An Energy-Efficient Digital ReRAM-Crossbar-Based CNN With Bitwise Parallelism

IEEE Journal on Exploratory Solid-State Computational Devices and Circuits

Rajiv V. Joshi¹  Zichuan Liu²  Hao Yu²  Leibin Ni² 

[1] IBM Thomas J. Watson Research Center, Yorktown Heights, NY, USA;School of Electrical and Electronic Engineering, Nanyang Technological University, Singapore;

关键词: Approximate computing;    neural network hardware;    nonvolatile memory (NVM);    resistive random access memory (ReRAM);    supervised learning;   

DOI  :  10.1109/JXCDC.2017.2697910

来源: DOAJ

【 摘 要 】

There is great attention to develop hardware accelerator with better energy efficiency, as well as throughput, than GPUs for convolutional neural network (CNN). The existing solutions have relatively limited parallelism as well as large power consumption (including leakage power). In this paper, we present a resistive random access memory (ReRAM)-accelerated CNN that can achieve significantly higher throughput and energy efficiency when the CNN is trained with binary constraints on both weights and activations, and is further mapped on a digital ReRAM-crossbar. We propose an optimized accelerator architecture tailored for bitwise convolution that features massive parallelism with high energy efficiency. Numerical experiment results show that the binary CNN accelerator on a digital ReRAM-crossbar achieves a peak throughput of 792 GOPS at the power consumption of 4.5 mW, which is 1.61 times faster and 296 times more energy-efficient than a high-end GPU.

【 授权许可】

Unknown