Representation models for parallel processes in parallel, cloud and distributed computing systems based on neuroprocessors

We consider models of logical organization of information processing systems based on neuroprocessors, and analyze their effectiveness. The following study is based on their known general properties and principles of operation. To solve these problems, we employ the set-theoretic approach, the theory of graphs and finite automata, the theory of planning parallel computational processes, and the theory of mathematical and system analysis. We propose analytical expressions to describe the efficiency of the functioning of neuroprocessor systems as functions of the execution time of subroutines, or of the number of neurons emulated when implementing neural networks without/with feedback.