Analysis and optimization of gyrokinetic toroidal simulations on homogenous and heterogenous platforms. (November 2013)

• Record Type: Journal Article
• Authors: Ibrahim, Khaled Z; Madduri, Kamesh; Williams, Samuel; Wang, Bei; Ethier, Stephane; Oliker, Leonid
• Is Part Of: International journal of high performance computing applications. Volume 27:Number 4(2013)
• Page Start: 454
• Page End: 473
• Publication Date: 2013-11
• Subjects: Particle-in-cell -- hybrid programming -- memory-centric multi-core tuning -- multi-core optimization -- GPU programming; High performance computing -- Periodicals; Supercomputers -- Periodicals; 004.1105
• DOI: 10.1177/1094342013492446
• Languages: English
• ISSNs: 1094-3420

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.