On regular methods for solving the inverse gravity problems on massively parallel computing systems

Several direct and iterative methods for solving the inverse gravity problems are proposed and numerically realized on the massively parallel computing system MVS-1000. As a result, we find the variable density in horizontal and curvilinear layers on the basis of gravitational data measured on the ground surface. The speed-up and efficiency of the parallel iterative algorithms are analyzed depending on the number of processors in use. This 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-2007; http://agora.guru.ru/pavt).