IPAC2015 - List of Authors (Amyx, K.M.)

Paper	Title	Page
MOPMA035	Current Status of the GPU-accelerated ELEGANT	623
	I.V. Pogorelov, J.R. King Tech-X, Boulder, Colorado, USA K.M. Amyx Sierra Nevada Corporation, Centennial, USA M. Borland, R. Soliday ANL, Argonne, Illinois, USA
	Funding: Work supported by the DOE Office of Science, Office of BES grant No. DE-SC0004585, and by Tech-X Corporation. This research used resources of the OLCF, supported under Contract DE-AC05-00OR22725. Efficient implementation of general-purpose particle tracking on GPUs can bring significant performance benefits to large-scale tracking simulations and direct (tracking-based) accelerator optimization techniques. This presentation is an update on the current status of our work on accelerating Argonne National Lab’s particle accelerator simulation code ELEGANT [] using CUDA-enabled GPU. We summarize the performance of beamline elements ported to GPU, and discuss optimization techniques for some important collective effects kernels. We also outline briefly our testing and code validation infrastructure within ELEGANT and present the latest results of scaling studies with realistic lattices of the GPU-accelerated version of the code. M. Borland, ‘‘elegant: A Flexible SDDS-compliant Code for Accelerator Simulation", APS LS-287, September 2000
DOI •	reference for this paper ※ https://doi.org/10.18429/JACoW-IPAC2015-MOPMA035
Export •	reference for this paper using ※ BibTeX, ※ LaTeX, ※ Text/Word, ※ RIS, ※ EndNote (xml)

Paper

Title

Page

Current Status of the GPU-accelerated ELEGANT

623

I.V. Pogorelov, J.R. King
Tech-X, Boulder, Colorado, USA
K.M. Amyx
Sierra Nevada Corporation, Centennial, USA
M. Borland, R. Soliday
ANL, Argonne, Illinois, USA

Funding: Work supported by the DOE Office of Science, Office of BES grant No. DE-SC0004585, and by Tech-X Corporation. This research used resources of the OLCF, supported under Contract DE-AC05-00OR22725.
Efficient implementation of general-purpose particle tracking on GPUs can bring significant performance benefits to large-scale tracking simulations and direct (tracking-based) accelerator optimization techniques. This presentation is an update on the current status of our work on accelerating Argonne National Lab’s particle accelerator simulation code ELEGANT [*] using CUDA-enabled GPU. We summarize the performance of beamline elements ported to GPU, and discuss optimization techniques for some important collective effects kernels. We also outline briefly our testing and code validation infrastructure within ELEGANT and present the latest results of scaling studies with realistic lattices of the GPU-accelerated version of the code.
* M. Borland, ‘‘elegant: A Flexible SDDS-compliant Code for Accelerator Simulation", APS LS-287, September 2000

DOI •

reference for this paper ※ https://doi.org/10.18429/JACoW-IPAC2015-MOPMA035

Export •

reference for this paper using ※ BibTeX, ※ LaTeX, ※ Text/Word, ※ RIS, ※ EndNote (xml)