An FPGA-based cortical and thalamic silicon neuronal network

A DSSN model is a neuron model which is designed to be implemented efficiently by digital arithmetic circuit. In our previous study, we expanded this model to support the neuronal activities of several cortical and thalamic neurons; Regular spiking, fast spiking, intrinsically bursting and low-threshold spike. In this paper, we report our implementation of this expanded DSSN model and a kinetic-model-based silicon synapse on an FPGA device. Here, synaptic efficacy was stored in block RAMs, and external connection was realized based on a bus that conform to the address event representation. We simulated our circuit by the Xilinx Vivado design suit.