Computational Diagnostics for Extreme Scale Toroidal Gyrokinetic Particle Simulations

POSTER

Abstract

The capability of monitoring processes in extreme scale simulation is crucial for extracting information about global and non-linear dynamics, as well as checking the integrity of the simulation. A set of post-processing computational diagnostics for toroidal gyrokinetic particle simulations is developed and optimized for efficient performance on multi- and many-core modern computational platforms. These diagnostics track the time evolution of parallel mode structure, radial profile, toroidal/poloidal spectra, and nonlinear energy transfer spectra. To demonstrate the performance of this diagnostic tool set, diagnosis results for drift wave turbulence with numerical dissipation are presented.

Authors

  • Yuan Shi

    • Princeton Plasma Physics Laboratory
  • Bei Wang

    • Princeton Institute of Computational Science and Engineering
  • Bruce Scott

    • Max-Planck-Institut f\"ur Plasmaphysik
  • William Tang

    • Princeton Plasma Physics Laboratory