Lawrence Berkeley National Laboratory Recent Work Title Analysis and optimization of gyrokinetic toroidal simulations on homogenous and heterogenous platforms Permalink https://escholarship.org/uc/item/48w9z3w5 Journal International Journal of High Performance Computing Applications, 27(4) ISSN 1094-3420 Authors Ibrahim, KZ Madduri, K Williams, S et al. Publication Date 2013-11-01 DOI 10.1177/1094342013492446 Peer reviewed eScholarship.org Powered by the California Digital Library University of California Analysis and Optimization of Gyrokinetic Toroidal Simulations on Homogenous and Heterogenous Platforms Khaled Z. Ibrahim1, Kamesh Madduri2, Samuel Williams1, Bei Wang3 , Stephane Ethier4, Leonid Oliker1 1CRD, Lawrence Berkeley National Laboratory, Berkeley, USA 2 The Pennsylvania State University, University Park, PA, USA 3 Princeton Institute of Computational Science and Engineering, Princeton University, Princeton, NJ, USA 4 Princeton Plasma Physics Laboratory, Princeton, NJ, USA f KZIbrahim, SWWilliams, LOliker
[email protected],
[email protected],
[email protected],
[email protected] Abstract—The Gyrokinetic Toroidal Code (GTC) uses the particle-in-cell method to efficiently simulate plasma microturbulence. This work presents novel analysis and optimization techniques to enhance the performance of GTC on large-scale machines. We introduce cell access analysis to better manage locality vs. synchronization tradeoffs on CPU and GPU-based architectures. Our optimized hybrid parallel implementation of GTC uses MPI, OpenMP, and nVidia CUDA, achieves up to a 2× speedup over the reference Fortran version on multiple parallel systems, and scales efficiently to tens of thousands of cores. 1 Introduction Continuing the scaling and increasing the efficiency of high performance machines are exacerbating the complexity and diversity of computing node architectures.