{"id":"https://openalex.org/W7117242934","doi":"https://doi.org/10.1109/mm.2025.3646859","title":"Phoebe: Measuring the Unmeasurable\u2014Demystifying Silent Data Corruptions in AI Accelerators Through Microarchitectural Modeling","display_name":"Phoebe: Measuring the Unmeasurable\u2014Demystifying Silent Data Corruptions in AI Accelerators Through Microarchitectural Modeling","publication_year":2025,"publication_date":"2025-12-25","ids":{"openalex":"https://openalex.org/W7117242934","doi":"https://doi.org/10.1109/mm.2025.3646859"},"language":null,"primary_location":{"id":"doi:10.1109/mm.2025.3646859","is_oa":false,"landing_page_url":"https://doi.org/10.1109/mm.2025.3646859","pdf_url":null,"source":{"id":"https://openalex.org/S59697426","display_name":"IEEE Micro","issn_l":"0272-1732","issn":["0272-1732","1937-4143"],"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 Micro","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/A5058376544","display_name":"Odysseas Chatzopoulos","orcid":null},"institutions":[{"id":"https://openalex.org/I200777214","display_name":"National and Kapodistrian University of Athens","ror":"https://ror.org/04gnjpq42","country_code":"GR","type":"education","lineage":["https://openalex.org/I200777214"]}],"countries":["GR"],"is_corresponding":false,"raw_author_name":"Odysseas Chatzopoulos","raw_affiliation_strings":["University of Athens, Athens, Greece"],"raw_orcid":"https://orcid.org/0009-0000-9801-4483","affiliations":[{"raw_affiliation_string":"University of Athens, Athens, Greece","institution_ids":["https://openalex.org/I200777214"]}]},{"author_position":"middle","author":{"id":"https://openalex.org/A5104601576","display_name":"Maria Trakosa","orcid":null},"institutions":[{"id":"https://openalex.org/I200777214","display_name":"National and Kapodistrian University of Athens","ror":"https://ror.org/04gnjpq42","country_code":"GR","type":"education","lineage":["https://openalex.org/I200777214"]}],"countries":["GR"],"is_corresponding":false,"raw_author_name":"Maria Trakosa","raw_affiliation_strings":["University of Athens, Athens, Greece"],"raw_orcid":"https://orcid.org/0009-0000-0177-142X","affiliations":[{"raw_affiliation_string":"University of Athens, Athens, Greece","institution_ids":["https://openalex.org/I200777214"]}]},{"author_position":"middle","author":{"id":"https://openalex.org/A5018336899","display_name":"Harish Dattatraya Dixit","orcid":"https://orcid.org/0009-0001-1163-5568"},"institutions":[{"id":"https://openalex.org/I4210114444","display_name":"Meta (United States)","ror":"https://ror.org/01zbnvs85","country_code":"US","type":"company","lineage":["https://openalex.org/I4210114444"]}],"countries":["US"],"is_corresponding":false,"raw_author_name":"Harish D. Dixit","raw_affiliation_strings":["Meta Platforms Inc., Menlo Park, CA, USA"],"raw_orcid":"https://orcid.org/0009-0001-1163-5568","affiliations":[{"raw_affiliation_string":"Meta Platforms Inc., Menlo Park, CA, USA","institution_ids":["https://openalex.org/I4210114444"]}]},{"author_position":"middle","author":{"id":"https://openalex.org/A5103061483","display_name":"Sriram Sankar","orcid":"https://orcid.org/0009-0008-4581-8371"},"institutions":[{"id":"https://openalex.org/I4210114444","display_name":"Meta (United States)","ror":"https://ror.org/01zbnvs85","country_code":"US","type":"company","lineage":["https://openalex.org/I4210114444"]}],"countries":["US"],"is_corresponding":false,"raw_author_name":"Sriram Sankar","raw_affiliation_strings":["Meta Platforms Inc., Menlo Park, CA, USA"],"raw_orcid":"https://orcid.org/0009-0008-4581-8371","affiliations":[{"raw_affiliation_string":"Meta Platforms Inc., Menlo Park, CA, USA","institution_ids":["https://openalex.org/I4210114444"]}]},{"author_position":"last","author":{"id":"https://openalex.org/A5007119083","display_name":"Dimitris