Special Session: Fault Criticality Assessment in AI Accelerators

The ubiquitous application of deep neural networks (DNN) has led to a rise in demand for AI accelerators. DNN-specific functional criticality analysis identifies faults that cause measurable and significant deviations from acceptable requirements such as the inferencing accuracy. This paper examines...

Celý popis

Uložené v:
Podrobná bibliografia
Vydané v:Proceedings - IEEE VLSI Test Symposium s. 1 - 4
Hlavní autori: Chaudhuri, Arjun, Talukdar, Jonti, Chakrabarty, Krishnendu
Médium: Konferenčný príspevok..
Jazyk:English
Vydavateľské údaje: IEEE 25.04.2022
Predmet:
ISSN:2375-1053
On-line prístup:Získať plný text
Tagy: Pridať tag
Žiadne tagy, Buďte prvý, kto otaguje tento záznam!
Abstract The ubiquitous application of deep neural networks (DNN) has led to a rise in demand for AI accelerators. DNN-specific functional criticality analysis identifies faults that cause measurable and significant deviations from acceptable requirements such as the inferencing accuracy. This paper examines the problem of classifying structural faults in the processing elements (PEs) of systolic-array accelerators. We first present a two-tier machine-learning (ML) based method to assess the functional criticality of faults. The problem of minimizing misclassification is addressed by utilizing generative adversarial networks (GANs). The two-tier ML/GAN-based criticality assessment method leads to less than 1% test escapes during functional criticality evaluation of structural faults. While supervised learning techniques can be used to accurately estimate fault criticality, it requires a considerable amount of ground truth for model training. We therefore describe a neural-twin framework for analyzing fault criticality with a negligible amount of ground-truth data. A recently proposed misclassification-driven training algorithm is used to sensitize and identify biases that are critical to the functioning of the accelerator for a given application workload. The proposed framework achieves up to 100% accuracy in fault-criticality classification in 16-bit and 32-bit PEs by using the criticality knowledge of only 2% of the total faults in a PE.
AbstractList The ubiquitous application of deep neural networks (DNN) has led to a rise in demand for AI accelerators. DNN-specific functional criticality analysis identifies faults that cause measurable and significant deviations from acceptable requirements such as the inferencing accuracy. This paper examines the problem of classifying structural faults in the processing elements (PEs) of systolic-array accelerators. We first present a two-tier machine-learning (ML) based method to assess the functional criticality of faults. The problem of minimizing misclassification is addressed by utilizing generative adversarial networks (GANs). The two-tier ML/GAN-based criticality assessment method leads to less than 1% test escapes during functional criticality evaluation of structural faults. While supervised learning techniques can be used to accurately estimate fault criticality, it requires a considerable amount of ground truth for model training. We therefore describe a neural-twin framework for analyzing fault criticality with a negligible amount of ground-truth data. A recently proposed misclassification-driven training algorithm is used to sensitize and identify biases that are critical to the functioning of the accelerator for a given application workload. The proposed framework achieves up to 100% accuracy in fault-criticality classification in 16-bit and 32-bit PEs by using the criticality knowledge of only 2% of the total faults in a PE.
Author Talukdar, Jonti
Chakrabarty, Krishnendu
Chaudhuri, Arjun
Author_xml – sequence: 1
  givenname: Arjun
  surname: Chaudhuri
  fullname: Chaudhuri, Arjun
  organization: Duke University,Department of Electrical and Computer Engineering,Durham,NC
– sequence: 2
  givenname: Jonti
  surname: Talukdar
  fullname: Talukdar, Jonti
  organization: Duke University,Department of Electrical and Computer Engineering,Durham,NC
– sequence: 3
  givenname: Krishnendu
  surname: Chakrabarty
  fullname: Chakrabarty, Krishnendu
  organization: Duke University,Department of Electrical and Computer Engineering,Durham,NC
