Energy-efficient XNOR-free in-memory BNN accelerator with input distribution regularization
Authors:
Open Access
- 2 November 2020
- conference paper
- Published by Association for Computing Machinery (ACM)
- p. 1-9
Abstract
SRAM-based in-memory Binary Neural Network (BNN) accelerators are garnering interests as a platform for energy-efficient edge neural network computing thanks to their compactness in terms of hardware and neural network parameter size. However, previous works had to modify SRAM cells to support XNOR operations on memory array resulting in limited area and energy efficiencies. In this work, we present a conversion method which replaces the signed inputs (+1/-1) of BNN with the unsigned inputs (1/0) without computation error, and vice versa. The method enables BNN computing on conventional 6T SRAM arrays and improves area and energy efficiencies. We also demonstrate that further energy saving is possible by skewing the distribution of binary input data based on regularization during network training. Evaluation results show that the proposed techniques improve the inference energy efficiency by up to 9.4x for various benchmarks over previous works.Keywords
Funding Information
- Ministry of Science and ICT (NRF-2016M3A7B4910249)
This publication has 12 references indexed in Scilit:
- NAND-Net: Minimizing Computational Complexity of In-Memory Processing for Binary Neural Networks2019
- In-memory batch-normalization for resistive memory based binary neural network hardware2019
- Input-Splitting of Large Neural Networks for Power-Efficient Accelerator with Resistive Crossbar Memory Array2018
- Parallelizing SRAM Arrays with Customized Bit-Cell for Binary Neural Networks2018
- A 65nm 4Kb algorithm-dependent computing-in-memory SRAM unit-macro with 2.3ns and 55.8TOPS/W fully parallel product-sum operation for binary DNN edge processors2018
- FINN2017
- XNOR-Net: ImageNet Classification Using Binary Convolutional Neural Networks2016
- Acceleration of Deep Neural Network Training with Resistive Cross-Point Devices: Design ConsiderationsFrontiers in Neuroscience, 2016
- ISAACACM SIGARCH Computer Architecture News, 2016
- Deep Residual Learning for Image Recognition2016