Visible to the public Exploiting Resiliency for Kernel-Wise CNN Approximation Enabled by Adaptive Hardware Design

TitleExploiting Resiliency for Kernel-Wise CNN Approximation Enabled by Adaptive Hardware Design
Publication TypeConference Paper
Year of Publication2021
AuthorsDe la Parra, Cecilia, El-Yamany, Ahmed, Soliman, Taha, Kumar, Akash, Wehn, Norbert, Guntoro, Andre
Conference Name2021 IEEE International Symposium on Circuits and Systems (ISCAS)
KeywordsAI Accelerator, CNN inference, Computer architecture, Hardware, image classification, neural network resiliency, Proposals, pubcrawl, Quantization (signal), resilience, Resiliency, Space exploration
AbstractEfficient low-power accelerators for Convolutional Neural Networks (CNNs) largely benefit from quantization and approximation, which are typically applied layer-wise for efficient hardware implementation. In this work, we present a novel strategy for efficient combination of these concepts at a deeper level, which is at each channel or kernel. We first apply layer-wise, low bit-width, linear quantization and truncation-based approximate multipliers to the CNN computation. Then, based on a state-of-the-art resiliency analysis, we are able to apply a kernel-wise approximation and quantization scheme with negligible accuracy losses, without further retraining. Our proposed strategy is implemented in a specialized framework for fast design space exploration. This optimization leads to a boost in estimated power savings of up to 34% in residual CNN architectures for image classification, compared to the base quantized architecture.
Citation Keyde_la_parra_exploiting_2021