Please use this identifier to cite or link to this item: https://hdl.handle.net/20.500.11851/8375
Full metadata record
DC FieldValueLanguage
dc.contributor.authorYüksel, I.E.-
dc.contributor.authorSalami, B.-
dc.contributor.authorErgin, Oğuz-
dc.contributor.authorÜnsal, O.S.-
dc.contributor.authorKestelman, A.C.-
dc.date.accessioned2022-01-15T13:02:42Z-
dc.date.available2022-01-15T13:02:42Z-
dc.date.issued2022-
dc.identifier.issn0278-0070-
dc.identifier.urihttps://doi.org/10.1109/TCAD.2021.3120073-
dc.identifier.urihttps://hdl.handle.net/20.500.11851/8375-
dc.description.abstractOn-chip memory (usually based on Static RAMs-SRAMs) are crucial components for various computing devices including heterogeneous devices, e.g, GPUs, FPGAs, ASICs to achieve high performance. Modern workloads such as Deep Neural Networks (DNNs) running on these heterogeneous fabrics are highly dependent on the on-chip memory architecture for efficient acceleration. Hence, improving the energy-efficiency of such memories directly leads to an efficient system. One of the common methods to save energy is undervolting i.e., supply voltage underscaling below the nominal level. Such systems can be safely undervolted without incurring faults down to a certain voltage limit. This safe range is also called voltage guardband. However, reducing voltage below the guardband level without decreasing frequency causes timing-based faults. In this paper, we propose MoRS, a framework that generates the first approximate undervolting fault model using real faults extracted from experimental undervolting studies on SRAMs to build the model. We inject the faults generated by MoRS into the on-chip memory of the DNN accelerator to evaluate the resilience of the system under the test. MoRS has the advantage of simplicity without any need for high-time overhead experiments while being accurate enough in comparison to a fully randomly-generated fault injection approach. We evaluate our experiment in popular DNN workloads by mapping weights to SRAMs and measure the accuracy difference between the output of the MoRS and the real data. Our results show that the maximum difference between real fault data and the output fault model of MoRS is 6.21%, whereas the maximum difference between real data and random fault injection model is 23.2%. In terms of average proximity to the real data, the output of MoRS outperforms the random fault injection approach by 3.21x. IEEEen_US
dc.language.isoenen_US
dc.publisherInstitute of Electrical and Electronics Engineers Inc.en_US
dc.relation.ispartofIEEE Transactions on Computer-Aided Design of Integrated Circuits and Systemsen_US
dc.rightsinfo:eu-repo/semantics/openAccessen_US
dc.subjectCircuit faultsen_US
dc.subjectData modelsen_US
dc.subjectFault-injectionen_US
dc.subjectIntegrated circuit modelingen_US
dc.subjectMathematical modelsen_US
dc.subjectModelingen_US
dc.subjectNeural Networksen_US
dc.subjectPower demanden_US
dc.subjectRandom access memoryen_US
dc.subjectSRAM.en_US
dc.subjectUndervoltingen_US
dc.subjectVoltageen_US
dc.subjectDeep neural networksen_US
dc.subjectEnergy efficiencyen_US
dc.subjectIntegrated circuitsen_US
dc.subjectMemory architectureen_US
dc.subjectProgram processorsen_US
dc.subjectSoftware testingen_US
dc.subjectStatic random access storageen_US
dc.subjectCircuit faultsen_US
dc.subjectFault injectionen_US
dc.subjectFault modelen_US
dc.subjectIntegrated circuit modelingen_US
dc.subjectModelingen_US
dc.subjectNeural-networksen_US
dc.subjectPower demandsen_US
dc.subjectRandom access memoryen_US
dc.subjectSRAM.en_US
dc.subjectUndervoltingen_US
dc.subjectTiming circuitsen_US
dc.titleMors: an Approximate Fault Modelling Framework for Reduced-Voltage Sramsen_US
dc.typeArticleen_US
dc.departmentFaculties, Faculty of Engineering, Department of Computer Engineeringen_US
dc.departmentFakülteler, Mühendislik Fakültesi, Bilgisayar Mühendisliği Bölümütr_TR
dc.identifier.wosWOS:000799624800009en_US
dc.identifier.scopus2-s2.0-85117839153en_US
dc.institutionauthorErgin, Oğuz-
dc.identifier.doi10.1109/TCAD.2021.3120073-
dc.authorscopusid57218846490-
dc.authorscopusid56029413900-
dc.authorscopusid6603141208-
dc.authorscopusid35612224700-
dc.authorscopusid56167359000-
dc.relation.publicationcategoryMakale - Uluslararası Hakemli Dergi - Kurum Öğretim Elemanıen_US
dc.identifier.scopusqualityQ1-
item.openairetypeArticle-
item.languageiso639-1en-
item.grantfulltextnone-
item.fulltextNo Fulltext-
item.openairecristypehttp://purl.org/coar/resource_type/c_18cf-
item.cerifentitytypePublications-
crisitem.author.dept02.3. Department of Computer Engineering-
Appears in Collections:Bilgisayar Mühendisliği Bölümü / Department of Computer Engineering
Scopus İndeksli Yayınlar Koleksiyonu / Scopus Indexed Publications Collection
WoS İndeksli Yayınlar Koleksiyonu / WoS Indexed Publications Collection
Show simple item record



CORE Recommender

WEB OF SCIENCETM
Citations

1
checked on Dec 21, 2024

Page view(s)

252
checked on Dec 23, 2024

Google ScholarTM

Check




Altmetric


Items in GCRIS Repository are protected by copyright, with all rights reserved, unless otherwise indicated.