{"id":"https://openalex.org/W4410584145","doi":"https://doi.org/10.23919/date64628.2025.10992712","title":"Distributed Inference with Minimal Off-Chip Traffic for Transformers on Low-Power MCUs","display_name":"Distributed Inference with Minimal Off-Chip Traffic for Transformers on Low-Power MCUs","publication_year":2025,"publication_date":"2025-03-31","ids":{"openalex":"https://openalex.org/W4410584145","doi":"https://doi.org/10.23919/date64628.2025.10992712"},"language":"en","primary_location":{"id":"doi:10.23919/date64628.2025.10992712","is_oa":false,"landing_page_url":"https://doi.org/10.23919/date64628.2025.10992712","pdf_url":null,"source":null,"license":null,"license_id":null,"version":"publishedVersion","is_accepted":true,"is_published":true,"raw_source_name":"2025 Design, Automation &amp;amp; Test in Europe Conference (DATE)","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/A5115024178","display_name":"Severin Bochem","orcid":null},"institutions":[{"id":"https://openalex.org/I35440088","display_name":"ETH Zurich","ror":"https://ror.org/05a28rw58","country_code":"CH","type":"education","lineage":["https://openalex.org/I2799323385","https://openalex.org/I35440088"]}],"countries":["CH"],"is_corresponding":true,"raw_author_name":"Severin Bochem","raw_affiliation_strings":["D-ITET, ETH Zurich,Switzerland"],"affiliations":[{"raw_affiliation_string":"D-ITET, ETH Zurich,Switzerland","institution_ids":["https://openalex.org/I35440088"]}]},{"author_position":"middle","author":{"id":"https://openalex.org/A5040009254","display_name":"Victor J. B. Jung","orcid":"https://orcid.org/0009-0001-7462-3468"},"institutions":[{"id":"https://openalex.org/I35440088","display_name":"ETH Zurich","ror":"https://ror.org/05a28rw58","country_code":"CH","type":"education","lineage":["https://openalex.org/I2799323385","https://openalex.org/I35440088"]}],"countries":["CH"],"is_corresponding":false,"raw_author_name":"Victor J.B. Jung","raw_affiliation_strings":["ETH Zurich,Integrated Systems Laboratory,Switzerland"],"affiliations":[{"raw_affiliation_string":"ETH Zurich,Integrated Systems Laboratory,Switzerland","institution_ids":["https://openalex.org/I35440088"]}]},{"author_position":"middle","author":{"id":"https://openalex.org/A5056197249","display_name":"Arpan Suravi Prasad","orcid":"https://orcid.org/0009-0009-6031-6668"},"institutions":[{"id":"https://openalex.org/I35440088","display_name":"ETH Zurich","ror":"https://ror.org/05a28rw58","country_code":"CH","type":"education","lineage":["https://openalex.org/I2799323385","https://openalex.org/I35440088"]}],"countries":["CH"],"is_corresponding":false,"raw_author_name":"Arpan Suravi Prasad","raw_affiliation_strings":["ETH Zurich,Integrated Systems Laboratory,Switzerland"],"affiliations":[{"raw_affiliation_string":"ETH Zurich,Integrated Systems Laboratory,Switzerland","institution_ids":["https://openalex.org/I35440088"]}]},{"author_position":"middle","author":{"id":"https://openalex.org/A5089543048","display_name":"Francesco Conti","orcid":null},"institutions":[{"id":"https://openalex.org/I9360294","display_name":"University of Bologna","ror":"https://ror.org/01111rn36","country_code":"IT","type":"education","lineage":["https://openalex.org/I9360294"]}],"countries":["IT"],"is_corresponding":false,"raw_author_name":"Francesco Conti","raw_affiliation_strings":["DEI, and Information Engineering, University of Bologna,Italy"],"affiliations":[{"raw_affiliation_string":"DEI, and Information Engineering, University of