{"id":"https://openalex.org/W4416922276","doi":"https://doi.org/10.1109/jssc.2025.3636451","title":"SparseCol: A 1320 BTOPS/W Precision-Scalable NPU Exploiting Training-Free Structured Bit-Level Sparsity and Dynamic Dataflow","display_name":"SparseCol: A 1320 BTOPS/W Precision-Scalable NPU Exploiting Training-Free Structured Bit-Level Sparsity and Dynamic Dataflow","publication_year":2025,"publication_date":"2025-12-02","ids":{"openalex":"https://openalex.org/W4416922276","doi":"https://doi.org/10.1109/jssc.2025.3636451"},"language":"en","primary_location":{"id":"doi:10.1109/jssc.2025.3636451","is_oa":false,"landing_page_url":"https://doi.org/10.1109/jssc.2025.3636451","pdf_url":null,"source":{"id":"https://openalex.org/S83637746","display_name":"IEEE Journal of Solid-State Circuits","issn_l":"0018-9200","issn":["0018-9200","1558-173X"],"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 Journal of Solid-State Circuits","raw_type":"journal-article"},"type":"article","indexed_in":["arxiv","crossref","datacite"],"open_access":{"is_oa":true,"oa_status":"green","oa_url":"https://arxiv.org/pdf/2606.16016","any_repository_has_fulltext":true},"authorships":[{"author_position":"first","author":{"id":"https://openalex.org/A5043501540","display_name":"Man Shi","orcid":"https://orcid.org/0000-0003-1675-6235"},"institutions":[{"id":"https://openalex.org/I99464096","display_name":"KU Leuven","ror":"https://ror.org/05f950310","country_code":"BE","type":"education","lineage":["https://openalex.org/I99464096"]}],"countries":["BE"],"is_corresponding":false,"raw_author_name":"Man Shi","raw_affiliation_strings":["ESAT, KU Leuven, Leuven, Belgium"],"raw_orcid":"https://orcid.org/0000-0003-1675-6235","affiliations":[{"raw_affiliation_string":"ESAT, KU Leuven, Leuven, Belgium","institution_ids":["https://openalex.org/I99464096"]}]},{"author_position":"middle","author":{"id":"https://openalex.org/A5074485526","display_name":"Vikram Jain","orcid":"https://orcid.org/0000-0002-1267-1683"},"institutions":[{"id":"https://openalex.org/I99464096","display_name":"KU Leuven","ror":"https://ror.org/05f950310","country_code":"BE","type":"education","lineage":["https://openalex.org/I99464096"]}],"countries":["BE"],"is_corresponding":false,"raw_author_name":"Vikram Jain","raw_affiliation_strings":["ESAT, KU Leuven, Leuven, Belgium"],"raw_orcid":"https://orcid.org/0000-0002-1267-1683","affiliations":[{"raw_affiliation_string":"ESAT, KU Leuven, Leuven, Belgium","institution_ids":["https://openalex.org/I99464096"]}]},{"author_position":"middle","author":{"id":"https://openalex.org/A5089286645","display_name":"Weijie Jiang","orcid":"https://orcid.org/0000-0003-3686-4639"},"institutions":[{"id":"https://openalex.org/I99464096","display_name":"KU Leuven","ror":"https://ror.org/05f950310","country_code":"BE","type":"education","lineage":["https://openalex.org/I99464096"]}],"countries":["BE"],"is_corresponding":false,"raw_author_name":"Weijie Jiang","raw_affiliation_strings":["ESAT, KU Leuven, Leuven, Belgium"],"raw_orcid":null,"affiliations":[{"raw_affiliation_string":"ESAT, KU Leuven, Leuven, Belgium","institution_ids":["https://openalex.org/I99464096"]}]},{"author_position":"middle","author":{"id":"https://openalex.org/A5066888122","display_name":"Chao Fang","orcid":"https://orcid.org/0000-0003-3430-1189"},"institutions":[{"id":"https://openalex.org/I99464096","display_name":"KU Leuven","ror":"https://ror.org/05f950310","country_code":"BE","type":"education","lineage":["https://openalex.org/I99464096"]}],"countries":["BE"],"is_corresponding":false,"raw_author_name":"Chao Fang","raw_affiliation_strings":["ESAT, KU Leuven, Leuven, Belgium"],"raw_orcid":"https://orcid.org/0000-0003-3430-1189","affiliations":[{"raw_affiliation_string":"ESAT, KU Leuven, Leuven, Belgium","institution_ids":["https://openalex.org/I99464096"]}]},{"author_position":"middle","author":{"id":"https://openalex.org/A5110800720","display_name":"Antony