Gizopoulos","orcid":"https://orcid.org/0000-0002-1613-9061"},"institutions":[{"id":"https://openalex.org/I200777214","display_name":"National and Kapodistrian University of Athens","ror":"https://ror.org/04gnjpq42","country_code":"GR","type":"education","lineage":["https://openalex.org/I200777214"]}],"countries":["GR"],"is_corresponding":false,"raw_author_name":"Dimitris Gizopoulos","raw_affiliation_strings":["University of Athens, Athens, Greece"],"raw_orcid":"https://orcid.org/0000-0002-1613-9061","affiliations":[{"raw_affiliation_string":"University of Athens, Athens, Greece","institution_ids":["https://openalex.org/I200777214"]}]}],"institutions":[],"countries_distinct_count":2,"institutions_distinct_count":2,"corresponding_author_ids":[],"corresponding_institution_ids":[],"apc_list":null,"apc_paid":null,"fwci":0.9253,"has_fulltext":false,"cited_by_count":2,"citation_normalized_percentile":{"value":0.79143009,"is_in_top_1_percent":false,"is_in_top_10_percent":false},"cited_by_percentile_year":{"min":97,"max":98},"biblio":{"volume":"46","issue":"1","first_page":"55","last_page":"65"},"is_retracted":false,"is_paratext":false,"is_xpac":false,"primary_topic":{"id":"https://openalex.org/T11005","display_name":"Radiation Effects in Electronics","score":0.9685999751091003,"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"}},"topics":[{"id":"https://openalex.org/T11005","display_name":"Radiation Effects in Electronics","score":0.9685999751091003,"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/T10054","display_name":"Parallel Computing and Optimization Techniques","score":0.004900000058114529,"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/T11424","display_name":"Security and Verification in Computing","score":0.004900000058114529,"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/reliability","display_name":"Reliability (semiconductor)","score":0.6601999998092651},{"id":"https://openalex.org/keywords/speedup","display_name":"Speedup","score":0.6499000191688538},{"id":"https://openalex.org/keywords/throughput","display_name":"Throughput","score":0.583299994468689},{"id":"https://openalex.org/keywords/design-space-exploration","display_name":"Design space exploration","score":0.5307000279426575},{"id":"https://openalex.org/keywords/microarchitecture","display_name":"Microarchitecture","score":0.5209000110626221},{"id":"https://openalex.org/keywords/key","display_name":"Key (lock)","score":0.5027999877929688},{"id":"https://openalex.org/keywords/software","display_name":"Software","score":0.476500004529953},{"id":"https://openalex.org/keywords/range","display_name":"Range (aeronautics)","score":0.4343000054359436},{"id":"https://openalex.org/keywords/vulnerability","display_name":"Vulnerability (computing)","score":0.43070000410079956}],"concepts":[{"id":"https://openalex.org/C41008148","wikidata":"https://www.wikidata.org/wiki/Q21198","display_name":"Computer science","level":0,"score":0.8604999780654907},{"id":"https://openalex.org/C43214815","wikidata":"https://www.wikidata.org/wiki/Q7310987","display_name":"Reliability (semiconductor)","level":3,"score":0.6601999998092651},{"id":"https://openalex.org/C68339613","wikidata":"https://www.wikidata.org/wiki/Q1549489","display_name":"Speedup","level":2,"score":0.6499000191688538},{"id":"https://openalex.org/C157764524","wikidata":"https://www.wikidata.org/wiki/Q1383412","display_name":"Throughput","level":3,"score":0.583299994468689},{"id":"https://openalex.org/C2776221188","wikidata":"https://www.wikidata.org/wiki/Q21072556","display_name":"Design space exploration","level":2,"score":0.5307000279426575},{"id":"https://openalex.org/C107598950","wikidata":"https://www.wikidata.org/wiki/Q259864","display_name":"Microarchitecture","level":2,"score":0.5209000110626221},{"id":"https://openalex.org/C149635348","wikidata":"https://www.wikidata.org/wiki/Q193040","display_name":"Embedded