SyncNN: Evaluating and Accelerating Spiking Neural Networks on FPGAs

Abstract

Compared to conventional artificial neural networks, Spiking Neural Networks (SNNs) are more biologically plausible and require less computation due to their event-driven nature of spiking neurons. However, the default asynchronous execution of SNNs also poses great challenges to accelerate their performance on FPGAs. In this work, we present a novel synchronous approach for rate encoding based SNNs, which is more hardware friendly than conventional asynchronous approaches. We first quantitatively evaluate and mathematically prove that the proposed synchronous approach and asynchronous implementation alternatives of rate encoding based SNNs are similar in terms of inference accuracy and we highlight the computational performance advantage of using SyncNN over asynchronous approach. We also design and implement the SyncNN framework to accelerate SNNs on Xilinx ARM-FPGA SoCs in a synchronous fashion. To improve the computation and memory access efficiency, we first quantize the network weights to 16-bit, 8-bit, and 4-bit fixed-point values with the SNN friendly quantization techniques. Moreover, we encode only the activated neurons by recording their positions and corresponding number of spikes to fully utilize the event-driven characteristics of SNNs, instead of using the common binary encoding (i.e., 1 for a spike and 0 for no spike). For the encoded neurons that have dynamic and irregular access patterns, we design parameterized compute engines to accelerate their performance on the FPGA, where we explore various parallelization strategies and memory access optimizations. Our experimental results on multiple Xilinx ARM-FPGA SoC boards demonstrate that our SyncNN is scalable to run multiple networks, such as LeNet, Network in Network, and VGG, on various datasets such as MNIST, SVHN, and CIFAR-10. SyncNN not only achieves competitive accuracy (99.6%) but also achieves state-of-the-art performance (13,086 frames per second) for the MNIST dataset. Finally, we compare the performance of SyncNN with conventional CNNs using the Vitis AI and find that SyncNN can achieve similar accuracy and better performance compared to Vitis AI for image classification using small networks.

Advancing energy efficiency of spiking neural network accelerator via dynamic predictive early stopping

“…On the other hand, SNNs' temporal information processing nature also presents challenges, leading to extended inference times and, in turn, limiting SNNs from achieving the expected high energy efficiency. This is exceptionally notable in conventional non-event-based tasks, where even though software SNNs have achieved state-of-the-art accuracies [4,5,6,7,8,9], hardware SNN accelerators [10,11,12,13,14,15] still fall short in energy efficiency compared to their CNN counterparts [16,17,18], preventing SNNs from being adopted as a general efficient solution.…”

neuroAIx-Framework: design of future neuroscience simulation systems exhibiting execution of the cortical microcircuit model 20× faster than biological real-time

“…Recent iterations of these systems target a broader range of tasks for example in the area of Machine Learning, and feature higher degrees of flexibility and efficiency (Mayr et al, 2019 ; Billaudelle et al, 2020 ). In addition, recent dedicated systems exist that target the simulation at higher degrees of abstraction (Wang et al, 2018 ) or aim at solving Machine Learning tasks (Panchapakesan et al, 2022 ). On one side, the mentioned variety together with advances in computational capabilities and the development of simulator-independent model description languages (e.g., PyNN by Davison et al, 2009 ) pushed the domain of computational neuroscience to study neural network models of increasing complexity.…”