Bologna,Italy","institution_ids":["https://openalex.org/I9360294"]}]},{"author_position":"last","author":{"id":"https://openalex.org/A5043408422","display_name":"Luca Benini","orcid":"https://orcid.org/0000-0001-8068-3806"},"institutions":[{"id":"https://openalex.org/I35440088","display_name":"ETH Zurich","ror":"https://ror.org/05a28rw58","country_code":"CH","type":"education","lineage":["https://openalex.org/I2799323385","https://openalex.org/I35440088"]}],"countries":["CH"],"is_corresponding":false,"raw_author_name":"Luca Benini","raw_affiliation_strings":["ETH Zurich,Integrated Systems Laboratory,Switzerland"],"affiliations":[{"raw_affiliation_string":"ETH Zurich,Integrated Systems Laboratory,Switzerland","institution_ids":["https://openalex.org/I35440088"]}]}],"institutions":[],"countries_distinct_count":2,"institutions_distinct_count":5,"corresponding_author_ids":["https://openalex.org/A5115024178"],"corresponding_institution_ids":["https://openalex.org/I35440088"],"apc_list":null,"apc_paid":null,"fwci":1.5567,"has_fulltext":false,"cited_by_count":1,"citation_normalized_percentile":{"value":0.83865105,"is_in_top_1_percent":false,"is_in_top_10_percent":false},"cited_by_percentile_year":{"min":91,"max":95},"biblio":{"volume":null,"issue":null,"first_page":"1","last_page":"7"},"is_retracted":false,"is_paratext":false,"is_xpac":false,"primary_topic":{"id":"https://openalex.org/T10829","display_name":"Interconnection Networks and Systems","score":0.9995999932289124,"subfield":{"id":"https://openalex.org/subfields/1705","display_name":"Computer Networks and Communications"},"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/T10829","display_name":"Interconnection Networks and Systems","score":0.9995999932289124,"subfield":{"id":"https://openalex.org/subfields/1705","display_name":"Computer Networks and Communications"},"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":0.9993000030517578,"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"}},{"id":"https://openalex.org/T10363","display_name":"Low-power high-performance VLSI design","score":0.9983000159263611,"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":0.584308922290802},{"id":"https://openalex.org/keywords/inference","display_name":"Inference","score":0.5735848546028137},{"id":"https://openalex.org/keywords/chip","display_name":"Chip","score":0.5587410926818848},{"id":"https://openalex.org/keywords/transformer","display_name":"Transformer","score":0.5075143575668335},{"id":"https://openalex.org/keywords/system-on-a-chip","display_name":"System on a chip","score":0.48765623569488525},{"id":"https://openalex.org/keywords/embedded-system","display_name":"Embedded system","score":0.4763879179954529},{"id":"https://openalex.org/keywords/microcontroller","display_name":"Microcontroller","score":0.44025516510009766},{"id":"https://openalex.org/keywords/automotive-engineering","display_name":"Automotive engineering","score":0.3554706871509552},{"id":"https://openalex.org/keywords/electrical-engineering","display_name":"Electrical engineering","score":0.2985433042049408},{"id":"https://openalex.org/keywords/engineering","display_name":"Engineering","score":0.2284248173236847},{"id":"https://openalex.org/keywords/artificial-intelligence","display_name":"Artificial intelligence","score":0.16522079706192017},{"id":"https://openalex.org/keywords/telecommunications","display_name":"Telecommunications","score":0.10107338428497314},{"id":"https://openalex.org/keywords/voltage","display_name":"Voltage","score":0.08611735701560974}],"concepts":[{"id":"https://openalex.org/C41008148","wikidata":"https://www.wikidata.org/wiki/Q21198","display_name":"Computer