Vortex Methods for Massively Parallel Computer Architectures

Chatelain, Philippe;Curioni, Alessandro;Bergdorf, Michael;Rossinelli, Diego;Koumoutsakos, Petros;et.al.
(2008) High Performance Computing for Computational Science - VECPAR 2008 — ISBN: [978-3-540-92858-4], p. 479-489, published

Files

pdfdocument.pdf
  • Restricted Access
  • Adobe PDF
  • 6.39 MB
Request a copy

Details

Authors
  • Chatelain, Philippeorcid-logoUCLouvain
    Author
  • Curioni, AlessandroIBM Research Division - Zurich Resaerch Laboratory, Rueschlikon, Switzerland
    Author
  • Bergdorf, MichaelIBM Research Division - Zurich Resaerch Laboratory, Rueschlikon, Switzerland
    Author
  • Rossinelli, DiegoIBM Research Division - Zurich Resaerch Laboratory, Rueschlikon, Switzerland
    Author
  • Koumoutsakos, PetrosETH Zurich
    Author
Show more
Abstract
We present Vortex Methods implemented in massively parallel computer architectures for the Direct Numerical Simulations of high Reynolds numbers flows. Periodic and non-periodic domains are considered leading to unprecedented simulations using billions of particles. We discuss the implementation performance of the method up to 16<em>k</em> IBM BG/L nodes and the evolutionary optimization of long wavelength instabilities in aircraft wakes.
Affiliations

Citations

  • APA
  • Chicago
  • FWB

Chatelain, P., Curioni, A., Bergdorf, M., Rossinelli, D., Andreoni, W., & Koumoutsakos, P. (2008). Vortex Methods for Massively Parallel Computer Architectures. In Palma, J.; Amestoy, P.; Daydé, M.; Mattoso, M. & Lopes, J. (ed.), High Performance Computing for Computational Science - VECPAR 2008 (p. p. 479-489). Springer. https://doi.org/10.1007/978-3-540-92859-1_42