system","level":1,"score":0.5108000040054321},{"id":"https://openalex.org/C26517878","wikidata":"https://www.wikidata.org/wiki/Q228039","display_name":"Key (lock)","level":2,"score":0.5027999877929688},{"id":"https://openalex.org/C2777904410","wikidata":"https://www.wikidata.org/wiki/Q7397","display_name":"Software","level":2,"score":0.476500004529953},{"id":"https://openalex.org/C118524514","wikidata":"https://www.wikidata.org/wiki/Q173212","display_name":"Computer architecture","level":1,"score":0.4544999897480011},{"id":"https://openalex.org/C204323151","wikidata":"https://www.wikidata.org/wiki/Q905424","display_name":"Range (aeronautics)","level":2,"score":0.4343000054359436},{"id":"https://openalex.org/C95713431","wikidata":"https://www.wikidata.org/wiki/Q631425","display_name":"Vulnerability (computing)","level":2,"score":0.43070000410079956},{"id":"https://openalex.org/C200601418","wikidata":"https://www.wikidata.org/wiki/Q2193887","display_name":"Reliability engineering","level":1,"score":0.4172999858856201},{"id":"https://openalex.org/C2775928411","wikidata":"https://www.wikidata.org/wiki/Q2041312","display_name":"Fault injection","level":3,"score":0.35269999504089355},{"id":"https://openalex.org/C2778309119","wikidata":"https://www.wikidata.org/wiki/Q5121614","display_name":"Circuit reliability","level":4,"score":0.3174000084400177},{"id":"https://openalex.org/C163258240","wikidata":"https://www.wikidata.org/wiki/Q25342","display_name":"Power (physics)","level":2,"score":0.30630001425743103},{"id":"https://openalex.org/C113775141","wikidata":"https://www.wikidata.org/wiki/Q428691","display_name":"Computer engineering","level":1,"score":0.30309998989105225},{"id":"https://openalex.org/C2779662365","wikidata":"https://www.wikidata.org/wiki/Q5416694","display_name":"Event (particle physics)","level":2,"score":0.30000001192092896},{"id":"https://openalex.org/C63540848","wikidata":"https://www.wikidata.org/wiki/Q3140932","display_name":"Fault tolerance","level":2,"score":0.29829999804496765},{"id":"https://openalex.org/C93682380","wikidata":"https://www.wikidata.org/wiki/Q2025226","display_name":"Static timing analysis","level":2,"score":0.2937999963760376},{"id":"https://openalex.org/C173608175","wikidata":"https://www.wikidata.org/wiki/Q232661","display_name":"Parallel computing","level":1,"score":0.28679999709129333},{"id":"https://openalex.org/C175551986","wikidata":"https://www.wikidata.org/wiki/Q47089","display_name":"Fault (geology)","level":2,"score":0.2838999927043915},{"id":"https://openalex.org/C168167062","wikidata":"https://www.wikidata.org/wiki/Q1117970","display_name":"Component (thermodynamics)","level":2,"score":0.2831000089645386},{"id":"https://openalex.org/C13164978","wikidata":"https://www.wikidata.org/wiki/Q600158","display_name":"Hardware acceleration","level":3,"score":0.26260000467300415},{"id":"https://openalex.org/C168522837","wikidata":"https://www.wikidata.org/wiki/Q679552","display_name":"Branch predictor","level":2,"score":0.2615000009536743},{"id":"https://openalex.org/C138852830","wikidata":"https://www.wikidata.org/wiki/Q2292993","display_name":"Design methods","level":2,"score":0.257999986410141},{"id":"https://openalex.org/C117447612","wikidata":"https://www.wikidata.org/wiki/Q1412670","display_name":"Software quality","level":4,"score":0.25290000438690186}],"mesh":[],"locations_count":1,"locations":[{"id":"doi:10.1109/mm.2025.3646859","is_oa":false,"landing_page_url":"https://doi.org/10.1109/mm.2025.3646859","pdf_url":null,"source":{"id":"https://openalex.org/S59697426","display_name":"IEEE Micro","issn_l":"0272-1732","issn":["0272-1732","1937-4143"],"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 