science","level":0,"score":0.584308922290802},{"id":"https://openalex.org/C2776214188","wikidata":"https://www.wikidata.org/wiki/Q408386","display_name":"Inference","level":2,"score":0.5735848546028137},{"id":"https://openalex.org/C165005293","wikidata":"https://www.wikidata.org/wiki/Q1074500","display_name":"Chip","level":2,"score":0.5587410926818848},{"id":"https://openalex.org/C66322947","wikidata":"https://www.wikidata.org/wiki/Q11658","display_name":"Transformer","level":3,"score":0.5075143575668335},{"id":"https://openalex.org/C118021083","wikidata":"https://www.wikidata.org/wiki/Q610398","display_name":"System on a chip","level":2,"score":0.48765623569488525},{"id":"https://openalex.org/C149635348","wikidata":"https://www.wikidata.org/wiki/Q193040","display_name":"Embedded system","level":1,"score":0.4763879179954529},{"id":"https://openalex.org/C173018170","wikidata":"https://www.wikidata.org/wiki/Q165678","display_name":"Microcontroller","level":2,"score":0.44025516510009766},{"id":"https://openalex.org/C171146098","wikidata":"https://www.wikidata.org/wiki/Q124192","display_name":"Automotive engineering","level":1,"score":0.3554706871509552},{"id":"https://openalex.org/C119599485","wikidata":"https://www.wikidata.org/wiki/Q43035","display_name":"Electrical engineering","level":1,"score":0.2985433042049408},{"id":"https://openalex.org/C127413603","wikidata":"https://www.wikidata.org/wiki/Q11023","display_name":"Engineering","level":0,"score":0.2284248173236847},{"id":"https://openalex.org/C154945302","wikidata":"https://www.wikidata.org/wiki/Q11660","display_name":"Artificial intelligence","level":1,"score":0.16522079706192017},{"id":"https://openalex.org/C76155785","wikidata":"https://www.wikidata.org/wiki/Q418","display_name":"Telecommunications","level":1,"score":0.10107338428497314},{"id":"https://openalex.org/C165801399","wikidata":"https://www.wikidata.org/wiki/Q25428","display_name":"Voltage","level":2,"score":0.08611735701560974}],"mesh":[],"locations_count":2,"locations":[{"id":"doi:10.23919/date64628.2025.10992712","is_oa":false,"landing_page_url":"https://doi.org/10.23919/date64628.2025.10992712","pdf_url":null,"source":null,"license":null,"license_id":null,"version":"publishedVersion","is_accepted":true,"is_published":true,"raw_source_name":"2025 Design, Automation &amp;amp; Test in Europe Conference (DATE)","raw_type":"proceedings-article"},{"id":"pmh:oai:cris.unibo.it:11585/1040756","is_oa":false,"landing_page_url":"https://hdl.handle.net/11585/1040756","pdf_url":null,"source":{"id":"https://openalex.org/S4306402579","display_name":"Archivio istituzionale della ricerca (Alma Mater Studiorum Universit\u00e0 di Bologna)","issn_l":null,"issn":null,"is_oa":false,"is_in_doaj":false,"is_core":false,"host_organization":"https://openalex.org/I4210117483","host_organization_name":"Istituto di Ematologia di Bologna","host_organization_lineage":["https://openalex.org/I4210117483"],"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":"info:eu-repo/semantics/conferenceObject"}],"best_oa_location":null,"sustainable_development_goals":[{"display_name":"Affordable and clean