The implementation of a neurocomputing environment based on a network of heterogeneous computers

With the development of network technology, distributed parallel computing with multi processors as its basis is becoming a new kind of efficient parallel computing method. We introduce a parallel neurocomputing environment-HCPC (heterogeneous computer parallel computing), which is realized in a network of UNIX workstations and microcomputers. Capabilities of HCPC are tested through the realization of an ART1 neural network model.