BookMark eNotj81KAzEYAKMo2NY-gQh5gV2__CeeXBarhYKHrV5LmnyByHZbNuuhb69gT3MYGJg5uRmOAxLyyKBmDNzT17ZTXAHUHDirnXGSM3VFls5YprWSDDSwazLjwqiKgRJ3ZF7KNwAHJWFGXroThux72mEp-Tg805X_6SfajnnKwfd5OtOmlD95wGGieaDNmjYhYI-jn45juSe3yfcFlxcuyOfqddu-V5uPt3XbbKrMQUyVtnth9zGG4Kz1MVntjRTe-AgyBm2Sx8BM4jIljS5FHZNMIhqnWNQ6oFiQh_9uRsTdacwHP553l2HxC63zTTE
ContentType Conference Proceeding
DBID 6IE
6IL
CBEJK
RIE
RIL
DOI 10.1109/VTS52500.2021.9794215
DatabaseName IEEE Electronic Library (IEL) Conference Proceedings
IEEE Proceedings Order Plan All Online (POP All Online) 1998-present by volume
IEEE Xplore All Conference Proceedings
IEEE Electronic Library (IEL)
IEEE Proceedings Order Plans (POP All) 1998-Present
DatabaseTitleList
Database_xml – sequence: 1
  dbid: RIE
  name: IEEE Electronic Library (IEL)
  url: https://ieeexplore.ieee.org/
  sourceTypes: Publisher
DeliveryMethod fulltext_linktorsrc
Discipline Engineering
EISBN 9781665410601
1665410604
EISSN 2375-1053
EndPage 4
ExternalDocumentID 9794215
Genre orig-research
GroupedDBID 6IE
6IL
6IN
AAWTH
ABLEC
ADZIZ
ALMA_UNASSIGNED_HOLDINGS
BEFXN
BFFAM
BGNUA
BKEBE
BPEOZ
CBEJK
CHZPO
IEGSK
M43
OCL
RIE
RIL
ID FETCH-LOGICAL-i203t-68b38bddcc988adf86a743a7ad04dc67faec17f24ff6e9fd6df4f3d7951d66ce3
IEDL.DBID RIE
ISICitedReferencesCount 5
ISICitedReferencesURI http://www.webofscience.com/api/gateway?GWVersion=2&SrcApp=Summon&SrcAuth=ProQuest&DestLinkType=CitingArticles&DestApp=WOS_CPL&KeyUT=000850232500027&url=https%3A%2F%2Fcvtisr.summon.serialssolutions.com%2F%23%21%2Fsearch%3Fho%3Df%26include.ft.matches%3Dt%26l%3Dnull%26q%3D
IngestDate Wed Aug 27 02:23:59 EDT 2025
IsPeerReviewed false
IsScholarly true
Language English
LinkModel DirectLink
MergedId FETCHMERGED-LOGICAL-i203t-68b38bddcc988adf86a743a7ad04dc67faec17f24ff6e9fd6df4f3d7951d66ce3
PageCount 4
ParticipantIDs ieee_primary_9794215
PublicationCentury 2000
PublicationDate 2022-April-25
PublicationDateYYYYMMDD 2022-04-25
PublicationDate_xml – month: 04
  year: 2022
  text: 2022-April-25
  day: 25
PublicationDecade 2020
PublicationTitle Proceedings - IEEE VLSI Test Symposium
PublicationTitleAbbrev VTS
PublicationYear 2022
Publisher IEEE
Publisher_xml – name: IEEE
SSID ssj0020540
Score 2.2153964
Snippet The ubiquitous application of deep neural networks (DNN) has led to a rise in demand for AI accelerators. DNN-specific functional criticality analysis...
SourceID ieee
SourceType Publisher
StartPage 1
SubjectTerms AI accelerators
Deep learning
Fault diagnosis
Neural networks
Supervised learning
Training
Very large scale integration
Title Special Session: Fault Criticality Assessment in AI Accelerators
URI https://ieeexplore.ieee.org/document/9794215
WOSCitedRecordID wos000850232500027&url=https%3A%2F%2Fcvtisr.summon.serialssolutions.com%2F%23%21%2Fsearch%3Fho%3Df%26include.ft.matches%3Dt%26l%3Dnull%26q%3D
hasFullText 1
inHoldings 1
isFullTextHit
isPrint
link http://cvtisr.summon.serialssolutions.com/2.0.0/link/0/eLvHCXMwlV1LSwMxEA61eNCLj1Z8k4NH02aTmIcni1j0Ugqt0ltJkwkUZFvarb_fZLuuCl68hYQQmATmm8x88yF0I2NMAspYAtIFIpzPiGVGEOm9lJ7GFetKsQk1GOjJxAwb6LbmwgBAWXwGnTQsc_l-4Tbpq6xr4uNhiVG-o5TccrXq4CpBj4qhk1HTfRuPUsKOxgiQZZ1q4y8FldKB9A_-d_Qhan8z8fCw9jFHqAH5Mdr_0USwhR4qCXk82rbYuMd9u3kv8JeIQYTZuFf338TzHPdecM-56G_KFPu6jV77T-PHZ1LpIpA5o7wgUs-4nnnvnNHa-qCljTjAKuup8E6qYMFlKjARggQTvPRBBO5VBFPxAhzwE9TMFzmcIixSWSbXYKUGAdqlrKjX1Mos44JTfYZayRbT5bb1xbQyw_nf0xdojyV2ABWE3V2iZrHawBXadR_FfL26Lu_rEyPDl3c
linkProvider IEEE
linkToHtml http://cvtisr.summon.serialssolutions.com/2.0.0/link/0/eLvHCXMwlV1LSwMxEB5KFdSLj1Z8m4NH02aTNJv1ZBFLi7UUWqW3kuYBBdlKu_X3m2zXVcGLt5AQApPAfJOZbz6AG-FjEhsnCluhHebaRFjRhGNhjBCG-BWlc7GJeDCQk0kyrMBtyYWx1ubFZ7YRhnku3yz0OnyVNRP_eGhglG-1OKdkw9Yqw6sAPgqOTkSS5ut4FFJ2xMeANGoUW39pqOQupLP_v8MPoP7NxUPD0sscQsWmR7D3o41gDe4LEXk02jTZuEMdtX7L0JeMgQfaqF124ETzFLV7qK219zh5kn1Vh5fO4_ihiwtlBDynhGVYyBmTM2O0TqRUxkmhPBJQsTKEGy1ip6yOYke5c8ImzgjjuGMm9nDKX4G27Biq6SK1J4B4KMxk0iohLbdSh7yokUSJKGKcEXkKtWCL6fum-cW0MMPZ39PXsNMdP_en_d7g6Rx2aeAKEI5p6wKq2XJtL2Fbf2Tz1fIqv7tPNaGavg
openUrl ctx_ver=Z39.88-2004&ctx_enc=info%3Aofi%2Fenc%3AUTF-8&rfr_id=info%3Asid%2Fsummon.serialssolutions.com&rft_val_fmt=info%3Aofi%2Ffmt%3Akev%3Amtx%3Abook&rft.genre=proceeding&rft.title=Proceedings+-+IEEE+VLSI+Test+Symposium&rft.atitle=Special+Session%3A+Fault+Criticality+Assessment+in+AI+Accelerators&rft.au=Chaudhuri%2C+Arjun&rft.au=Talukdar%2C+Jonti&rft.au=Chakrabarty%2C+Krishnendu&rft.date=2022-04-25&rft.pub=IEEE&rft.eissn=2375-1053&rft.spage=1&rft.epage=4&rft_id=info:doi/10.1109%2FVTS52500.2021.9794215&rft.externalDocID=9794215