First Experiences with ab initio Molecular Dynamics on OpenPOWER: The Case of CPMD

  • Valéry Weber
  • A. Cristiano I. Malossi
  • Ivano Tavernelli
  • Teodoro Laino
  • Costas Bekas
  • Manish Modani
  • Nina Wilner
  • Tom Heller
  • Alessandro Curioni
Conference paper
Part of the Lecture Notes in Computer Science book series (LNCS, volume 9945)

Abstract

In this article, we present the algorithmic adaptation and code re-engineering required for porting highly successful and popular planewave codes to next-generation heterogeneous OpenPOWER architectures that foster acceleration and high bandwidth links to GPUs. Here we focus on CPMD as the most representative software for ab initio molecular dynamics simulations. We have ported the construction of the electronic density, the application of the potential to the wavefunctions and the orthogonalization procedure to the GPU. The different GPU kernels consist mainly of fast Fourier transforms (FFT) and basic linear algebra operations (BLAS). The performance of the new implementation obtained on Firestone (POWER8/Tesla) is discussed. We show that the communication between the host and the GPU contributes a large fraction of the total run time. We expect a strong attenuation of the communication bottleneck when the NVLink high-speed interconnect will be available.

Keywords

CPMD POWER8 CUDA NVlink FFT Gram–Schmidt 

References

  1. 1.
    CPMD ver. 4.1: Copyright IBM Corp.1990–2016, Copyright MPI für Festkörperforschung Stuttgart (1997–2001). http://www.cpmd.org
  2. 2.
    Bekas, C., Curioni, A.: Very large scale wavefunction orthogonalization in density functional theory electronic structure calculations. Comput. Phys. Commun. 181(6), 1057–1068 (2010)CrossRefMATHGoogle Scholar
  3. 3.
    Goedecker, S.: Fast radix 2, 3, 4, and 5 kernels for fast Fourier transformations on computers with overlapping multiply-add instructions. SIAM J. Sci. Comput. 18(6), 1605–1611 (1997)MathSciNetCrossRefMATHGoogle Scholar
  4. 4.
    Goedecker, S., Teter, M., Hutter, J.: Separable dual-space Gaussian pseudopotentials. Phys. Rev. B 54, 1703–1710 (1996)CrossRefGoogle Scholar
  5. 5.
    Weber, V., Bekas, C., Laino, T., Curioni, A., Bertsch, A., Futral, S.: Shedding light on lithium/air batteries using millions of threads on the BG/Q supercomputer. In: 2014 IEEE 28th International Parallel and Distributed Processing Symposium, pp. 735–744 (2014)Google Scholar

Copyright information

© Springer International Publishing AG 2016

Authors and Affiliations

  • Valéry Weber
    • 1
  • A. Cristiano I. Malossi
    • 1
  • Ivano Tavernelli
    • 1
  • Teodoro Laino
    • 1
  • Costas Bekas
    • 1
  • Manish Modani
    • 2
  • Nina Wilner
    • 3
  • Tom Heller
    • 3
  • Alessandro Curioni
    • 1
  1. 1.IBM Research–ZurichZurichSwitzerland
  2. 2.IBM IndiaBengaluruIndia
  3. 3.IBMArmonkUSA

Personalised recommendations