An FPGA-based systolic architecture for the high speed simulation of spiking neural networks is presented. The design is an implementation of Izhikevich's neuron model and employs optimizations for the typical case where neuron activity is low. Since execution time required is related to the activity level, performance of the design can be improved by an order of magnitude.
No takes yet. Share an insight, caveat, or question.
Cheung et al. (2009) studied this question.