Advertisement

591 TFLOPS Multi-trillion Particles Simulation on SuperMUC

  • Wolfgang Eckhardt
  • Alexander Heinecke
  • Reinhold Bader
  • Matthias Brehm
  • Nicolay Hammer
  • Herbert Huber
  • Hans-Georg Kleinhenz
  • Jadran Vrabec
  • Hans Hasse
  • Martin Horsch
  • Martin Bernreuther
  • Colin W. Glass
  • Christoph Niethammer
  • Arndt Bode
  • Hans-Joachim Bungartz
Part of the Lecture Notes in Computer Science book series (LNCS, volume 7905)

Abstract

Anticipating large-scale molecular dynamics simulations (MD) in nano-fluidics, we conduct performance and scalability studies of an optimized version of the code ls1 mardyn. We present our implementation requiring only 32 Bytes per molecule, which allows us to run the, to our knowledge, largest MD simulation to date. Our optimizations tailored to the Intel Sandy Bridge processor are explained, including vectorization as well as shared-memory parallelization to make use of Hyperthreading. Finally we present results for weak and strong scaling experiments on up to 146016 Cores of SuperMUC at the Leibniz Supercomputing Centre, achieving a speed-up of 133k times which corresponds to an absolute performance of 591.2 TFLOPS.

Keywords

molecular dynamics simulations highly scalable simulation vectorization Intel AVX SuperMUC 

Preview

Unable to display preview. Download preview PDF.

Unable to display preview. Download preview PDF.

