SIMSCRIPT Solutions


SIMSCRIPT II.5: COMPUTER ARCHITECTURE and NETWORK ANALYSIS


Neural Network Simulation of an "NFLOP"

San Diego State University; San Diego, CA

This model simulates an artificial neural network representing an NFLOP. An NFLOP is a generalization of a flip-flop, having more than two elements. Each element of the NFLOP is represented by one neuron. By selecting the sub-field "NEURON", a detailed schematic of the simulated neuron may be viewed.

Initially, the n elements of the NFLOP have random output voltages, representing the initial strength of each neuron. The simulation then solves the system of n nonlinear ordinary differential equations, which determine the time history of each neuron's strength.

During the animated execution, the relative strength of each neuron is displayed as well as the elapsed time in seconds.

Alternatives Considered: None