Joseph","orcid":null},"institutions":[{"id":"https://openalex.org/I4210134434","display_name":"NXP (Belgium)","ror":"https://ror.org/031v4g827","country_code":"BE","type":"company","lineage":["https://openalex.org/I109147379","https://openalex.org/I4210134434"]}],"countries":["BE"],"is_corresponding":false,"raw_author_name":"Antony Joseph","raw_affiliation_strings":["NXP Semiconductor, Leuven, Belgium"],"raw_orcid":null,"affiliations":[{"raw_affiliation_string":"NXP Semiconductor, Leuven, Belgium","institution_ids":["https://openalex.org/I4210134434"]}]},{"author_position":"middle","author":{"id":"https://openalex.org/A5076274517","display_name":"Wim Dehaene","orcid":"https://orcid.org/0000-0002-6792-7965"},"institutions":[{"id":"https://openalex.org/I99464096","display_name":"KU Leuven","ror":"https://ror.org/05f950310","country_code":"BE","type":"education","lineage":["https://openalex.org/I99464096"]}],"countries":["BE"],"is_corresponding":false,"raw_author_name":"Wim Dehaene","raw_affiliation_strings":["ESAT, KU Leuven, Leuven, Belgium"],"raw_orcid":"https://orcid.org/0000-0002-6792-7965","affiliations":[{"raw_affiliation_string":"ESAT, KU Leuven, Leuven, Belgium","institution_ids":["https://openalex.org/I99464096"]}]},{"author_position":"last","author":{"id":"https://openalex.org/A5012150553","display_name":"Marian Verhelst","orcid":"https://orcid.org/0000-0003-3495-9263"},"institutions":[{"id":"https://openalex.org/I99464096","display_name":"KU Leuven","ror":"https://ror.org/05f950310","country_code":"BE","type":"education","lineage":["https://openalex.org/I99464096"]}],"countries":["BE"],"is_corresponding":false,"raw_author_name":"Marian Verhelst","raw_affiliation_strings":["ESAT, KU Leuven, Leuven, Belgium"],"raw_orcid":"https://orcid.org/0000-0003-3495-9263","affiliations":[{"raw_affiliation_string":"ESAT, KU Leuven, Leuven, Belgium","institution_ids":["https://openalex.org/I99464096"]}]}],"institutions":[],"countries_distinct_count":1,"institutions_distinct_count":2,"corresponding_author_ids":[],"corresponding_institution_ids":[],"apc_list":null,"apc_paid":null,"fwci":0.0,"has_fulltext":false,"cited_by_count":0,"citation_normalized_percentile":{"value":0.28488261,"is_in_top_1_percent":false,"is_in_top_10_percent":false},"cited_by_percentile_year":null,"biblio":{"volume":"61","issue":"7","first_page":"3759","last_page":"3772"},"is_retracted":false,"is_paratext":false,"is_xpac":false,"primary_topic":{"id":"https://openalex.org/T10036","display_name":"Advanced Neural Network Applications","score":0.3695000112056732,"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":0.3695000112056732,"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":0.15880000591278076,"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/T10363","display_name":"Low-power high-performance VLSI design","score":0.08709999918937683,"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/dataflow","display_name":"Dataflow","score":0.7134000062942505},{"id":"https://openalex.org/keywords/scalability","display_name":"Scalability","score":0.6837000250816345},{"id":"https://openalex.org/keywords/computation","display_name":"Computation","score":0.6442000269889832},{"id":"https://openalex.org/keywords/overhead","display_name":"Overhead (engineering)","score":0.4629000127315521},{"id":"https://openalex.org/keywords/leverage","display_name":"Leverage (statistics)","score":0.451200008392334},{"id":"https://openalex.org/keywords/exploit","display_name":"Exploit","score":0.435699999332428},{"id":"https://openalex.org/keywords/scheduling","display_name":"Scheduling (production processes)","score":0.39489999413490295},{"id":"https://openalex.org/keywords/convolutional-neural-network","display_name":"Convolutional neural