|MOPWO067||Beam Dynamics Simulations with a GPU-accelerated Version of Elegant||1040|
Funding: Work supported by the US DOE Office of Science, Office of Basic Energy Sciences under grant number DE-SC0004585, and by Tech-X Corporation
Large scale particle tracking and tracking-based lattice optimization simulations can derive significant benefit from efficient implementation of general-purpose particle tracking on GPUs. We present the latest results of our work on accelerating Argonne National Lab's accelerator simulation code ELEGANT*,** using CUDA-enabled GPUs. A sufficiently large number of Elegant beamline elements has been ported to GPUs to allow the GPU-accelerated simulation of realistic test lattices. We will identify some of performance-limiting factors, and briefly discuss optimization techniques for efficient utilization of the device memory space, with an emphasis on register usage. We also present a novel hardware-assisted technique for efficiently calculating a histogram from a large distribution of particle coordinates, and compare this to data-parallel implementations.
* M. Borland, Elegant: A Flexible SDDS-compliant Code for Accelerator Simulation, APS LS-287, September 2000
** Y. Wang, M. Borland, in Proc. of PAC07, THPAN095 (2007)