Please use this identifier to cite or link to this item:
https://hdl.handle.net/10356/85536
Title: | An Energy-Efficient Digital ReRAM-Crossbar-Based CNN With Bitwise Parallelism | Authors: | Ni, Leibin Liu, Zichuan Yu, Hao Joshi, Rajiv V. |
Keywords: | Approximate computing Neural network hardware |
Issue Date: | 2017 | Source: | Ni, L., Liu, Z., Yu, H., & Joshi, R. V. (2017). An Energy-Efficient Digital ReRAM-Crossbar-Based CNN With Bitwise Parallelism. IEEE Journal on Exploratory Solid-State Computational Devices and Circuits, 3, 37-46. | Series/Report no.: | IEEE Journal on Exploratory Solid-State Computational Devices and Circuits | Abstract: | 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. | URI: | https://hdl.handle.net/10356/85536 http://hdl.handle.net/10220/43795 |
DOI: | 10.1109/JXCDC.2017.2697910 | Schools: | School of Electrical and Electronic Engineering | Rights: | © 2017 IEEE. Personal use of this material is permitted. Permission from IEEE must be obtained for all other uses, in any current or future media, including reprinting/republishing this material for advertising or promotional purposes, creating new collective works, for resale or redistribution to servers or lists, or reuse of any copyrighted component of this work in other works. The published version is available at: [http://dx.doi.org/10.1109/JXCDC.2017.2697910]. | Fulltext Permission: | open | Fulltext Availability: | With Fulltext |
Appears in Collections: | EEE Journal Articles |
Files in This Item:
File | Description | Size | Format | |
---|---|---|---|---|
An Energy-Efficient Digital ReRAM-Crossbar-Based CNN With Bitwise Parallelism.pdf | 2.36 MB | Adobe PDF | View/Open |
SCOPUSTM
Citations
10
57
Updated on Mar 21, 2024
Web of ScienceTM
Citations
10
48
Updated on Oct 26, 2023
Page view(s) 50
612
Updated on Mar 28, 2024
Download(s) 20
194
Updated on Mar 28, 2024
Google ScholarTM
Check
Altmetric
Items in DR-NTU are protected by copyright, with all rights reserved, unless otherwise indicated.