{"id":"https://openalex.org/W4416157473","doi":"https://doi.org/10.1109/asp-dac66049.2026.11420756","title":"Precision-Scalable Microscaling Datapaths with Optimized Reduction Tree for Efficient NPU Integration","display_name":"Precision-Scalable Microscaling Datapaths with Optimized Reduction Tree for Efficient NPU Integration","publication_year":2026,"publication_date":"2026-01-19","ids":{"openalex":"https://openalex.org/W4416157473","doi":"https://doi.org/10.1109/asp-dac66049.2026.11420756"},"language":null,"primary_location":{"id":"doi:10.1109/asp-dac66049.2026.11420756","is_oa":false,"landing_page_url":"https://doi.org/10.1109/asp-dac66049.2026.11420756","pdf_url":null,"source":null,"license":null,"license_id":null,"version":"publishedVersion","is_accepted":true,"is_published":true,"raw_source_name":"2026 31st Asia and South Pacific Design Automation Conference (ASP-DAC)","raw_type":"proceedings-article"},"type":"conference-paper","indexed_in":["arxiv","crossref","datacite"],"open_access":{"is_oa":true,"oa_status":"green","oa_url":"https://lirias.kuleuven.be/handle/20.500.12942/777340","any_repository_has_fulltext":true},"authorships":[{"author_position":"first","author":{"id":"https://openalex.org/A5120076836","display_name":"Stef Cuyckens","orcid":null},"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":"Stef Cuyckens","raw_affiliation_strings":["KU Leuven,ESAT-MICAS,Belgium"],"raw_orcid":null,"affiliations":[{"raw_affiliation_string":"KU Leuven,ESAT-MICAS,Belgium","institution_ids":["https://openalex.org/I99464096"]}]},{"author_position":"middle","author":{"id":"https://openalex.org/A5061006105","display_name":"Xiaoling Yi","orcid":"https://orcid.org/0009-0000-9527-7595"},"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":"Xiaoling Yi","raw_affiliation_strings":["KU Leuven,ESAT-MICAS,Belgium"],"raw_orcid":null,"affiliations":[{"raw_affiliation_string":"KU Leuven,ESAT-MICAS,Belgium","institution_ids":["https://openalex.org/I99464096"]}]},{"author_position":"middle","author":{"id":"https://openalex.org/A5065013150","display_name":"Robin Geens","orcid":null},"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":"Robin Geens","raw_affiliation_strings":["KU Leuven,ESAT-MICAS,Belgium"],"raw_orcid":null,"affiliations":[{"raw_affiliation_string":"KU Leuven,ESAT-MICAS,Belgium","institution_ids":["https://openalex.org/I99464096"]}]},{"author_position":"middle","author":{"id":"https://openalex.org/A5099841785","display_name":"Joren Dumoulin","orcid":"https://orcid.org/0009-0005-0692-1227"},"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":"Joren Dumoulin","raw_affiliation_strings":["KU Leuven,ESAT-MICAS,Belgium"],"raw_orcid":null,"affiliations":[{"raw_affiliation_string":"KU Leuven,ESAT-MICAS,Belgium","institution_ids":["https://openalex.org/I99464096"]}]},{"author_position":"middle","author":{"id":null,"display_name":"Martin Wiesner","orcid":null},"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":"Martin Wiesner","raw_affiliation_strings":["KU Leuven,ESAT-MICAS,Belgium"],"raw_orcid":null,"affiliations":[{"raw_affiliation_string":"KU Leuven,ESAT-MICAS,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":["KU Leuven,ESAT-MICAS,Belgium"],"raw_orcid":null,"affiliations":[{"raw_affiliation_string":"KU Leuven,ESAT-MICAS,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":["KU Leuven,ESAT-MICAS,Belgium"],"raw_orcid":null,"affiliations":[{"raw_affiliation_string":"KU Leuven,ESAT-MICAS,Belgium","institution_ids":["https://openalex.org/I99464096"]}]}],"institutions":[],"countries_distinct_count":1,"institutions_distinct_count":1,"corresponding_author_ids":[],"corresponding_institution_ids":["https://openalex.org/I99464096"],"apc_list":null,"apc_paid":null,"fwci":14.1002,"has_fulltext":true,"cited_by_count":1,"citation_normalized_percentile":{"value":0.95184971,"is_in_top_1_percent":false,"is_in_top_10_percent":true},"cited_by_percentile_year":{"min":94,"max":97},"biblio":{"volume":null,"issue":null,"first_page":"611","last_page":"617"},"is_retracted":false,"is_paratext":false,"is_xpac":false,"primary_topic":{"id":"https://openalex.org/T10036","display_name":"Advanced