We describe a programmable multi-chip VLSI neuronal system that can be used
for exploring spike-based information processing models. The system consis
ts of a silicon retina, a PIC microcontroller, and a transceiver chip whose
integrate-and-fire neurons are connected in a soft winner-take-all archite
cture. The circuit on this multi-neuron chip approximates a cortical microc
ircuit. The neurons can be configured for different computational propertie
s by the virtual connections of a selected set of pixels on the silicon ret
ina. The virtual wiring between the different chips is effected by an event
-driven communication protocol that uses asynchronous digital pulses, simil
ar to spikes in a neuronal system. We used the multi-chip spike-based syste
m to synthesize orientation-tuned neurons using both a feedforward model an
d a feedback model. The performance of our analog hardware spiking model ma
tched the experimental observations and digital simulations of continuous-v
alued neurons. The multi-chip VLSI system has advantages over computer neur
onal models in that it is real-time, and the computational time does not sc
ale with the size of the neuronal network. (C) 2001 Elsevier Science Ltd. A
ll rights reserved.