DATA PROCESSING

Parallel-pipeline computations for data flows processing

A. A. Talalaev

Ailamazyan Program Systems Institute of Russian Academy of Sciences

Abstract: This paper consider some principles of the parallel-pipeline computation model for target information flows processing. This principles are used in development of program system of space appointment based on the artificial neural networks. Offered decisions will allow to provide effective computations on multiprocessor computing systems.

Keywords: program system, parallel-pipeline computations, cluster, process control.