Neural Network Applications","score":0.2370000034570694,"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":0.2370000034570694,"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/T10320","display_name":"Neural Networks and Applications","score":0.12380000203847885,"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/T12808","display_name":"Ferroelectric and Negative Capacitance Devices","score":0.10540000349283218,"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/reduction","display_name":"Reduction (mathematics)","score":0.6284999847412109},{"id":"https://openalex.org/keywords/quantization","display_name":"Quantization (signal processing)","score":0.5278000235557556},{"id":"https://openalex.org/keywords/tree","display_name":"Tree (set theory)","score":0.47909998893737793},{"id":"https://openalex.org/keywords/throughput","display_name":"Throughput","score":0.44589999318122864},{"id":"https://openalex.org/keywords/inference","display_name":"Inference","score":0.4178999960422516},{"id":"https://openalex.org/keywords/modular-design","display_name":"Modular design","score":0.3831999897956848},{"id":"https://openalex.org/keywords/artificial-neural-network","display_name":"Artificial neural network","score":0.38029998540878296}],"concepts":[{"id":"https://openalex.org/C41008148","wikidata":"https://www.wikidata.org/wiki/Q21198","display_name":"Computer science","level":0,"score":0.7657999992370605},{"id":"https://openalex.org/C111335779","wikidata":"https://www.wikidata.org/wiki/Q3454686","display_name":"Reduction (mathematics)","level":2,"score":0.6284999847412109},{"id":"https://openalex.org/C28855332","wikidata":"https://www.wikidata.org/wiki/Q198099","display_name":"Quantization (signal processing)","level":2,"score":0.5278000235557556},{"id":"https://openalex.org/C113174947","wikidata":"https://www.wikidata.org/wiki/Q2859736","display_name":"Tree (set theory)","level":2,"score":0.47909998893737793},{"id":"https://openalex.org/C157764524","wikidata":"https://www.wikidata.org/wiki/Q1383412","display_name":"Throughput","level":3,"score":0.44589999318122864},{"id":"https://openalex.org/C2776214188","wikidata":"https://www.wikidata.org/wiki/Q408386","display_name":"Inference","level":2,"score":0.4178999960422516},{"id":"https://openalex.org/C11413529","wikidata":"https://www.wikidata.org/wiki/Q8366","display_name":"Algorithm","level":1,"score":0.39980000257492065},{"id":"https://openalex.org/C101468663","wikidata":"https://www.wikidata.org/wiki/Q1620158","display_name":"Modular design","level":2,"score":0.3831999897956848},{"id":"https://openalex.org/C50644808","wikidata":"https://www.wikidata.org/wiki/Q192776","display_name":"Artificial neural network","level":2,"score":0.38029998540878296},{"id":"https://openalex.org/C97137487","wikidata":"https://www.wikidata.org/wiki/Q729138","display_name":"Integer (computer science)","level":2,"score":0.33970001339912415},{"id":"https://openalex.org/C113775141","wikidata":"https://www.wikidata.org/wiki/Q428691","display_name":"Computer engineering","level":1,"score":0.3366999924182892},{"id":"https://openalex.org/C149635348","wikidata":"https://www.wikidata.org/wiki/Q193040","display_name":"Embedded system","level":1,"score":0.30889999866485596},{"id":"https://openalex.org/C530198007","wikidata":"https://www.wikidata.org/wiki/Q80831","display_name":"Integrated circuit","level":2,"score":0.29190000891685486},{"id":"https://openalex.org/C79403827","wikidata":"https://www.wikidata.org/wiki/Q3988","display_name":"Real-time computing","level":1,"score":0.2782999873161316},{"id":"https://openalex.org/C163797641","wikidata":"https://www.wikidata.org/wiki/Q2067937","display_name":"Tree structure","level":3,"score":0.2700999975204468},{"id":"https://openalex.org/C173608175","wikidata":"https://www.wikidata.org/wiki/Q232661","display_name":"Parallel