Vortex Methods for Massively Parallel Computer Architectures

Authors:
Philippe Chatelain;Alessandro Curioni;Michael Bergdorf;Diego Rossinelli;Wanda Andreoni;Petros Koumoutsakos
Affiliations:
Computational Science and Engineering Laboratory, ETH Zurich, Switzerland CH-8092;Computational Sciences, IBM Research Division - Zurich Research Laboratory, Rueschlikon, Switzerland CH-8003;Computational Science and Engineering Laboratory, ETH Zurich, Switzerland CH-8092;Computational Science and Engineering Laboratory, ETH Zurich, Switzerland CH-8092;Computational Sciences, IBM Research Division - Zurich Research Laboratory, Rueschlikon, Switzerland CH-8003;Computational Science and Engineering Laboratory, ETH Zurich, Switzerland CH-8092
Venue:
High Performance Computing for Computational Science - VECPAR 2008
Year:
2008

Citing 7
Cited 1

Computer simulation using particles

Computer simulation using particles
Artificial viscosity models for vortex and particle methods

Journal of Computational Physics
Inviscid axisymmetrization of an elliptical vortex

Journal of Computational Physics
A vortex particle method for two-dimensional compressible flow

Journal of Computational Physics
Remeshed smoothed particle hydrodynamics for the simulation of viscous and heat conducting flows

Journal of Computational Physics
Reducing the time complexity of the derandomized evolution strategy with covariance matrix adaptation (CMA-ES)

Evolutionary Computation
PPM: a highly efficient parallel particle-mesh library for the simulation of continuum systems

Journal of Computational Physics

Abstractions and Middleware for Petascale Computing and Beyond

International Journal of Distributed Systems and Technologies

Quantified Score

Hi-index	0.00

Visualization

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 16k IBM BG/L nodes and the evolutionary optimization of long wavelength instabilities in aircraft wakes.