Processing Hard Sphere Collisions on a GPU Using OpenCL
Computer Science, Trinity University, San Antonio, TX, USA
The 2014 International Conference on Parallel and Distributed Processing Techniques and Applications (PDPTA’14), 2014
@article{langbert2014processing,
title={Processing Hard Sphere Collisions on a GPU Using OpenCL},
author={Langbert, Zachary and Lewis, Mark C},
year={2014}
}
Physically accurate hard sphere collisions are inherently sequential as the order in which collisions occur can have a significant impact on the resulting system. This makes processing hard sphere collisions on parallel hardware challenging. We present an approach to solving this problem that can be implemented using OpenCL that runs on current hardware. This approach makes significant use of atomic operations to prevent race conditions, even across thread groups. We find that an unoptimized implementation of the approach provides speed on modest GPUs that is on par with our earlier OpenMP parallel CPU approach and the OpenCL running on a CPU is faster than the OpenMP code. Full timing results using commodity GPU and using OpenCL on multi-core chips are presented.
November 18, 2014 by hgpu