computing","level":1,"score":0.26269999146461487},{"id":"https://openalex.org/C2742236","wikidata":"https://www.wikidata.org/wiki/Q924713","display_name":"Efficient energy use","level":2,"score":0.25839999318122864},{"id":"https://openalex.org/C2776175482","wikidata":"https://www.wikidata.org/wiki/Q1195816","display_name":"Transfer (computing)","level":2,"score":0.25679999589920044},{"id":"https://openalex.org/C154945302","wikidata":"https://www.wikidata.org/wiki/Q11660","display_name":"Artificial intelligence","level":1,"score":0.25529998540878296},{"id":"https://openalex.org/C2778820799","wikidata":"https://www.wikidata.org/wiki/Q3454688","display_name":"Cost reduction","level":2,"score":0.2551000118255615},{"id":"https://openalex.org/C14580979","wikidata":"https://www.wikidata.org/wiki/Q876049","display_name":"Very-large-scale integration","level":2,"score":0.2524000108242035}],"mesh":[],"locations_count":4,"locations":[{"id":"doi:10.1109/asp-dac66049.2026.11420756","is_oa":false,"landing_page_url":"https://doi.org/10.1109/asp-dac66049.2026.11420756","pdf_url":null,"source":null,"license":null,"license_id":null,"version":"publishedVersion","is_accepted":true,"is_published":true,"raw_source_name":"2026 31st Asia and South Pacific Design Automation Conference (ASP-DAC)","raw_type":"proceedings-article"},{"id":"pmh:oai:lirias2repo.kuleuven.be:20.500.12942/777340","is_oa":true,"landing_page_url":"https://lirias.kuleuven.be/handle/20.500.12942/777340","pdf_url":null,"source":{"id":"https://openalex.org/S7407055369","display_name":"Lirias","issn_l":null,"issn":null,"is_oa":false,"is_in_doaj":false,"is_core":false,"host_organization":null,"host_organization_name":null,"host_organization_lineage":[],"host_organization_lineage_names":[],"type":"repository"},"license":"other-oa","license_id":"https://openalex.org/licenses/other-oa","version":"acceptedVersion","is_accepted":true,"is_published":false,"raw_source_name":"arXiv","raw_type":"info:eu-repo/semantics/preprint"},{"id":"pmh:oai:arXiv.org:2511.06313","is_oa":true,"landing_page_url":"http://arxiv.org/abs/2511.06313","pdf_url":"https://arxiv.org/pdf/2511.06313","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":"cc-by","license_id":"https://openalex.org/licenses/cc-by","version":"submittedVersion","is_accepted":false,"is_published":false,"raw_source_name":null,"raw_type":"text"},{"id":"doi:10.48550/arxiv.2511.06313","is_oa":true,"landing_page_url":"https://doi.org/10.48550/arxiv.2511.06313","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":"cc-by","license_id":"https://openalex.org/licenses/cc-by","version":null,"is_accepted":false,"is_published":null,"raw_source_name":null,"raw_type":"Preprint"}],"best_oa_location":{"id":"pmh:oai:lirias2repo.kuleuven.be:20.500.12942/777340","is_oa":true,"landing_page_url":"https://lirias.kuleuven.be/handle/20.500.12942/777340","pdf_url":null,"source":{"id":"https://openalex.org/S7407055369","display_name":"Lirias","issn_l":null,"issn":null,"is_oa":false,"is_in_doaj":false,"is_core":false,"host_organization":null,"host_organization_name":null,"host_organization_lineage":[],"host_organization_lineage_names":[],"type":"repository"},"license":"other-oa","license_id":"https://openalex.org/licenses/other-oa","version":"acceptedVersion","is_accepted":true,"is_published":false,"raw_source_name":"arXiv","raw_type":"info:eu-repo/semantics/preprint"},"sustainable_development_goals":[],"awards":[{"id":"https://openalex.org/G1841200666","display_name":"Seamless design of smart edge processors","funder_award_id":"101070374","funder_id":"https://openalex.org/F4320320300","funder_display_name":"European Commission"},{"id":"https://openalex.org/G1884731767","display_name":null,"funder_award_id":"101088865","funder_id":"https://openalex.org/F4320320300","funder_display_name":"European Commission"},{"id":"https://openalex.org/G434337789","display_name":null,"funder_award_id":"HORIZON2020","funder_id":"https://openalex.org/F4320320300","funder_display_name":"European