network","score":0.3822000026702881},{"id":"https://openalex.org/keywords/bottleneck","display_name":"Bottleneck","score":0.3677999973297119}],"concepts":[{"id":"https://openalex.org/C41008148","wikidata":"https://www.wikidata.org/wiki/Q21198","display_name":"Computer science","level":0,"score":0.8317000269889832},{"id":"https://openalex.org/C96324660","wikidata":"https://www.wikidata.org/wiki/Q205446","display_name":"Dataflow","level":2,"score":0.7134000062942505},{"id":"https://openalex.org/C48044578","wikidata":"https://www.wikidata.org/wiki/Q727490","display_name":"Scalability","level":2,"score":0.6837000250816345},{"id":"https://openalex.org/C45374587","wikidata":"https://www.wikidata.org/wiki/Q12525525","display_name":"Computation","level":2,"score":0.6442000269889832},{"id":"https://openalex.org/C173608175","wikidata":"https://www.wikidata.org/wiki/Q232661","display_name":"Parallel computing","level":1,"score":0.550599992275238},{"id":"https://openalex.org/C2779960059","wikidata":"https://www.wikidata.org/wiki/Q7113681","display_name":"Overhead (engineering)","level":2,"score":0.4629000127315521},{"id":"https://openalex.org/C153083717","wikidata":"https://www.wikidata.org/wiki/Q6535263","display_name":"Leverage (statistics)","level":2,"score":0.451200008392334},{"id":"https://openalex.org/C165696696","wikidata":"https://www.wikidata.org/wiki/Q11287","display_name":"Exploit","level":2,"score":0.435699999332428},{"id":"https://openalex.org/C113775141","wikidata":"https://www.wikidata.org/wiki/Q428691","display_name":"Computer engineering","level":1,"score":0.40639999508857727},{"id":"https://openalex.org/C118524514","wikidata":"https://www.wikidata.org/wiki/Q173212","display_name":"Computer architecture","level":1,"score":0.40070000290870667},{"id":"https://openalex.org/C206729178","wikidata":"https://www.wikidata.org/wiki/Q2271896","display_name":"Scheduling (production processes)","level":2,"score":0.39489999413490295},{"id":"https://openalex.org/C81363708","wikidata":"https://www.wikidata.org/wiki/Q17084460","display_name":"Convolutional neural network","level":2,"score":0.3822000026702881},{"id":"https://openalex.org/C2780513914","wikidata":"https://www.wikidata.org/wiki/Q18210350","display_name":"Bottleneck","level":2,"score":0.3677999973297119},{"id":"https://openalex.org/C9390403","wikidata":"https://www.wikidata.org/wiki/Q3966","display_name":"Computer hardware","level":1,"score":0.33009999990463257},{"id":"https://openalex.org/C48372109","wikidata":"https://www.wikidata.org/wiki/Q3913","display_name":"Binary number","level":2,"score":0.32359999418258667},{"id":"https://openalex.org/C75684735","wikidata":"https://www.wikidata.org/wiki/Q858810","display_name":"Big data","level":2,"score":0.3059999942779541},{"id":"https://openalex.org/C50644808","wikidata":"https://www.wikidata.org/wiki/Q192776","display_name":"Artificial neural network","level":2,"score":0.30489999055862427},{"id":"https://openalex.org/C42935608","wikidata":"https://www.wikidata.org/wiki/Q190411","display_name":"Field-programmable gate array","level":2,"score":0.2985000014305115},{"id":"https://openalex.org/C184596265","wikidata":"https://www.wikidata.org/wiki/Q2651576","display_name":"Model of computation","level":3,"score":0.2856000065803528},{"id":"https://openalex.org/C162319229","wikidata":"https://www.wikidata.org/wiki/Q175263","display_name":"Data structure","level":2,"score":0.28279998898506165},{"id":"https://openalex.org/C118505674","wikidata":"https://www.wikidata.org/wiki/Q42586063","display_name":"Encoder","level":2,"score":0.28029999136924744},{"id":"https://openalex.org/C108583219","wikidata":"https://www.wikidata.org/wiki/Q197536","display_name":"Deep learning","level":2,"score":0.2766000032424927},{"id":"https://openalex.org/C164620267","wikidata":"https://www.wikidata.org/wiki/Q376953","display_name":"Adder","level":3,"score":0.27300000190734863},{"id":"https://openalex.org/C489000","wikidata":"https://www.wikidata.org/wiki/Q747385","display_name":"Data