Receptive-Field and Switch-Matrices based ReRAM Accelerator with Low Digital-Analog Conversion for CNNs
Yingxun Fu1,a, Xun Liu1,b, Jiwu Shu2, Zhirong Shen3, Shiye Zhang1,c, Jun Wu1,d and Li Ma1,e
1College of Information Science, North China University of Technology, Beijing, China
amooncape@hotmail.com
b1363937238@qq.com
czangxy 2020@qq.com
dWuJune000@outlook.com
emali@ncut.edu.cn
2Department of Computer Science and Technology, Tsinghua University, Beijing, China
shujw@tsinghua.edu.cn
3Computer Science Department, Xiamen University, Xiamen, China
zhirong.shen2601@gmail.com
ABSTRACT
Process-in-Memory (PIM) based accelerator becomes one of the best solutions for the execution of convolution neural networks (CNN). Resistive random access memory (ReRAM) is a classic type of non-volatile random-access memory, which is very suitable for implementing PIM architectures. However, existing ReRAM-based accelerators mainly consider to improve the calculation efficiency, but ignore the fact that the digital-analog signal conversion process spends a lot of energy and executing time. In this paper, we propose a novel ReRAM-based accelerator named Receptive-Field and Switch-Matrices based CNN Accelerator (RFSM). In RFSM, we first propose a receptive-field based convolution strategy to analyze the data relationships, and then gives a dynamic and configurable crossbar combination method to reduce the digital-analog conversion operations. The evaluation result shows that, compared to existing works, RFSM gains up to 6.7x higher speedup and 7.1x lower energy consumption.
Keywords: CNN, PIM, Receptive-Field, Crossbar, Switch-Matrices.