References

  1. 1.
    Buchholz, M., Bungartz, H.-J., Vrabec, J.: Software design for a highly parallel molecular dynamics simulation framework in chemical engineering. Journal of Computational Science 2(2), 124–129 (2011)CrossRefGoogle Scholar
  2. 2.
    Eckhardt, W., Heinecke, A.: An efficient vectorization of linked-cell particle simulations. In: ACM International Conference on Computing Frontiers, Cagliari, pp. 241–243 (May 2012)Google Scholar
  3. 3.
    Eckhardt, W., Neckel, T.: Memory-efficient implementation of a rigid-body molecular dynamics simulation. In: Proceedings of the 11th International Symposium on Parallel and Distributed Computing - ISPDC 2012, Munich, pp. 103–110. IEEE (2012)Google Scholar
  4. 4.
    Germann, T.C., Kadau, K.: Trillion-atom molecular dynamics becomes a reality. International Journal of Modern Physics C 19(09), 1315–1319 (2008)MATHCrossRefGoogle Scholar
  5. 5.
    Gou, C., Kuzmanov, G., Gaydadjiev, G.N.: SAMS multi-layout memory: providing multiple views of data to boost SIMD performance. In: Proceedings of the 24th ACM International Conference on Supercomputing, ICS 2010, pp. 179–188. ACM, New York (2010)CrossRefGoogle Scholar
  6. 6.
    Heinecke, A., Pflüger, D.: Emerging architectures enable to boost massively parallel data mining using adaptive sparse grids. International Journal of Parallel Programming 41(3), 357–399 (2013)CrossRefGoogle Scholar
  7. 7.
    Heinecke, A., Trinitis, C.: Cache-oblivious matrix algorithms in the age of multi- and many-cores. Concurrency and Computation: Practice and Experience (2013); accepted for publicationGoogle Scholar
  8. 8.
    Horsch, M., Vrabec, J., Bernreuther, M., Grottel, S., Reina, G., Wix, A., Schaber, K., Hasse, H.: Homogeneous nucleation in supersaturated vapors of methane, ethane, and carbon dioxide predicted by brute force molecular dynamics. The Journal of Chemical Physics 128(16), 164510 (2008)CrossRefGoogle Scholar
  9. 9.
    Kabadshow, I., Dachsel, H., Hammond, J.: Poster: Passing the three trillion particle limit with an error-controlled fast multipole method. In: Proceedings of the 2011 Companion on High Performance Computing Networking, Storage and Analysis Companion, SC 2011 Companion, pp. 73–74. ACM, New York (2011)CrossRefGoogle Scholar
  10. 10.
    Kadau, K., Germann, T.C., Lomdahl, P.S.: Molecular dynamics comes of age: 320 billion atom simulation on bluegene/l. International Journal of Modern Physics C 17(12), 1755–1761 (2006)CrossRefGoogle Scholar
  11. 11.
    Lindahl, E., Hess, B., van der Spoel, D.: Gromacs 3.0: a package for molecular simulation and trajectory analysis. Journal of Molecular Modeling 7, 306–317 (2001)Google Scholar
  12. 12.
    Olivier, S., Prins, J., Derby, J., Vu, K.: Porting the gromacs molecular dynamics code to the cell processor. In: IEEE International Parallel and Distributed Processing Symposium, IPDPS 2007, pp. 1–8 (March 2007)Google Scholar
  13. 13.
    Peng, L., Kunaseth, M., Dursun, H., Nomura, K.-i., Wang, W., Kalia, R., Nakano, A., Vashishta, P.: Exploiting hierarchical parallelisms for molecular dynamics simulation on multicore clusters. The Journal of Supercomputing 57, 20–33 (2011)CrossRefGoogle Scholar
  14. 14.
    Piazza, T., Jiang, H., Hammarlund, P., Singhal, R.: Technology Insight: Intel(R) Next Generation Microarchitecture Code Name Haswell (September 2012)Google Scholar
  15. 15.
    Rahimian, A., Lashuk, I., Veerapaneni, S., Chandramowlishwaran, A., Malhotra, D., Moon, L., Sampath, R., Shringarpure, A., Vetter, J., Vuduc, R., Zorin, D., Biros, G.: Petascale direct numerical simulation of blood flow on 200k cores and heterogeneous architectures. In: Proceedings of the 2010 ACM/IEEE International Conference for High Performance Computing, Networking, Storage and Analysis, SC 2010, pp. 1–11. IEEE Computer Society, Washington, DC (2010)CrossRefGoogle Scholar
  16. 16.
    Rapaport, D.C.: The Art of Molecular Dynamics Simulation. Cambridge University Press (2004)Google Scholar
  17. 17.
    Rapaport, D.C.: Multibillion-atom molecular dynamics simulation: Design considerations for vector-parallel processing. Computer Physics Communications 174(7), 521–529 (2006)MathSciNetMATHCrossRefGoogle Scholar
  18. 18.
    Roth, J., Gähler, F., Trebin, H.-R.: A molecular dynamics run with 5 180 116 000 particles. International Journal of Modern Physics C 11(02), 317–322 (2000)CrossRefGoogle Scholar
  19. 19.
    Vrabec, J., Kedia, G.K., Fuchs, G., Hasse, H.: Comprehensive study of the vapour-liquid coexistence of the truncated and shifted lennard-jones fluid including planar and spherical interface properties. Molecular Physics 104(9), 1509–1527 (2006)CrossRefGoogle Scholar

Copyright information

© Springer-Verlag Berlin Heidelberg 2013

Authors and Affiliations

  • Wolfgang Eckhardt
    • 1
  • Alexander Heinecke
    • 1
  • Reinhold Bader
    • 2
  • Matthias Brehm
    • 2
  • Nicolay Hammer
    • 2
  • Herbert Huber
    • 2
  • Hans-Georg Kleinhenz
    • 2
  • Jadran Vrabec
    • 3
  • Hans Hasse
    • 4
  • Martin Horsch
    • 4
  • Martin Bernreuther
    • 5
  • Colin W. Glass
    • 5
  • Christoph Niethammer
    • 5
  • Arndt Bode
    • 1
    • 2
  • Hans-Joachim Bungartz
    • 1
    • 2
  1. 1.Technische Universität MünchenGarchingGermany
  2. 2.Leibniz-Rechenzentrum der Bayerischen Akademie der WissenschaftenGarchingGermany
  3. 3.University of PaderbornPaderbornGermany
  4. 4.Laboratory of Engineering Thermodynamics (LTD)TU KaiserslauternKaiserslauternGermany
  5. 5.High Performance Computing Centre Stuttgart (HLRS)StuttgartGermany

Personalised recommendations