event-based simulation of continuous time spiking neural network

摘要

In a particular embodiment of the present disclosure, ordered and unordered for faster event sort schedules and multidimensional with the accelerator, multi-schedule architecture, as well as to provide a continuous-time neural network model prediction (in the future) and the forward (update / anticipating / catch-up) as a rule or method of event operation of modeling including the formation, it is to provide a method and apparatus for event-based simulation of continuous-time neural network. In this way, the advantage, include faster simulation of spiking neural network (orders of magnitude), as well as continuous time neurons, synapses, and describes the behavior of a typical neuron network and a method for modeling .