Compilation-semantic verification of time-parameterized multi-parallel programs for information management systems

Abstract

Abstract: The article provides a definition of time-parameterized multi-parallel programs, which (in contrast to the generally accepted interpretation of parallel programs) contain specifications for the start of operations/functions, as well as the subsets of such operations/functions. The need to develop new methods for verifying time-parameterized parallel programs is substantiated. The stages of performing the task of compilation-semantic verification of time-parameterized multi-parallel programs in the interests of creating systems of automatic synthesis of highly efficient parallel programs for computer systems of various classes are presented. Topicality. The need to solve the scientific problem of developing new methods for verifying time-parameterized multi-parallel programs, focused on the application of the format of semantic-numerical specification structures, is necessary. The goal is to describe the method of compilation-semantic verification of time-parameterized multi-parallel programs in the interests of creating systems of automatic synthesis of highly efficient parallel programs for computer systems of various classes. Research method. Compilation-semantic verification of time-parameterized multiparallel programs. Results. The real time is used as one of the main parameters of the formal synthesis of parallel programs and time-parallel processes corresponding to them. The actually used methods of parallel data processing are clearly presented in the constructions of parallel programs. The units of measurement (semantics) of the processed data are clearly presented in the constructions of parallel programs. An example illustrating the main stages of the method is given. Conclusions. The developed method of compilation-semantic verification of time-parameterized multi-parallel programs takes into account the groups of factors that have a significant impact on the effectiveness of software tools of parallel computing systems.