Simulation neural networks using C*

Simulations of two neural network models are used to illustrate the benefits of using C* as a programming language to create applications which utilize massive, fine-grained parallel-processing architectures such as the Connection Machine. An overview of the C* language is given, and general data structures are discussed. Two neural network models are presented and contrasted: a relaxation energy model and a forward propagation model. The use of C* for simulation purposes and advantages of simulating neural networks on large parallel processors are discussed.<>