Efficient Inter-process Communication in Parallel Implementation of Grid-Characteristic Method
We consider application of Message Passing Interface (MPI) to parallelize a numerical code that solves the seismic wave equation with the grid-characteristic method. The problem is time reducing of communication between MPI processes during computation. The solution to this problem is of great interest, when multiple computational grids are involved. The difficulty of problem is increased with independent distribution of nodes across processes. General approach to handle contacts is formulated in terms of interpolation of nodes from some part of one grid to some part of another. In this chapter, we propose an efficient algorithm and give an example of its implementation. The implementation is not restricted to Cartesian grids, so it is possible to make accurate simulations taking into account earth surface topography and complex geometry of contacts between geological layers.
KeywordsParallel programming Grid-characteristic method MPI Structured grids
This work has been carried out using computing resources of the federal collective usage center Complex for Simulation and Data Processing for Mega-science Facilities at NRC “Kurchatov Institute”, http://ckp.nrcki.ru/.
The reported study was funded by RFBR according to the research project No. 18-07-00914 A.
- 1.Moczo, P., Kristek, J., Galis, M., Pazak, P., Balazovjech, M.: The finite-difference and finite-element modeling of seismic wave propagation and earthquake motion. Acta. Phys. Slovaca. Rev. Tutorials 57(2), 177–406 (2010)Google Scholar
- 4.Martin, K., Dumbser, M.: An arbitrary high-order Discontinuous Galerkin method for elastic waves on unstructured meshes—I. The two-dimensional isotropic case with external source terms. Geophys. J. Int. 166(2), 855–877Google Scholar
- 7.Micikevicius, P.: 3D finite difference computation on GPUs using CUDA. In: Proceedings of 2nd Workshop on General Purpose Processing on Graphics Processing Units. ACM, pp. 79–84 (2009)Google Scholar
- 8.Martin, R., Komatitsch, D., Blitz, C., Le Goff, N.: Simulation of seismic wave propagation in an asteroid based upon an unstructured MPI spectral-element method: blocking and non-blocking communication strategies. In: International Conference on High Performance Computing for Computational Science (VECPAR 2008), pp. 350–363. Springer, Berlin, Heidelberg (2008)CrossRefGoogle Scholar
- 9.Voinov, O.Y., Golubev V.I., Petrov I.B.: Elastic imaging using multiprocessor computer systems. In: CEUR Workshop Proceedings, vol. 1787, pp. 491–495 (2016)Google Scholar
- 12.Khokhlov, N., Ivanov, A., Zhdanov, M., Petrov, I., Ryabinkin, E.: Applying OpenCL Technology for Modelling Seismic Processes Using Grid-characteristic Methods. In: International Conference on Distributed Computer and Communication Networks (DCCN 2016), pp. 577–588. Springer, Cham (2016)Google Scholar
- 14.Gabriel, E., Fagg, G.E., Bosilca, G., Angskun, T., Dongarra, J.J., Squyres, J.M., Sahay, V., Kambadur, P., Barrett, B., Lumsdaine, A., Castain, R.H., Daniel, D.J., Graham, R.L., Woodall, T.S.: Open MPI: goals, concept, and design of a next generation MPI implementation. In: European Parallel Virtual Machine/Message Passing Interface Users Group Meeting (EuroPVM/MPI 2004), pp. 97–104. Springer, Berlin, Heidelberg (2004)CrossRefGoogle Scholar
- 15.LeVeque, R.J.: Finite Volume Methods for Hyperbolic Problems, vol. 31. Cambridge University PressGoogle Scholar
- 18.Golubev, V.I., Petrov, I.B.: Improving the simulation accuracy of seismic responses from curvilinear geological boundaries. In: Geomodel 2016: 18th Science and Applied Research Conference on Oil and Gas Geological Exploration and Development. https://doi.org/10.3997/2214-4609.201602221. (In Russian)