Please use this identifier to cite or link to this item: https://hdl.handle.net/20.500.11851/8602
Full metadata record
DC FieldValueLanguage
dc.contributor.authorKoc, Fahrettin-
dc.contributor.authorSalami, Behzad-
dc.contributor.authorErgin, Oğuz-
dc.contributor.authorUnsal, Osman-
dc.contributor.authorKestelman, Adrian Cristal-
dc.date.accessioned2022-07-30T16:41:53Z-
dc.date.available2022-07-30T16:41:53Z-
dc.date.issued2022-
dc.identifier.citationKoc, F., Salami, B., Ergin, O., Unsal, O., & Kestelman, A. C. (2022). Can We Trust Undervolting in FPGA-Based Deep Learning Designs at Harsh Conditions?. IEEE Micro, 42(3), 57-65.en_US
dc.identifier.issn0272-1732-
dc.identifier.issn1937-4143-
dc.identifier.urihttps://doi.org/10.1109/MM.2022.3153891-
dc.identifier.urihttps://hdl.handle.net/20.500.11851/8602-
dc.description.abstractAs more neural networks on field-programmable gate arrays (FPGAs) are used in a wider context, the importance of power efficiency increases. However, the focus on power should never compromise application accuracy. One technique to increase power efficiency is reducing the FPGAs' supply voltage (undervolting), which can cause accuracy problems. Therefore, careful design-time considerations are required for correct configuration without hindering the target accuracy. This fact becomes especially important for autonomous systems, edge computing, or data centers. This study reveals the impact of undervolting in harsh environmental conditions on the accuracy and power efficiency of convolutional neural network benchmarks. We perform comprehensive testing in a calibrated infrastructure at controlled temperatures (between -40 degrees C and 50 degrees C) and four distinct humidity levels (50%, 60%, 70%, and 80%) for off-the-shelf FPGAs. We show that the voltage guard-band shift with temperature is linear and propose new reliable undervolting designs providing a 65% increase in power-efficiency Giga-OPs per second (GOPS/W).en_US
dc.language.isoenen_US
dc.publisherIEEE Computer Socen_US
dc.relation.ispartofIEEE Microen_US
dc.rightsinfo:eu-repo/semantics/closedAccessen_US
dc.titleCan We Trust Undervolting in Fpga-Based Deep Learning Designs at Harsh Conditions?en_US
dc.typeArticleen_US
dc.departmentFakülteler, Mühendislik Fakültesi, Bilgisayar Mühendisliği Bölümüen_US
dc.departmentFaculties, Faculty of Engineering, Department of Computer Engineeringen_US
dc.identifier.volume42en_US
dc.identifier.issue3en_US
dc.identifier.startpage57en_US
dc.identifier.endpage65en_US
dc.identifier.wosWOS:000798189700008en_US
dc.identifier.scopus2-s2.0-85125346365en_US
dc.institutionauthorErgin, Oğuz-
dc.identifier.doi10.1109/MM.2022.3153891-
dc.authorscopusid54684331500-
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)

112
checked on Dec 23, 2024

Google ScholarTM

Check




Altmetric


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