When Single Event Upset Meets Deep Neural Networks: Observations, Explorations, and Remedies
This addresses hardware-level robustness for neural networks in security-sensitive systems, representing an incremental advance by focusing on a specific hardware vulnerability.
The paper tackles the problem of Single Event Upset (SEU) induced parameter perturbations in deep neural networks, which can degrade accuracy by up to 28%, and proposes remedies that reduce this degradation to 0.27% with minimal hardware overhead.
Deep Neural Network has proved its potential in various perception tasks and hence become an appealing option for interpretation and data processing in security sensitive systems. However, security-sensitive systems demand not only high perception performance, but also design robustness under various circumstances. Unlike prior works that study network robustness from software level, we investigate from hardware perspective about the impact of Single Event Upset (SEU) induced parameter perturbation (SIPP) on neural networks. We systematically define the fault models of SEU and then provide the definition of sensitivity to SIPP as the robustness measure for the network. We are then able to analytically explore the weakness of a network and summarize the key findings for the impact of SIPP on different types of bits in a floating point parameter, layer-wise robustness within the same network and impact of network depth. Based on those findings, we propose two remedy solutions to protect DNNs from SIPPs, which can mitigate accuracy degradation from 28% to 0.27% for ResNet with merely 0.24-bit SRAM area overhead per parameter.