A TTFS-based energy and utilization efficient neuromorphic CNN accelerator

Frontiers in Neuroscience

Neuroscience

Kyle Timothy Ng Chu¹  Miao Yu²  Tingting Xiang²  Yaswanth Tavva²  Trevor E. Carlson²  Venkata Pavan Kumar Miriyala²  Burin Amornpaisannon²  Srivatsa P.³ 

[1] Centre for Quantum Technologies, National University of Singapore, Singapore, Singapore;School of Computing, Department of Computer Science, National University of Singapore, Singapore, Singapore;School of Interactive Computing, Georgia Institute of Technology, Atlanta, GA, United States;

关键词: artificial neural networks (ANNs);    brain-inspired networks;    neuromorphic hardware;    spiking neural networks (SNNs);    time-to-first-spike;   

DOI  :  10.3389/fnins.2023.1121592

 received in 2022-12-12, accepted in 2023-04-10,  发布年份 2023

来源: Frontiers

【 摘 要 】

Spiking neural networks (SNNs), which are a form of neuromorphic, brain-inspired AI, have the potential to be a power-efficient alternative to artificial neural networks (ANNs). Spikes that occur in SNN systems, also known as activations, tend to be extremely sparse, and low in number. This minimizes the number of data accesses typically needed for processing. In addition, SNN systems are typically designed to use addition operations which consume much less energy than the typical multiply and accumulate operations used in DNN systems. The vast majority of neuromorphic hardware designs support rate-based SNNs, where the information is encoded by spike rates. Generally, rate-based SNNs can be inefficient as a large number of spikes will be transmitted and processed during inference. One coding scheme that has the potential to improve efficiency is the time-to-first-spike (TTFS) coding, where the information isn't presented through the frequency of spikes, but instead through the relative spike arrival time. In TTFS-based SNNs, each neuron can only spike once during the entire inference process, and this results in high sparsity. The activation sparsity of TTFS-based SNNs is higher than rate-based SNNs, but TTFS-based SNNs have yet to achieve the same accuracy as rate-based SNNs. In this work, we propose two key improvements for TTFS-based SNN systems: (1) a novel optimization algorithm to improve the accuracy of TTFS-based SNNs and (2) a novel hardware accelerator for TTFS-based SNNs that uses a scalable and low-power design. Our work in TTFS coding and training improves the accuracy of TTFS-based SNNs to achieve state-of-the-art results on the MNIST and Fashion-MNIST datasets. Meanwhile, our work reduces the power consumption by at least 2.4×, 25.9×, and 38.4× over the state-of-the-art neuromorphic hardware on MNIST, Fashion-MNIST, and CIFAR10, respectively.

【 授权许可】

Unknown   
Copyright © 2023 Yu, Xiang, P., Chu, Amornpaisannon, Tavva, Miriyala and Carlson.