energy","score":0.5,"id":"https://metadata.un.org/sdg/7"}],"awards":[],"funders":[],"has_content":{"pdf":false,"grobid_xml":false},"content_urls":null,"referenced_works_count":31,"referenced_works":["https://openalex.org/W2889339831","https://openalex.org/W2896180420","https://openalex.org/W2899663614","https://openalex.org/W3034457371","https://openalex.org/W3197527602","https://openalex.org/W3209945042","https://openalex.org/W3211490618","https://openalex.org/W4200106037","https://openalex.org/W4224035735","https://openalex.org/W4226165489","https://openalex.org/W4281708879","https://openalex.org/W4308760184","https://openalex.org/W4313496563","https://openalex.org/W4321636575","https://openalex.org/W4322718191","https://openalex.org/W4376652984","https://openalex.org/W4381586770","https://openalex.org/W4384918448","https://openalex.org/W4385245566","https://openalex.org/W4386655575","https://openalex.org/W4390632065","https://openalex.org/W4391462022","https://openalex.org/W4398757502","https://openalex.org/W4401752634","https://openalex.org/W4404101614","https://openalex.org/W4406014253","https://openalex.org/W6755207826","https://openalex.org/W6846659131","https://openalex.org/W6849146884","https://openalex.org/W6856574389","https://openalex.org/W6861768399"],"related_works":["https://openalex.org/W4316095964","https://openalex.org/W2383001583","https://openalex.org/W4206178588","https://openalex.org/W3094491777","https://openalex.org/W3214715529","https://openalex.org/W4287635093","https://openalex.org/W2065289416","https://openalex.org/W2017236304","https://openalex.org/W2115579119","https://openalex.org/W2136854845"],"abstract_inverted_index":{"Contextual":[0],"Artificial":[1],"Intelligence":[2],"(AI)":[3],"based":[4],"on":[5,75,111],"emerging":[6],"Transformer":[7],"models":[8],"is":[9,53,163],"predicted":[10],"to":[11,71,153,177],"drive":[12],"the":[13,54,62,103,107,159],"next":[14],"technology":[15],"revolution":[16],"in":[17,48,61],"interactive":[18],"wearable":[19,77],"devices":[20,37,78,93],"such":[21],"as":[22],"new-generation":[23],"smart":[24],"glasses.":[25],"By":[26],"coupling":[27],"numerous":[28],"sensors":[29],"with":[30,79,97,166],"small,":[31],"low-power":[32,76],"Micro-Controller":[33],"Units":[34],"(MCUs),":[35],"these":[36],"will":[38],"enable":[39],"on-device":[40],"intelligence":[41],"and":[42,94,142],"sensor":[43],"control.":[44],"A":[45],"major":[46],"bottleneck":[47],"this":[49,65],"class":[50],"of":[51,57,87,114,126,131,139,149],"systems":[52],"small":[55],"amount":[56],"on-chip":[58,99],"memory":[59],"available":[60],"MCUs.":[63,118],"In":[64],"paper,":[66],"we":[67],"propose":[68],"a":[69,84,112,129,136,154,167,178],"methodology":[70],"deploy":[72],"real-world":[73],"Transformers":[74],"minimal":[80],"off-chip":[81],"traffic":[82],"exploiting":[83],"distributed":[85,120,160],"system":[86,113,121],"MCUs,":[88],"partitioning":[89],"inference":[90],"across":[91],"multiple":[92],"enabling":[95],"execution":[96],"stationary":[98],"weights.":[100],"We":[101],"validate":[102],"scheme":[104],"by":[105],"deploying":[106],"TinyLlama-42M":[108],"decoder-only":[109],"model":[110],"8":[115],"parallel":[116],"ultra-low-power":[117],"The":[119],"achieves":[122],"an":[123,143],"energy":[124],"consumption":[125],"0.64":[127],"mJ,":[128],"latency":[130],"0.54":[132],"ms":[133],"per":[134],"inference,":[135],"super-linear":[137,168],"speedup":[138,171],"26.1":[140],"x,":[141,151],"Energy":[144],"Delay":[145],"Product":[146],"(EDP)":[147],"improvement":[148],"27.2":[150],"compared":[152,176],"single-chip":[155,179],"system.":[156,180],"On":[157],"MobileBERT,":[158],"system's":[161],"runtime":[162],"38.8":[164],"ms,":[165],"4.7":[169],"\u00d7":[170],"when":[172],"using":[173],"4":[174],"MCUs":[175]},"counts_by_year":[{"year":2025,"cited_by_count":1}],"updated_date":"2025-11-06T03:46:38.306776","created_date":"2025-10-10T00:00:00"}
