Implementation of movie-based matrix algorithms on OpenMP platform

The convenience and programmer's productivity are the main point of visual programming systems and languages. From the other side, the parallel programming is mainly focused on reaching the high performance by optimization of executable code. The Movie-based Programming is based not only on the introduction of special symbols and images with semantic support, but also on a series of images that can present dynamical features of algorithms. The presented paper describes a technique of OpenMP parallelization of Movie-based algorithms in order to obtain the suitable program performance. The results of numerical experiments are also presented showing applicability of the proposed technique including implementation, code validity checking and performance testing.