Navegar

Colección

Datos Estadísticas

Artículo

La versión final de este artículo es de uso interno de la institución.
Consulte el artículo en la página del editor
Consulte la política de Acceso Abierto del editor

Abstract:

A hybrid scheme that utilizes MPI for distributed memory parallelism and OpenMP for shared memory parallelism is presented. The work is motivated by the desire to achieve exceptionally high Reynolds numbers in pseudospectral computations of fluid turbulence on emerging petascale, high core-count, massively parallel processing systems. The hybrid implementation derives from and augments a well-tested scalable MPI-parallelized pseudospectral code. The hybrid paradigm leads to a new picture for the domain decomposition of the pseudospectral grids, which is helpful in understanding, among other things, the 3D transpose of the global data that is necessary for the parallel fast Fourier transforms that are the central component of the numerical discretizations. Details of the hybrid implementation are provided, and performance tests illustrate the utility of the method. It is shown that the hybrid scheme achieves good scalability up to ∼20,000 compute cores with a maximum efficiency of 89%, and a mean of 79%. Data are presented that help guide the choice of the optimal number of MPI tasks and OpenMP threads in order to maximize code performance on two different platforms. © 2011 Elsevier B.V. All rights reserved.

Registro:

Documento: Artículo
Título:A hybrid MPI-OpenMP scheme for scalable parallel pseudospectral computations for fluid turbulence
Autor:Mininni, P.D.; Rosenberg, D.; Reddy, R.; Pouquet, A.
Filiación:Institute for Mathematics Applied to Geosciences, National Center for Atmospheric Research, P.O. Box 3000, Boulder, CO 80307-3000, United States
Pittsburgh Supercomputing Center, 300 S. Craig Street, Pittsburgh, PA 15213, United States
Departamento de Física, Facultad de Ciencias Exactas y Naturales, Ciudad Universitaria, 1428 Buenos Aires, Argentina
Palabras clave:Computational fluids; MPI; Numerical simulation; OpenMP; Parallel scalability; Central component; Code performance; Computational fluid; Discretizations; Distributed Memory; Domain decompositions; Fluid turbulence; Global data; High Reynolds number; Hybrid implementation; Hybrid paradigm; Hybrid scheme; Massively parallel processing systems; Maximum Efficiency; MPI; OpenMP; Optimal number; Parallel scalability; Performance tests; Petascale; Pseudospectral; Shared memory parallelism; Application programming interfaces (API); Domain decomposition methods; Fast Fourier transforms; Fluids; Optimal systems; Reynolds number; Scalability; Turbulence; Parallel processing systems
Año:2011
Volumen:37
Número:6-7
Página de inicio:316
Página de fin:326
DOI: http://dx.doi.org/10.1016/j.parco.2011.05.004
Título revista:Parallel Computing
Título revista abreviado:Parallel Comput
ISSN:01678191
CODEN:PACOE
Registro:https://bibliotecadigital.exactas.uba.ar/collection/paper/document/paper_01678191_v37_n6-7_p316_Mininni

Referencias:

