|
Numerical methods and programming, 2008, Volume 9, Issue 1, Pages 129–140
(Mi vmp427)
|
|
|
|
This article is cited in 1 scientific paper (total in 1 paper)
Вычислительные методы и приложения
Parallel algorithms for solving the inverse gravity problem
and the distant communication between the MVS-1000 and the user
E. N. Akimovaa, D. V. Gemaidinovb a Institute of Mathematics and Mechanics, Ural Branch of the Russian Academy of Sciences, Ekaterinburg
b Ural State Technical University, Ekaterinburg
Abstract:
A special Web-server for starting the programs that realize the parallel algorithms for solving the inverse gravity problem (i.e., the problem of allocating the anomalies of a gravitational field in a horizontal or curvilinear layer and the problem of finding the variable density in a layer) is developed for the massively parallel computing system MVS-1000/32 through the Web-interface. A number of parallel direct and iterative methods are located on this Web-server. The parallel algorithms are tested for a model problem in the case of a synthetic gravitational field. The preliminary gravity data processing problem and the problem of finding the variable density in a layer are solved. An analysis of the speed up and efficiency of the parallel algorithms is performed depending on the number of processors. The work was supported by the Russian Foundation for Basic Research (project N 06-01-00116). The paper was prepared on the basis of the authors' report at the International Conference
on Parallel Computing Technologies (PaVT-2008; http://agora.guru.ru/pavt).
Keywords:
inverse gravity problem, direct and iterative methods, parallel algorithms, massively parallel
computing systems, Web-server, efficiency and speed up.
Citation:
E. N. Akimova, D. V. Gemaidinov, “Parallel algorithms for solving the inverse gravity problem
and the distant communication between the MVS-1000 and the user”, Num. Meth. Prog., 9:1 (2008), 129–140
Linking options:
https://www.mathnet.ru/eng/vmp427 https://www.mathnet.ru/eng/vmp/v9/i1/p129
|
Statistics & downloads: |
Abstract page: | 136 | Full-text PDF : | 47 |
|