Cluster Parallelism of the Neuroprocessor Systems Based on the Fuzzy Logic

Abstract

The article is devoted to the development of the neuroprocessor system (NPS) model based on the clustering and the analysis of the obtained fuzzy structures. The production model of explicit and implicit parallelism is suggested. The given model could be used for the intelligent compiler development and for the choice of the optimal technical parameters, such as NPS modules number, the firmware memory volume, the system productivity etc.