  • Canuto, C., Hussaini, M.Y., Quateroni, A., Zang, T.A., (1988) Spectral Methods in Fluid Dynamics, , Springer New York
  • Dmitruk, P., Wang, L.-P., Matthaeus, W.H., Zhang, R., Seckel, D., Scalable parallel FFT for spectral simulations on a Beowulf cluster (2001) Parallel Computing, 27 (14), pp. 1921-1936. , DOI 10.1016/S0167-8191(01)00120-X, PII S016781910100120X
  • Donzis, D.A., Yeung, P.K., Pekurovksy, D., Turbulence simulations at o(104) core counts (2008) TeraGrid G08 Conference, , Las Vegas, NV Science track paper
  • Frigo, M., Johnson, S.G., The design and implementation of FFTW3 (1998) Proceedings of the IEEE International Conference Acoustics Speech and Signal Processing, 3, p. 1381
  • Frigo, M., Johnson, S.G., The design and implementation of FFTW3 (2005) Proceedings of the IEEE, 93 (2), pp. 216-231. , DOI 10.1109/JPROC.2004.840301, Program Generation, Optimization and Platform Adaptation
  • Gomez, D.O., Mininni, P.D., Dmitruk, P., Parallel simulations in turbulent MHD (2005) Physica Scripta T, T116, pp. 123-127. , DOI 10.1238/Physica.Topical.116a00123, International Workshop on Theoretical Plasma Physics: Modern Plasma Science
  • Gottlieb, D., Hussaini, M.Y., Orszag, S.A., (1984) Spectral Methods for Partial Differential Equations, , SIAM Philadelphia
  • Gottlieb, D., Orszag, S.A., (1977) Numerical Analysis of Spectral Methods: Theory and Application, , SIAM Philadelphia
  • Hager, G., Jost, G., Rabenseifner, R., Communication characteristics and hybrid MPI/OpenMP parallel programming on clusters of multi-core SMP nodes (2009) Cray User Group Proceedings, , http://www.cug.org/5-publications/proceedings_attendee_lists/CUG09CD/ CUG2009/pages/1-program/final_program/20.tuesday.html
  • Kolmogorov, A.N., Dissipation of energy in locally isotropic turbulence (1941) Dokl. Akad. Nauk SSSR, 32, pp. 16-18
  • Kolmogorov, A.N., The local structure of turbulence in incompressible viscous fluid for very large Reynolds number (1941) Dokl. Akad. Nauk SSSR, 30, pp. 9-13
  • Mininni, P.D., Alexakis, A., Pouquet, A., Nonlocal interactions in hydrodynamic turbulence at high Reynolds numbers: The slow emergence of scaling laws (2008) Phys. Rev. e, 77, p. 036306
  • Orszag, S.A., Comparison of pseudospectral and spectral approximation (1972) Stud. Appl. Math., 51, pp. 253-259
  • Patterson, G., Orszag, S.A., Spectral calculations of isotropic turbulence: Efficient removal of aliasing interactions (1971) Phys. Fluids, 14, pp. 2538-2541
  • Takahashi, D., A hybrid MPI/OpenMP implementation of a parallel 3-D FFT on SMP clusters (2006) Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics), 3911, pp. 970-977. , Parallel Processing and Applied Mathematics - 6th International Conference, PPAM 2005, Revised Selected Papers
  • Yeung, P.K., Donzis, D.A., Sreenivasan, K.R., High Reynolds number simulation of turbulent mixing (2005) Phys. Fluids, 17, p. 081703
  • Yilmaz, E., Payli, R.U., Akay, H.U., Ecer, A., Hybrid parallelism for CFD simulations: Combining MPI with OpenMP (2007) Lecture Notes in Computational Science and Engineering, 67, pp. 401-408. , Parallel Computational Fluid Dynamics

Citas:

---------- APA ----------
Mininni, P.D., Rosenberg, D., Reddy, R. & Pouquet, A. (2011) . A hybrid MPI-OpenMP scheme for scalable parallel pseudospectral computations for fluid turbulence. Parallel Computing, 37(6-7), 316-326.
http://dx.doi.org/10.1016/j.parco.2011.05.004
---------- CHICAGO ----------
Mininni, P.D., Rosenberg, D., Reddy, R., Pouquet, A. "A hybrid MPI-OpenMP scheme for scalable parallel pseudospectral computations for fluid turbulence" . Parallel Computing 37, no. 6-7 (2011) : 316-326.
http://dx.doi.org/10.1016/j.parco.2011.05.004
---------- MLA ----------
Mininni, P.D., Rosenberg, D., Reddy, R., Pouquet, A. "A hybrid MPI-OpenMP scheme for scalable parallel pseudospectral computations for fluid turbulence" . Parallel Computing, vol. 37, no. 6-7, 2011, pp. 316-326.
http://dx.doi.org/10.1016/j.parco.2011.05.004
---------- VANCOUVER ----------
Mininni, P.D., Rosenberg, D., Reddy, R., Pouquet, A. A hybrid MPI-OpenMP scheme for scalable parallel pseudospectral computations for fluid turbulence. Parallel Comput. 2011;37(6-7):316-326.
http://dx.doi.org/10.1016/j.parco.2011.05.004