Comparison of scalable fast methods for long-range interactions

Axel Arnold, Florian Fahrenberger, Christian Holm, Olaf Lenz, Matthias Bolten, Holger Dachsel, Rene Halver, Ivo Kabadshow, Franz Gähler, Frederik Heber, Julian Iseringhausen, Michael Hofmann, Michael Pippig, Daniel Potts und Godehard Sutmann
In: Phys. Rev. E (Dez. 2013), 88:6(063308)
 

Abstract

Based on a parallel scalable library for Coulomb interactions in particle systems, a comparison between the fast multipole method (FMM), multigrid-based methods, fast Fourier transform (FFT)-based methods, and a Maxwell solver is provided for the case of three-dimensional periodic boundary conditions. These methods are directly compared with respect to complexity, scalability, performance, and accuracy. To ensure comparable conditions for all methods and to cover typical applications, we tested all methods on the same set of computers using identical benchmark systems. Our findings suggest that, depending on system size and desired accuracy, the FMM- and FFT-based methods are most efficient in performance and stability.

Bilder

Bibtex

@ARTICLE{Arnold:2013,
    author = {Arnold, Axel and Fahrenberger, Florian and Holm, Christian and Lenz, Olaf and Bolten, Matthias and
              Dachsel, Holger and Halver, Rene and Kabadshow, Ivo and G{\"a}hler, Franz and Heber, Frederik and
              Iseringhausen, Julian and Hofmann, Michael and Pippig, Michael and Potts, Daniel and Sutmann,
              Godehard},
     pages = {063308},
     title = {Comparison of scalable fast methods for long-range interactions},
   journal = {Phys. Rev. E},
    volume = {88},
    number = {6},
      year = {2013},
     month = dec,
  abstract = {Based on a parallel scalable library for Coulomb interactions in particle systems, a comparison
              between the fast multipole method (FMM), multigrid-based methods, fast Fourier transform (FFT)-based
              methods, and a Maxwell solver is provided for the case of three-dimensional periodic boundary
              conditions. These methods are directly compared with respect to complexity, scalability,
              performance, and accuracy. To ensure comparable conditions for all methods and to cover typical
              applications, we tested all methods on the same set of computers using identical benchmark systems.
              Our findings suggest that, depending on system size and desired accuracy, the FMM- and FFT-based
              methods are most efficient in performance and stability.},
       doi = {10.1103/PhysRevE.88.063308}
}