flow diagram","level":2,"score":0.26910001039505005},{"id":"https://openalex.org/C2779602883","wikidata":"https://www.wikidata.org/wiki/Q15544750","display_name":"Memory architecture","level":2,"score":0.2687999904155731},{"id":"https://openalex.org/C99821215","wikidata":"https://www.wikidata.org/wiki/Q1136583","display_name":"Swap (finance)","level":2,"score":0.2662999927997589},{"id":"https://openalex.org/C120314980","wikidata":"https://www.wikidata.org/wiki/Q180634","display_name":"Distributed computing","level":1,"score":0.26600000262260437},{"id":"https://openalex.org/C11413529","wikidata":"https://www.wikidata.org/wiki/Q8366","display_name":"Algorithm","level":1,"score":0.2653999924659729},{"id":"https://openalex.org/C66322947","wikidata":"https://www.wikidata.org/wiki/Q11658","display_name":"Transformer","level":3,"score":0.2547999918460846},{"id":"https://openalex.org/C80444323","wikidata":"https://www.wikidata.org/wiki/Q2878974","display_name":"Theoretical computer science","level":1,"score":0.2542000114917755}],"mesh":[],"locations_count":4,"locations":[{"id":"doi:10.1109/jssc.2025.3636451","is_oa":false,"landing_page_url":"https://doi.org/10.1109/jssc.2025.3636451","pdf_url":null,"source":{"id":"https://openalex.org/S83637746","display_name":"IEEE Journal of Solid-State Circuits","issn_l":"0018-9200","issn":["0018-9200","1558-173X"],"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 Journal of Solid-State Circuits","raw_type":"journal-article"},{"id":"pmh:oai:lirias2repo.kuleuven.be:20.500.12942/778666","is_oa":false,"landing_page_url":"https://lirias.kuleuven.be/handle/20.500.12942/778666","pdf_url":null,"source":{"id":"https://openalex.org/S4306401954","display_name":"Lirias (KU Leuven)","issn_l":null,"issn":null,"is_oa":false,"is_in_doaj":false,"is_core":false,"host_organization":"https://openalex.org/I99464096","host_organization_name":"KU Leuven","host_organization_lineage":["https://openalex.org/I99464096"],"host_organization_lineage_names":[],"type":"repository"},"license":null,"license_id":null,"version":"acceptedVersion","is_accepted":true,"is_published":false,"raw_source_name":"Ieee Journal Of Solid-State Circuits, vol. 61 (7)","raw_type":"info:eu-repo/semantics/article"},{"id":"pmh:oai:arXiv.org:2606.16016","is_oa":true,"landing_page_url":"https://arxiv.org/abs/2606.16016","pdf_url":"https://arxiv.org/pdf/2606.16016","source":{"id":"https://openalex.org/S4306400194","display_name":"arXiv (Cornell University)","issn_l":"2331-8422","issn":["2331-8422"],"is_oa":true,"is_in_doaj":false,"is_core":false,"host_organization":"https://openalex.org/I205783295","host_organization_name":"Cornell University","host_organization_lineage":["https://openalex.org/I205783295"],"host_organization_lineage_names":[],"type":"repository"},"license":null,"license_id":null,"version":"submittedVersion","is_accepted":false,"is_published":false,"raw_source_name":null,"raw_type":"text"},{"id":"doi:10.48550/arxiv.2606.16016","is_oa":true,"landing_page_url":"https://doi.org/10.48550/arxiv.2606.16016","pdf_url":null,"source":{"id":"https://openalex.org/S4306400194","display_name":"arXiv (Cornell University)","issn_l":"2331-8422","issn":["2331-8422"],"is_oa":true,"is_in_doaj":false,"is_core":false,"host_organization":"https://openalex.org/I205783295","host_organization_name":"Cornell University","host_organization_lineage":["https://openalex.org/I205783295"],"host_organization_lineage_names":[],"type":"repository"},"license":null,"license_id":null,"version":null,"is_accepted":false,"is_published":null,"raw_source_name":null,"raw_type":"Preprint"}],"best_oa_location":{"id":"pmh:oai:arXiv.org:2606.16016","is_oa":true,"landing_page_url":"https://arxiv.org/abs/2606.16016","pdf_url":"https://arxiv.org/pdf/2606.16016","source":{"id":"https://openalex.org/S4306400194","display_name":"arXiv (Cornell University)","issn_l":"2331-8422","issn":["2331-8422"],"is_oa":true,"is_in_doaj":false,"is_core":false,"host_organization":"https://openalex.org/I205783295","host_organization_name":"Cornell University","host_organization_lineage":["https://openalex.org/I205783295"],"host_organization_lineage_names":[],"type":"repository"},"license":null,"license_id":null,"version":"submittedVersion","is_accepted":false,"is_published":false,"raw_source_name":null,"raw_type":"text"},"sustainable_development_goals":[],"awards":[],"funders":[{"id":"https://openalex.org/F4320313460","display_name":"Agentschap Innoveren en Ondernemen","ror":"https://ror.org/032xdry56"}],"has_content":{"grobid_xml":false,"pdf":false},"content_urls":null,"referenced_works_count":0,"referenced_works":[],"related_works":[],"abstract_inverted_index":{"Bit-serial":[0],"computation":[1,43,60,90],"enables":[2],"sequential":[3],"processing":[4],"of":[5,66,192,208],"data":[6,84],"at":[7],"the":[8,42,95],"bit":[9,102],"level,":[10],"providing":[11],"several":[12],"advantages,":[13],"such":[14],"as":[15],"scalable":[16],"computational":[17],"precision.":[18],"This":[19],"approach":[20],"has":[21],"gained":[22],"significant":[23],"attention,":[24],"especially":[25],"for":[26],"exploiting":[27],"bit-level":[28,83],"sparsity":[29,103],"(BLS)":[30],"in":[31,153,158,190],"AI":[32,111],"workloads.":[33],"While":[34],"current":[35],"bit-serial":[36,155],"processors":[37,189],"leverage":[38],"BLS":[39],"to":[40,81,99,123],"eliminate":[41],"associated":[44],"with":[45],"zero":[46],"bits,":[47],"they":[48,54,70],"face":[49],"a":[50,141],"fundamental":[51],"tradeoff:":[52],"either":[53],"suffer":[55],"from":[56,75],"low":[57],"memory-access":[58],"and":[59,85,89,119,202,211],"efficiency":[61,182,193],"caused":[62],"by":[63,131],"irregular":[64],"patterns":[65],"non-zero":[67],"bits":[68],"or":[69],"incur":[71],"substantial":[72],"area":[73],"overhead":[74],"complex":[76],"online":[77],"scheduling":[78],"mechanisms":[79],"required":[80],"reorganize":[82],"preserve":[86],"memory":[87],"access":[88],"regularity.":[91],"Therefore,":[92],"we":[93],"present":[94],"SparseCol":[96,126,139,162],"processor,":[97],"designed":[98],"harness":[100],"extensive":[101],"while":[104,183],"maintaining":[105,184],"high":[106],"hardware":[107,148],"utilization":[108],"across":[109],"various":[110],"applications,":[112],"including":[113],"convolutional":[114],"neural":[115],"networks":[116],"(CNNs),":[117],"RNNs,":[118],"Transformers.":[120],"In":[121],"contrast":[122],"traditional":[124],"methods,":[125],"exploits":[127],"structured":[128],"BLS,":[129],"denoted":[130],"bit-column":[132],"sparsity,":[133],"without":[134],"requiring":[135],"any":[136],"re-training.":[137],"Furthermore,":[138],"implements":[140],"dynamic":[142],"dataflow":[143],"(DF)":[144],"architecture":[145],"that":[146],"tackles":[147],"under-utilization":[149],"issues":[150],"commonly":[151],"found":[152],"existing":[154],"solutions.":[156],"Fabricated":[157],"16-nm":[159],"CMOS":[160],"node,":[161],"delivers":[163],"1320":[164],"BTOPS/W":[165,210],"(BTOPS":[166],"represents":[167],"binary":[168],"tera-operations":[169],"per":[170],"second,":[171],"calculated":[172],"as$\\#\\text":[173],"{Wbits}":[174],"\\times":[175,178],"\\#\\text":[176],"{Abits}":[177],"\\text":[179],"{TOPS}$)":[180],"peak":[181],"accuracy,":[185],"outperforming":[186],"SoTA":[187],"sparse":[188],"terms":[191],"by$6.8{\\times":[194],"}$.":[195],"Comprehensive":[196],"evaluations":[197],"on":[198],"CNN":[199],"classification":[200],"tasks":[201],"Transformer":[203],"architectures":[204],"demonstrate":[205],"system-level":[206],"efficiencies":[207],"745.02":[209],"850.5":[212],"BTOPS/W,":[213],"respectively.":[214]},"counts_by_year":[],"updated_date":"2026-07-29T09:40:50.615796","created_date":"2025-12-02T00:00:00"}