Mathematical Framework for Optimizing Crossbar Allocation for ReRAM-based CNN Accelerators
-
Published:2023-12-18
Issue:1
Volume:29
Page:1-24
-
ISSN:1084-4309
-
Container-title:ACM Transactions on Design Automation of Electronic Systems
-
language:en
-
Short-container-title:ACM Trans. Des. Autom. Electron. Syst.
Author:
Li Wanqian1ORCID,
Han Yinhe2ORCID,
Chen Xiaoming2ORCID
Affiliation:
1. Institute of Computing Technology, Chinese Academy of Sciences; University of Chinese Academy of Sciences, China
2. Institute of Computing Technology, Chinese Academy of Sciences, China
Abstract
The resistive random-access memory (ReRAM) has widely been used to accelerate convolutional neural networks (CNNs) thanks to its analog in-memory computing capability. ReRAM crossbars not only store layers’ weights, but also perform in-situ matrix-vector multiplications which are core operations of CNNs. To boost the performance of ReRAM-based CNN accelerators, crossbars can be duplicated to explore more intra-layer parallelism. The crossbar allocation scheme can significantly influence both the computing throughput and bandwidth requirements of ReRAM-based CNN accelerators. Under the resource constraints (i.e., crossbars and memory bandwidths), how to find the optimal number of crossbars for each layer to maximize the inference performance for an entire CNN is an unsolved problem. In this work, we find the optimal crossbar allocation scheme by mathematically modeling the problem as a constrained optimization problem and solving it with a dynamic programming based solver. Experiments demonstrate that our model for CNN inference time is almost precise, and the proposed framework can obtain solutions with near-optimal inference time. We also emphasize that communication (i.e., data access) is an important factor and must also be considered when determining the optimal crossbar allocation scheme.
Funder
National Key R&D Program of China
National Natural Science Foundation of China
Key Research Program of Frontier Sciences, CAS
Strategic Priority Research Program of CAS
Innovation Funding of ICT, CAS
Publisher
Association for Computing Machinery (ACM)
Subject
Electrical and Electronic Engineering,Computer Graphics and Computer-Aided Design,Computer Science Applications