Conference proceeding
Characterizing Deep Learning Neural Network Failures between Algorithmic Inaccuracy and Transient Hardware Faults
2022 IEEE 27TH PACIFIC RIM INTERNATIONAL SYMPOSIUM ON DEPENDABLE COMPUTING (PRDC), pp.54-67
IEEE Pacific Rim International Symposium on Dependable Computing
01/01/2022
DOI: 10.1109/PRDC55274.2022.00020
Abstract
Deep Neural Networks (DNNs) have been widely deployed in safety-critical applications such as autonomous vehicles, healthcare, and space applications. Though DNN models have long suffered intrinsic algorithmic inaccuracies, the increasing number of hardware transient faults in computer systems has been raising safety and reliability concerns in safety-critical applications. This paper investigates the impact of DNN misclassifications that caused by hardware transient faults and intrinsic algorithmic inaccuracy in safety-critical applications. We first extend a state-of-the-art fault injector for TensorFlow application, TENSORFI, to support fault injections on modern DNN models in a scalable way, then characterize the outcome classes of the models, analyzing them based on safety related metrics. Finally, we conduct a large-scale fault injection experiment to measure the failures according to the metrics and study their impact on safety. We observe that failures caused by hardware transient faults could have much more significant impact (up to 4 times higher probability) on safety-critical applications than that of the DNN algorithmic inaccuracies, advocating the potential needs to protect DNNs from hardware faults in safety-critical applications.
Details
- Title: Subtitle
- Characterizing Deep Learning Neural Network Failures between Algorithmic Inaccuracy and Transient Hardware Faults
- Creators
- Sabuj Laskar - University of IowaMd Hasanur Rahman - Univ Iowa, Iowa City, IA 52242 USABohan Zhang - University of IowaGuanpeng Li - University of Iowa
- Resource Type
- Conference proceeding
- Publication Details
- 2022 IEEE 27TH PACIFIC RIM INTERNATIONAL SYMPOSIUM ON DEPENDABLE COMPUTING (PRDC), pp.54-67
- Publisher
- IEEE
- Series
- IEEE Pacific Rim International Symposium on Dependable Computing
- DOI
- 10.1109/PRDC55274.2022.00020
- ISSN
- 1555-094X
- eISSN
- 2473-3105
- Number of pages
- 14
- Language
- English
- Date published
- 01/01/2022
- Academic Unit
- Computer Science
- Record Identifier
- 9984411083502771
Metrics
1 Record Views