vmaitComputational Mathematics and Information TechnologiesComputational Mathematics and Information Technologies2587-8999Донской государственный технический университет10.23947/2587-8999-2022-1-1-18-28vmait-24Research ArticleСтатьиПрименение сеточно-характеристического метода для решения задач распространения динамических волновых возмущений с использованием систем HPCApplication of the grid-characteristic method for solving the problems of the propagation of dynamic waves using HPC systemsХохловН. И.KhokhlovN. I.

Хохлов Николай Игоревич - кандидат физико-технических наук.

141701, Московская облаcть, Долгопрудный, Институтский пер., 9

Nikolay I. Khokhlov - PhD, Moscow Institute of Physics and Technology.

9 Institutskiy per., Dolgoprudny, Moscow Region

khokhlov.ni@mipt.ruМосковский физико-технический институтРоссияMoscow Institute of Physics and TechnologyRussian Federation202220022023611828Copyright © Хохлов Н.И., 20222022Хохлов Н.И.Khokhlov N.I.Данная работа распространяется под лицензией Creative Commons Attribution 4.0.This work is licensed under a Creative Commons Attribution 4.0 License.https://www.cmit-journal.ru/jour/article/view/24

В работе рассматривается применение различных современных технологий высокопроизводительных вычислений для ускорения численного решения задач распространения динамических волновых возмущений с использованием сеточно-характеристического метода. Рассматриваются технологии как для центральных процессоров (CPU), так и для графических процессоров (GPU). Приведены сравнительные результаты применения технологий MPI, OpenMP, CUDA. В качестве примеров работы разработанного программного комплекса приводится ряд примеров расчета задач сейсмики и геофизики. Отдельно рассмотрен вопрос распараллеливания задач с наличием контактов многих сеток и топографии дневной поверхности, используя криволинейные сетки.

The paper considers the application of various modern technologies of high-performance computing to accelerate the numerical solution of the problems of propagation of dynamic wave disturbances using the grid-characteristic method. Technologies are considered both for central processing units (CPUs) and for graphic processors (GPUs). Comparative results of applying MPI, OpenMP, CUDA technologies are presented. As examples of the work of the developed software package, a number of examples of calculating the problems of seismic and geophysics are given. Separately, the issue of parallelizing problems with the presence of contacts of many grids and the topography of the day surface using curvilinear grids is considered.

