High Rayleigh Number Mantle Convection on GPU
Supercomputing Institute Research Report, University of Minnesota
University of Minnesota, Supercomputing Institute Research Report, UMSI 2011/28, 2011
@article{sanchez2011high,
title={High Rayleigh Number Mantle Convection on GPU},
author={Sanchez, D. and Gonzalez, C. and Yuen, D. and Wright, G. and Barnett, G.},
journal={Yuen, D., Wang, J., Johnsson, L., Chi, C., Shi, Y., GPU Solutions to Multi-Scale Problems in Science and Engineering, Springer},
year={2011}
}
We implemented two- and three-dimensional Rayleigh-Benard convection on Nvidia GPUs by utilizing a 2nd-order finite difference method. By exploiting the massive parallelism of GPU using both CUDA for C and optimized CUBLAS routines, we have on a single Fermi GPU run simultaneous of Raileigh number up to 6×10^10 (on a mesh of 2000×4000 uniform grid points) in two dimensions and up to 10^7 (on a mesh of 450x450x225 uniform grid points) for three dimensions. On Nvidia Tesla C2070 GPUs, these implementations enjoy single-precision performance of 535 GFLOP/s and 100 GFLOP/s respectively, and double-precision performance of 230 GFLOP/s and 70 GFLOP/s respectively.
November 25, 2011 by hgpu
