A Survey on Modeling and Improving Reliability of DNN Algorithms and Accelerators

Mittal, Sparsh (2020) A Survey on Modeling and Improving Reliability of DNN Algorithms and Accelerators. Elsevier.

[img] Text
Systems_Architecture.pdf
Restricted to Registered users only

Download (3MB) | Request a copy

Abstract

As DNNs become increasingly common in mission-critical applications, ensuring their reliable operation has become crucial. Conventional resilience techniques fail to account for the unique characteristics of DNN algorithms/accelerators, and hence, they are infeasible or ineffective. In this paper, we present a survey of techniques for studying and optimizing the reliability of DNN accelerators and architectures. The reliability issues we cover include soft/hard errors arising due to process variation, voltage scaling, timing errors, DRAM errors due to refresh rate scaling and thermal effects, etc. We organize the research projects on several categories to bring out their key attributes. This paper underscores the importance of designing for reliability as the first principle, and not merely retrofit for it.

[error in script]
IITH Creators:
IITH CreatorsORCiD
Mittal, Sparshhttp://orcid.org/0000-0002-2908-993X
Item Type: Other
Additional Information: The author declares no competing interest. Sparsh Mittal received the B.Tech. degree in electronics and communications engineering from IIT, Roorkee, India and the Ph.D. degree in computer engineering from Iowa State University (ISU), USA. He worked as a Post-Doctoral Research Associate at Oak Ridge National Lab (ORNL), USA for 3 years. He is currently working as an assistant professor at IIT Hyderabad, India. He was the graduating topper of his batch in B.Tech and has received fellowship from ISU and performance award from ORNL. Sparsh has published more than 80 papers in top conferences and journals. My research has been covered by several technical news websites, e.g. Phys.org, InsideHPC, Primeur Magazine, StorageSearch, Data- Compression.info, TechEnablement, ScientificComputing, SemiEngineering, ReRAM forum and HPCWire. His research interests include accelerators for neural networks, architectures for machine learning, nonvolatile memory, and GPU architectures. His webpage is http://www.iith.ac.in/~sparsh/ .
Uncontrolled Keywords: Deep learning, Deep neural networks, Fault-injection, Permanent fault, Review, Transient fault, Indexed in Scopus
Subjects: Computer science
Divisions: Department of Computer Science & Engineering
Depositing User: Team Library
Date Deposited: 02 Dec 2019 04:37
Last Modified: 25 Oct 2022 11:42
URI: http://raiithold.iith.ac.in/id/eprint/7074
Publisher URL: http://doi.org/10.1016/j.sysarc.2019.101689
OA policy: https://v2.sherpa.ac.uk/id/publication/11438
Related URLs:

Actions (login required)

View Item View Item
Statistics for RAIITH ePrint 7074 Statistics for this ePrint Item