Micro","raw_type":"journal-article"}],"best_oa_location":null,"sustainable_development_goals":[{"display_name":"Peace, Justice and strong institutions","score":0.8170267343521118,"id":"https://metadata.un.org/sdg/16"}],"awards":[{"id":"https://openalex.org/G2739299204","display_name":"Digital Autonomy for RISC-V in Europe (Specific Grant Agreement 1)","funder_award_id":"101202459","funder_id":"https://openalex.org/F4320320300","funder_display_name":"European Commission"},{"id":"https://openalex.org/G4846513126","display_name":null,"funder_award_id":"101097224","funder_id":"https://openalex.org/F4320334322","funder_display_name":"HORIZON EUROPE Framework Programme"},{"id":"https://openalex.org/G5556781112","display_name":null,"funder_award_id":"101070238","funder_id":"https://openalex.org/F4320334322","funder_display_name":"HORIZON EUROPE Framework Programme"}],"funders":[{"id":"https://openalex.org/F4320320300","display_name":"European Commission","ror":"https://ror.org/00k4n6c32"},{"id":"https://openalex.org/F4320334322","display_name":"HORIZON EUROPE Framework Programme","ror":null}],"has_content":{"pdf":false,"grobid_xml":false},"content_urls":null,"referenced_works_count":12,"referenced_works":["https://openalex.org/W2100597484","https://openalex.org/W2752470767","https://openalex.org/W2969553121","https://openalex.org/W3105084904","https://openalex.org/W4249144718","https://openalex.org/W4320002940","https://openalex.org/W4376126263","https://openalex.org/W4381746679","https://openalex.org/W4393592433","https://openalex.org/W4409248377","https://openalex.org/W4410638920","https://openalex.org/W4413320771"],"related_works":[],"abstract_inverted_index":{"AI":[0,41,70],"accelerators":[1,42,71],"are":[2],"vital":[3],"for":[4,61,156],"the":[5,62,65,73,81,129,151],"efficient":[6],"execution":[7],"of":[8,64,67,75,91,119,153],"data-heavy":[9],"workloads":[10],"across":[11,147],"various":[12],"environments,":[13],"but":[14],"increased":[15],"design":[16,92,158],"complexity":[17],"and":[18,50,99,122],"shrinking":[19],"manufacturing":[20],"processes":[21],"lead":[22],"to":[23,47,86],"severe":[24],"reliability":[25,36,66],"issues":[26],"(with":[27],"insidious":[28],"behavior":[29],"like":[30],"silent":[31],"data":[32],"corruptions\u2014SDCs).":[33],"While":[34],"CPU":[35],"has":[37],"been":[38],"analyzed":[39],"significantly,":[40],"require":[43],"tailored":[44],"analysis":[45],"due":[46],"their":[48],"diverse":[49,140],"evolving":[51],"designs.":[52],"In":[53],"this":[54],"work,":[55],"we":[56],"present":[57],"a":[58,88,116],"comprehensive":[59],"methodology":[60,155],"evaluation":[63],"systolic":[68],"array-based":[69],"harnessing":[72],"power":[74],"microarchitectural":[76],"modeling.":[77],"Our":[78],"approach":[79],"captures":[80],"entire":[82],"compute":[83],"stack\u2014from":[84],"software":[85],"gates\u2014across":[87],"wide":[89,117],"range":[90,118],"points":[93],"including":[94],"dataflows,":[95],"datatypes,":[96],"memory":[97],"hierarchies,":[98],"array":[100],"sizes.":[101],"We":[102],"support":[103],"statistical":[104],"fault":[105,120],"injection":[106,130],"on":[107,139],"key":[108],"hardware":[109],"structures":[110],"(SRAMs,":[111],"registers,":[112],"functional":[113],"units),":[114],"supporting":[115],"models,":[121],"implement":[123],"speedup":[124],"techniques":[125],"that":[126],"significantly":[127],"boost":[128],"throughput":[131],"(more":[132],"than":[133],"1000\u00d7)":[134],"without":[135],"sacrificing":[136],"accuracy.":[137],"Evaluations":[138],"ML":[141],"models":[142],"reveal":[143],"SDC":[144],"vulnerability":[145],"trends":[146],"accelerator":[148],"design-points":[149],"highlighting":[150],"effectiveness":[152],"our":[154],"broad":[157],"space":[159],"exploration.":[160]},"counts_by_year":[{"year":2026,"cited_by_count":2}],"updated_date":"2026-03-27T05:58:40.876381","created_date":"2025-12-25T00:00:00"}