"],"raw_orcid":"https://orcid.org/0000-0002-6846-8721","affiliations":[{"raw_affiliation_string":"Georgia Institute of Technology, Atlanta, GA, USA","institution_ids":["https://openalex.org/I130701444"]}]},{"author_position":"middle","author":{"id":"https://openalex.org/A5010477922","display_name":"Andrei Bersatti","orcid":"https://orcid.org/0000-0002-8664-5533"},"institutions":[{"id":"https://openalex.org/I130701444","display_name":"Georgia Institute of Technology","ror":"https://ror.org/01zkghx44","country_code":"US","type":"education","lineage":["https://openalex.org/I130701444"]}],"countries":["US"],"is_corresponding":false,"raw_author_name":"Andrei Bersatti","raw_affiliation_strings":["Georgia Institute of Technology, Atlanta, GA, USA"],"raw_orcid":"https://orcid.org/0000-0002-8664-5533","affiliations":[{"raw_affiliation_string":"Georgia Institute of Technology, Atlanta, GA, USA","institution_ids":["https://openalex.org/I130701444"]}]},{"author_position":"middle","author":{"id":"https://openalex.org/A5082772077","display_name":"Moinuddin K. Qureshi","orcid":"https://orcid.org/0000-0002-1314-9096"},"institutions":[{"id":"https://openalex.org/I130701444","display_name":"Georgia Institute of Technology","ror":"https://ror.org/01zkghx44","country_code":"US","type":"education","lineage":["https://openalex.org/I130701444"]}],"countries":["US"],"is_corresponding":false,"raw_author_name":"Moinuddin Qureshi","raw_affiliation_strings":["Georgia Institute of Technology, Atlanta, GA, USA"],"raw_orcid":null,"affiliations":[{"raw_affiliation_string":"Georgia Institute of Technology, Atlanta, GA, USA","institution_ids":["https://openalex.org/I130701444"]}]},{"author_position":"last","author":{"id":"https://openalex.org/A5000822269","display_name":"Hyesoon Kim","orcid":"https://orcid.org/0000-0002-6061-7825"},"institutions":[{"id":"https://openalex.org/I130701444","display_name":"Georgia Institute of Technology","ror":"https://ror.org/01zkghx44","country_code":"US","type":"education","lineage":["https://openalex.org/I130701444"]}],"countries":["US"],"is_corresponding":false,"raw_author_name":"Hyesoon Kim","raw_affiliation_strings":["Georgia Institute of Technology, Atlanta, GA, USA"],"raw_orcid":"https://orcid.org/0000-0002-6061-7825","affiliations":[{"raw_affiliation_string":"Georgia Institute of Technology, Atlanta, GA, USA","institution_ids":["https://openalex.org/I130701444"]}]}],"institutions":[],"countries_distinct_count":1,"institutions_distinct_count":1,"corresponding_author_ids":[],"corresponding_institution_ids":["https://openalex.org/I130701444"],"apc_list":null,"apc_paid":null,"fwci":0.097,"has_fulltext":false,"cited_by_count":2,"citation_normalized_percentile":{"value":0.39994518,"is_in_top_1_percent":false,"is_in_top_10_percent":false},"cited_by_percentile_year":{"min":90,"max":95},"biblio":{"volume":"20","issue":"2","first_page":"126","last_page":"129"},"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/T10052","display_name":"Medical Image Segmentation Techniques","score":0.996399998664856,"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/T10627","display_name":"Advanced Image and Video Retrieval Techniques","score":0.9958000183105469,"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"}}],"keywords":[{"id":"https://openalex.org/keywords/quantization","display_name":"Quantization (signal processing)","score":0.7816643118858337},{"id":"https://openalex.org/keywords/computer-science","display_name":"Computer science","score":0.7692434787750244},{"id":"https://openalex.org/keywords/artificial-neural-network","display_name":"Artificial neural network","score":0.7301448583602905},{"id":"https://openalex.org/keywords/cluster-analysis","display_name":"Cluster