A biophysically based dendrite model using programmable floating-gate devices

This paper presents our programmable floating-gate dendrite model based upon our transistor channel models. We are developing a flexible, low-power, real-time computational engine for studying, understanding, and implementing neural computation. Here we present a systematic study of the dynamics seen in this silicon model, emphasizing the connection to conventional neurobiological modeling of dendrites. These results provide a foundation or future work on larger systems.