High-Performance Spiking Neural Network Simulator

摘要

Simulation of neural networks is a significant task for contemporary artificial intelligence research. Despite the availability of modern processing hardware, the task is still too demanding to be done in a sequential way. Therefore, a parallel computation approach is almost always necessary. Modern graphical accelerators (GPUs) represent highly parallel machines with a significant computational performance that can be unleashed only under certain conditions including threads scheduling, proper sources occupation, aligned data access, communication management, etc. We have proposed a novel acceleration approach for large neural networks. It is using a GPU and incorporating biologically highly precise spiking neurons that can imitate real biological neurons. The simulator can be, for example, used for research of communication dynamics of large neural networks with tens of thousands of spiking neurons.