analysis","score":0.7146027088165283},{"id":"https://openalex.org/keywords/bottleneck","display_name":"Bottleneck","score":0.6651909351348877},{"id":"https://openalex.org/keywords/exploit","display_name":"Exploit","score":0.5404051542282104},{"id":"https://openalex.org/keywords/artificial-intelligence","display_name":"Artificial intelligence","score":0.5105499029159546},{"id":"https://openalex.org/keywords/vector-quantization","display_name":"Vector quantization","score":0.5086191892623901},{"id":"https://openalex.org/keywords/algorithm","display_name":"Algorithm","score":0.4070969820022583},{"id":"https://openalex.org/keywords/pattern-recognition","display_name":"Pattern recognition (psychology)","score":0.3863109350204468}],"concepts":[{"id":"https://openalex.org/C28855332","wikidata":"https://www.wikidata.org/wiki/Q198099","display_name":"Quantization (signal processing)","level":2,"score":0.7816643118858337},{"id":"https://openalex.org/C41008148","wikidata":"https://www.wikidata.org/wiki/Q21198","display_name":"Computer science","level":0,"score":0.7692434787750244},{"id":"https://openalex.org/C50644808","wikidata":"https://www.wikidata.org/wiki/Q192776","display_name":"Artificial neural network","level":2,"score":0.7301448583602905},{"id":"https://openalex.org/C73555534","wikidata":"https://www.wikidata.org/wiki/Q622825","display_name":"Cluster analysis","level":2,"score":0.7146027088165283},{"id":"https://openalex.org/C2780513914","wikidata":"https://www.wikidata.org/wiki/Q18210350","display_name":"Bottleneck","level":2,"score":0.6651909351348877},{"id":"https://openalex.org/C165696696","wikidata":"https://www.wikidata.org/wiki/Q11287","display_name":"Exploit","level":2,"score":0.5404051542282104},{"id":"https://openalex.org/C154945302","wikidata":"https://www.wikidata.org/wiki/Q11660","display_name":"Artificial intelligence","level":1,"score":0.5105499029159546},{"id":"https://openalex.org/C199833920","wikidata":"https://www.wikidata.org/wiki/Q612536","display_name":"Vector quantization","level":2,"score":0.5086191892623901},{"id":"https://openalex.org/C11413529","wikidata":"https://www.wikidata.org/wiki/Q8366","display_name":"Algorithm","level":1,"score":0.4070969820022583},{"id":"https://openalex.org/C153180895","wikidata":"https://www.wikidata.org/wiki/Q7148389","display_name":"Pattern recognition (psychology)","level":2,"score":0.3863109350204468},{"id":"https://openalex.org/C149635348","wikidata":"https://www.wikidata.org/wiki/Q193040","display_name":"Embedded system","level":1,"score":0.0},{"id":"https://openalex.org/C38652104","wikidata":"https://www.wikidata.org/wiki/Q3510521","display_name":"Computer security","level":1,"score":0.0}],"mesh":[],"locations_count":1,"locations":[{"id":"doi:10.1109/lca.2021.3108505","is_oa":false,"landing_page_url":"https://doi.org/10.1109/lca.2021.3108505","pdf_url":null,"source":{"id":"https://openalex.org/S17643076","display_name":"IEEE Computer Architecture Letters","issn_l":"1556-6056","issn":["1556-6056","1556-6064","2473-2575"],"is_oa":false,"is_in_doaj":false,"is_core":true,"host_organization":"https://openalex.org/P4310319808","host_organization_name":"Institute of Electrical and Electronics Engineers","host_organization_lineage":["https://openalex.org/P4310319808"],"host_organization_lineage_names":["Institute of Electrical and Electronics Engineers"],"type":"journal"},"license":null,"license_id":null,"version":"publishedVersion","is_accepted":true,"is_published":true,"raw_source_name":"IEEE Computer Architecture Letters","raw_type":"journal-article"}],"best_oa_location":null,"sustainable_development_goals":[],"awards":[{"id":"https://openalex.org/G869176301","display_name":"CSR: Small:Collaborative Research: Decentralized