Performance Analysis of a Four-neuron SET-WTA Neural Network with Lateral Inhibition and Realistic Interconnections

摘要

The development of processing in nanoscale chips is being threatened by the necessary miniaturization of interconnections. In this work, a RC interconnection model is implemented in order to study the electrical performance of a four-neuron SET-WTA neural network. The interconnection's influence upon the circuit behavior is analyzed, through simulations with SIMON, and interconnection-related limits for the operational frequency are established.