сеточно-характеристический методсейсмикагеофизикаMPIOpenMPCUDAgrid-characteristic methodseismicgeophysicsMPIOpenMPCUDAReferencesV. A. Biryukov, V. A. Miryakha, I. B. Petrov, and N. I. Khokhlov, Simulation of elastic wave propagation in geological media: Intercomparison of three numerical methods // Comput. Math. Math. Phys., Vol. 56, No. 6, 2016.V. A. Biryukov, V. A. Miryakha, I. B. Petrov, and N. I. Khokhlov, Simulation of elastic wave propagation in geological media: Intercomparison of three numerical methods // Comput. Math. Math. Phys., Vol. 56, No. 6, 2016.Голубев В.И., Петров И.Б., Хохлов Н.И. Численное моделирование сейсмической активности сеточно-характеристическим методом // Журнал вычислительной математики и математической физики, 2013. Т. 53, № 10. С. 1709 – 1720.Голубев В.И., Петров И.Б., Хохлов Н.И. Численное моделирование сейсмической активности сеточно-характеристическим методом // Журнал вычислительной математики и математической физики, 2013. Т. 53, № 10. С. 1709 – 1720.P.L. Roe. Characteristic-Based Schemes for the Euler Equations // Annual Review of Fluid Mechanics. 1986. No.18. pp. 337-365.P.L. Roe. Characteristic-Based Schemes for the Euler Equations // Annual Review of Fluid Mechanics. 1986. No.18. pp. 337-365.LeVeque R. J. Finite volume methods for hyperbolic problems. Vol. 31. Cambridge university press, 2002.LeVeque R. J. Finite volume methods for hyperbolic problems. Vol. 31. Cambridge university press, 2002.LeVeque R. J. Finite difference methods for ordinary and partial differential equations: steady-state and time-dependent problems. Vol. 98. Siam, 2007.LeVeque R. J. Finite difference methods for ordinary and partial differential equations: steady-state and time-dependent problems. Vol. 98. Siam, 2007.Strang G. On the construction and comparison of difference schemes // SIAM Journalon Numerical Analysis. 1968. Vol. 5, No. 3. pp. 506–517.Strang G. On the construction and comparison of difference schemes // SIAM Journalon Numerical Analysis. 1968. Vol. 5, No. 3. pp. 506–517.Courant R., Isaacson E., Rees M. On the solution of nonlinear hyperbolic differential equations by finite differences // Communications on Pure and Applied Mathematics. 1952. Vol. 5, No. 3. pp. 243–255.Courant R., Isaacson E., Rees M. On the solution of nonlinear hyperbolic differential equations by finite differences // Communications on Pure and Applied Mathematics. 1952. Vol. 5, No. 3. pp. 243–255.Rusanov V. V. Difference schemes of the third order of accuracy for the forward calculation of discontinuous solutions // Doklady Akademii Nauk. Vol. 180. Russian Academy of Sciences. 1968. pp. 1303-1305.Rusanov V. V. Difference schemes of the third order of accuracy for the forward calculation of discontinuous solutions // Doklady Akademii Nauk. Vol. 180. Russian Academy of Sciences. 1968. pp. 1303-1305.Favorskaya A. V., Petrov I. B. Grid-characteristic method // Innovations in Wave Processes Modelling and Decision Making / ed. by A. V. Favorskaya, I. B. Petrov. Springer, 2018. pp. 117–160.Favorskaya A. V., Petrov I. B. Grid-characteristic method // Innovations in Wave Processes Modelling and Decision Making / ed. by A. V. Favorskaya, I. B. Petrov. Springer, 2018. pp. 117–160.Dagum L., Menon R. OpenMP: an industry standard API for shared-memory programming // IEEE computational science and engineering. 1998. Vol. 5. No. 1. pp. 46–55.Dagum L., Menon R. OpenMP: an industry standard API for shared-memory programming // IEEE computational science and engineering. 1998. Vol. 5. No. 1. pp. 46–55.Nakata N., Tsuji T., Matsuoka T. Acceleration of computation speed for elastic wave simulation using a Graphic Processing Unit // Exploration Geophysics. 2011. Vol. 42, No. 1. pp. 98–104.Nakata N., Tsuji T., Matsuoka T. Acceleration of computation speed for elastic wave simulation using a Graphic Processing Unit // Exploration Geophysics. 2011. Vol. 42, No. 1. pp. 98–104.Weiss R. M., Shragge J. Solving 3D anisotropic elastic wave equations on parallel GPU devices // Geophysics. 2013. Vol. 78, No. 2. F7–F15.Weiss R. M., Shragge J. Solving 3D anisotropic elastic wave equations on parallel GPU devices // Geophysics. 2013. Vol. 78, No. 2. F7–F15.Finite-difference staggered grids in GPUs for anisotropic elastic wave propagation simulation / F. Rubio [et al.] // Computers & geosciences. 2014. Vol. 70. pp. 181–189.Finite-difference staggered grids in GPUs for anisotropic elastic wave propagation simulation / F. Rubio [et al.] // Computers & geosciences. 2014. Vol. 70. pp. 181–189.Komatitsch D., Michéa D., Erlebacher G. Porting a high-order finite-element earthquake modeling application to NVIDIA graphics cards using CUDA // Journal of Parallel and Distributed Computing. 2009. Vol. 69, No. 5. pp. 451-460.Komatitsch D., Michéa D., Erlebacher G. Porting a high-order finite-element earthquake modeling application to NVIDIA graphics cards using CUDA // Journal of Parallel and Distributed Computing. 2009. Vol. 69, No. 5. pp. 451-460.Modeling the propagation of elastic waves using spectral elements on a cluster of 192 GPUs / D. Komatitsch [et al.] // Computer Science-Research and Development. 2010. Vol. 25, No. 1/2. pp. 75–82.Modeling the propagation of elastic waves using spectral elements on a cluster of 192 GPUs / D. Komatitsch [et al.] // Computer Science-Research and Development. 2010. Vol. 25, No. 1/2. pp. 75–82.Message P Forum. 1994. MPI: a Message-Passing Interface Standard. Technical Report. University of Tennessee, Knoxville, TN, USA.Message P Forum. 1994. MPI: a Message-Passing Interface Standard. Technical Report. University of Tennessee, Knoxville, TN, USA.Якобовский М.В. Введение в параллельные методы решения задач // Предисл.: В. А. Садовничий. М.: Издательство Московского университета, 2013. 328 с.Якобовский М.В. Введение в параллельные методы решения задач // Предисл.: В. А. Садовничий. М.: Издательство Московского университета, 2013. 328 с.Ivanov A.M., Khokhlov N.I. Efficient Inter-process Communication in Parallel Implementation of Grid-Characteristic Method, 2019. pp. 91–102.Ivanov A.M., Khokhlov N.I. Efficient Inter-process Communication in Parallel Implementation of Grid-Characteristic Method, 2019. pp. 91–102.Ivanov A.M., Khokhlov N.I. Parallel implementation of the grid-characteristic method in the case of explicit contact boundaries // Comput. Res. Model. 2018. V. 10. № 5. pp. 667–678.Ivanov A.M., Khokhlov N.I. Parallel implementation of the grid-characteristic method in the case of explicit contact boundaries // Comput. Res. Model. 2018. V. 10. № 5. pp. 667–678.Khokhlov N. и др. Solution of Large-scale Seismic Modeling Problems // Procedia Comput. Sci. 2015. V. 66. pp. 191–199.Khokhlov N. и др. Solution of Large-scale Seismic Modeling Problems // Procedia Comput. Sci. 2015. V. 66. pp. 191–199.Khokhlov N. и др. Applying OpenCL Technology for Modelling Seismic Processes Using Grid-Characteristic Methods, 2016. pp. 577–588.Khokhlov N. и др. Applying OpenCL Technology for Modelling Seismic Processes Using Grid-Characteristic Methods, 2016. pp. 577–588.

The authors declare that there are no conflicts of interest present.