Real-Time Machine Learning Systems on Near-User Edge Devices","funder_award_id":"1815047","funder_id":"https://openalex.org/F4320306076","funder_display_name":"National Science Foundation"}],"funders":[{"id":"https://openalex.org/F4320306076","display_name":"National Science Foundation","ror":"https://ror.org/021nxhr62"},{"id":"https://openalex.org/F4320308737","display_name":"Facebook","ror":"https://ror.org/01zbnvs85"}],"has_content":{"grobid_xml":false,"pdf":false},"content_urls":null,"referenced_works_count":24,"referenced_works":["https://openalex.org/W2053987390","https://openalex.org/W2183341477","https://openalex.org/W2194775991","https://openalex.org/W2803113791","https://openalex.org/W2889797931","https://openalex.org/W2896457183","https://openalex.org/W2908510526","https://openalex.org/W2923014074","https://openalex.org/W2947629474","https://openalex.org/W2953235111","https://openalex.org/W2962761403","https://openalex.org/W2963310665","https://openalex.org/W2963341956","https://openalex.org/W2963709416","https://openalex.org/W2994782216","https://openalex.org/W3000364559","https://openalex.org/W4288346545","https://openalex.org/W6750615492","https://openalex.org/W6751528251","https://openalex.org/W6753770798","https://openalex.org/W6755207826","https://openalex.org/W6757817989","https://openalex.org/W6763653508","https://openalex.org/W6771680064"],"related_works":["https://openalex.org/W2608884419","https://openalex.org/W2037550275","https://openalex.org/W3139547257","https://openalex.org/W2162675966","https://openalex.org/W1489554298","https://openalex.org/W1971930133","https://openalex.org/W4225773308","https://openalex.org/W2556145167","https://openalex.org/W2334580970","https://openalex.org/W127943105"],"abstract_inverted_index":{"Advancements":[0],"in":[1,20,160],"modern":[2],"deep":[3],"learning":[4],"have":[5,86],"shown":[6],"that":[7,74,104,143],"deeper":[8],"networks":[9,25,84],"with":[10,92,157],"larger":[11],"datasets":[12],"can":[13,85],"achieve":[14],"state":[15,63],"of":[16,31,41,52,77,126,142],"the":[17,28,38,50,75,78,112,120,140,148],"art":[18],"results":[19],"many":[21],"different":[22],"tasks.":[23],"As":[24],"become":[26],"deeper,":[27],"memory":[29,149],"requirement":[30],"neural":[32,53,68,83],"network":[33,54,69],"training":[34,152],"proves":[35],"to":[36,110,133,155],"be":[37,90],"primary":[39],"bottleneck":[40],"single-machine":[42],"training.":[43],"In":[44],"this":[45,106],"letter,":[46],"we":[47],"first":[48],"study":[49],"characteristics":[51],"weight,":[55],"gradient,":[56],"feature":[57],"map,":[58,60],"gradient":[59],"and":[61,123,128],"optimizer":[62],"distributions":[64,89],"for":[65],"some":[66],"popular":[67],"architectures.":[70],"Our":[71,115,145],"investigation":[72],"shows":[73],"majority":[76],"data":[79,113],"structures":[80],"used":[81],"by":[82,153],"their":[87],"value":[88],"approximated":[91],"normal":[93,108],"distributions.":[94],"We":[95],"then":[96],"introduce":[97],"Smart":[98],"Quantization":[99],"(SmaQ),":[100],"a":[101],"quantization":[102,117],"scheme":[103,146],"exploits":[105],"observed":[107],"distribution":[109],"quantize":[111],"structures.":[114],"dynamic":[116],"method":[118],"calculates":[119],"sampled":[121],"mean":[122],"standard":[124],"deviation":[125],"tensors":[127],"quantizes":[129],"each":[130],"tensor":[131],"element":[132],"6":[134],"or":[135],"8":[136],"bits":[137],"based":[138],"on":[139],"z-score":[141],"value.":[144],"reduces":[147],"usage":[150],"during":[151],"up":[154],"6.7x":[156],"minor":[158],"losses":[159],"accuracy.":[161]},"counts_by_year":[{"year":2025,"cited_by_count":1},{"year":2023,"cited_by_count":1}],"updated_date":"2026-06-26T08:34:08.712188","created_date":"2025-10-10T00:00:00"}