Commission"}],"funders":[{"id":"https://openalex.org/F4320320300","display_name":"European Commission","ror":"https://ror.org/00k4n6c32"},{"id":"https://openalex.org/F4320321730","display_name":"Fonds Wetenschappelijk Onderzoek","ror":"https://ror.org/03qtxy027"},{"id":"https://openalex.org/F4320322308","display_name":"KU Leuven","ror":"https://ror.org/05f950310"},{"id":"https://openalex.org/F4320327336","display_name":"Vlaamse regering","ror":null},{"id":"https://openalex.org/F4320334678","display_name":"European Research Council","ror":"https://ror.org/0472cxd90"}],"has_content":{"grobid_xml":false,"pdf":false},"content_urls":null,"referenced_works_count":29,"referenced_works":["https://openalex.org/W4411614560","https://openalex.org/W4205808347","https://openalex.org/W4406457544","https://openalex.org/W4226256030","https://openalex.org/W4402543980","https://openalex.org/W4386867450","https://openalex.org/W4387967984","https://openalex.org/W2920242713","https://openalex.org/W3036878841","https://openalex.org/W4409262453","https://openalex.org/W3090049941","https://openalex.org/W4393141038","https://openalex.org/W4416963812","https://openalex.org/W3213775513","https://openalex.org/W4393406920","https://openalex.org/W4391429102","https://openalex.org/W3016829100","https://openalex.org/W4413361539","https://openalex.org/W4408151419","https://openalex.org/W4409248488","https://openalex.org/W4312531383","https://openalex.org/W4416963670","https://openalex.org/W4414197189","https://openalex.org/W2987956272","https://openalex.org/W4400315128","https://openalex.org/W3013692244","https://openalex.org/W2593564159","https://openalex.org/W4220837341","https://openalex.org/W4401211881"],"related_works":[],"abstract_inverted_index":{"Emerging":[0],"continual":[1],"learning":[2],"applications":[3],"necessitate":[4],"next-generation":[5],"neural":[6],"processing":[7],"unit":[8],"(NPU)":[9],"platforms":[10],"to":[11,109,116,135],"support":[12],"both":[13,82,126],"training":[14],"and":[15,28,60,113,129,132,148,155,163],"inference":[16,27],"operations.":[17],"The":[18],"promising":[19],"Microscaling":[20],"(MX)":[21],"standard":[22],"enables":[23],"narrow":[24,50],"bit-widths":[25],"for":[26,32,74,152],"large":[29],"dynamic":[30],"ranges":[31],"training.":[33],"However,":[34],"existing":[35],"MX":[36,75,120],"multiply-accumulate":[37],"(MAC)":[38],"designs":[39],"face":[40],"a":[41,69,158],"critical":[42],"trade-off:":[43],"integer":[44],"accumulation":[45,55,87],"requires":[46],"expensive":[47],"conversions":[48],"from":[49,57],"floating-point":[51],"products,":[52],"while":[53],"FP32":[54],"suffers":[56],"quantization":[58],"losses":[59],"costly":[61],"normalization.":[62],"To":[63],"address":[64],"these":[65,99],"limitations,":[66],"we":[67,93],"propose":[68],"hybrid":[70],"precision-scalable":[71,119],"reduction":[72],"tree":[73],"MACs":[76,100],"that":[77],"combines":[78],"the":[79,102,136],"benefits":[80],"of":[81,98,145,160],"approaches,":[83],"enabling":[84],"efficient":[85,111],"mixed-precision":[86],"with":[88,157],"controlled":[89],"accuracy":[90],"relaxation.":[91],"Moreover,":[92],"integrate":[94],"an":[95,142],"8x8":[96],"array":[97],"into":[101],"state-of-the-art":[103],"(SotA)":[104],"NPU":[105],"integration":[106],"platform,":[107],"SNAX,":[108],"provide":[110],"control":[112],"data":[114],"transfer":[115],"our":[117,124],"optimized":[118],"datapath.":[121],"We":[122],"evaluate":[123],"design":[125],"on":[127],"MAC":[128],"system":[130,140],"level":[131],"compare":[133],"it":[134],"SotA.":[137],"Our":[138],"integrated":[139],"achieves":[141],"energy":[143],"efficiency":[144],"657,":[146],"1438-1675,":[147],"4065":[149],"GOPS/W,":[150],"respectively,":[151],"MXINT8,":[153],"MXFP8/6,":[154],"MXFP4,":[156],"throughput":[159],"64,":[161],"256,":[162],"512":[164],"GOPS.":[165]},"counts_by_year":[{"year":2026,"cited_by_count":1}],"updated_date":"2026-08-18T07:49:30.821534","created_date":"2025-11-12T00:00:00"}