Abstract: The work presented by authors is devoted to features of implementation of algorithms of calculus mathematics with the aid of free and proprietary compilers. The analysis of speed of consecutive and parallel applications was carried out on the example of task of matrix multiplication, the solution of systems of the linear algebraic equations by exact and iterative methods. New technologies of parallel programming (auto parallelization, coarrays) are implemented only in modern proprietary compilers. Results of the analysis of high-speed performance of parallel programs of the high-precision calculations developed by authors are presented. Recommendations about the features of realization of task of calculus mathematics were worked out with the aid of various programming languages and compilers. Free and proprietary compilers allow to develop highly effective consecutive applications. Advantages of proprietary compilers are shown when developing parallel applications first of all due to support of the new technologies (auto parallelization, coarrays).
No Comments.