A High Energy-Efficiency Multi-core Neuromorphic Architecture for Deep SNN Training
This addresses the need for energy-efficient edge training in dynamically changing environments, representing a novel advancement in neuromorphic computing.
The paper tackles the problem of enabling direct training of spiking neural networks (SNNs) on neuromorphic architectures for edge applications, achieving a high energy efficiency of 1.05 TFLOPS/W at FP16 on 28nm and reducing DRAM access by 55-85% compared to an A100 GPU in SNN training.
There is a growing necessity for edge training to adapt to dynamically changing environment. Neuromorphic computing represents a significant pathway for high-efficiency intelligent computation in energy-constrained edges, but existing neuromorphic architectures lack the ability of directly training spiking neural networks (SNNs) based on backpropagation. We develop a multi-core neuromorphic architecture with Feedforward-Propagation, Back-Propagation, and Weight-Gradient engines in each core, supporting high efficient parallel computing at both the engine and core levels. It combines various data flows and sparse computation optimization by fully leveraging the sparsity in SNN training, obtaining a high energy efficiency of 1.05TFLOPS/W@ FP16 @ 28nm, 55 ~ 85% reduction of DRAM access compared to A100 GPU in SNN trainings, and a 20-core deep SNN training and a 5-worker federated learning on FPGAs. Our study develops the first multi-core neuromorphic architecture supporting the direct SNN training, facilitating the neuromorphic computing in edge-learnable applications.