Improving the Performance of CNN Accelerator Architecture under the Impact of Process Variations
Open Access
- 9 September 2023
- journal article
- research article
- Published by Association for Computing Machinery (ACM) in ACM Transactions on Design Automation of Electronic Systems
- Vol. 28 (5) , 1-21
Abstract
Convolutional neural network (CNN) accelerators are popular specialized platforms for efficient CNN processing. As semiconductor manufacturing technology scales down to nano scale, process variation dramatically affects the chip’s quality. Process variation causes delay variation within the chip due to transistor parameter differences. CNN accelerators adopt a large number of processing elements (PEs) for parallel computing, which are highly susceptible to process variation effects. Fast CNN processing desires consistent performance among PEs; otherwise the processing speed is limited by the slowest PE within the chip. In this work, we first quantitatively model and analyze the impact of process variation on CNN accelerators’ operating frequency. We further analyze the utilization of CNN accelerators and the characteristics of CNN models. We then leverage the PE underutilization to propose a sub-matrix reformation mechanism and leverage the pixel similarity of images to propose a weight transfer technique. Both techniques are able to tolerate the low-frequency PEs and achieve performance improvement at chip level. Furthermore, a novel resilience-aware mapping technique that exploits the diversity in the importance of weights is also proposed to improve the performance. Evaluation results show that our techniques are able to achieve significant processing speed improvement with negligible accuracy loss.Keywords
Funding Information
- Jilin Provincial Scientific and Technological Development Program (20220201139GX)
This publication has 24 references indexed in Scilit:
- In-Datacenter Performance Analysis of a Tensor Processing Unit2017
- FlexFlow: A Flexible Dataflow Accelerator Architecture for Convolutional Neural Networks2017
- OpenRAM: an open-source memory compiler2016
- ShiDianNao2015
- Mitigating the Susceptibility of GPGPUs Register File to Process Variations2015
- ImageNet Large Scale Visual Recognition ChallengeInternational Journal of Computer Vision, 2015
- DaDianNao: A Machine-Learning Supercomputer2014
- AxNN2014
- Gradient-based learning applied to document recognitionProceedings of the IEEE, 1998
- A clustered failure model for the memory array reconfiguration problemIEEE Transactions on Computers, 1993