Please use this identifier to cite or link to this item:
Title: Can We Trust Undervolting in FPGA-Based Deep Learning Designs at Harsh Conditions?
Authors: Koc, Fahrettin
Salami, Behzad
Ergin, Oğuz
Unsal, Osman
Kestelman, Adrian Cristal
Issue Date: 2022
Publisher: IEEE Computer Soc
Source: Koc, 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.
Abstract: As 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).
ISSN: 0272-1732
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 full item record

CORE Recommender

Page view(s)

checked on Dec 26, 2022

Google ScholarTM



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