Biologically compatible neural networks with reconfigurable hardware

This paper presents a reconfigurable hardware neuro-simulator specifically designed to emulate biophysically accurate and biologically compatible neural networks. The platform is based on FPGA technology which is used to create real-time custom neuroprocessors with floating point accuracy and a novel hybrid time-event driven architecture for synaptic integration. Through a series of experiments the dynamics of the neuroprocessors are evaluated and compared with real neuron responses. The problem of interconnecting neurons with individual synapses is tackled with a novel synaptic architecture where all incoming synapses are merged efficiently in one single accumulative process without losing biological information. The case studies demonstrate the suitability of conductance-based models and FPGA platforms to simulate living organisms' behaviour in a biological compatible context.