Computing Prestack Kirchhoff Time Migration on General Purpose GPU
School of Computer Science and Engineering, Beihang University, Xueyuan Road, No. 37, Haidian District, Beijing 100083, China
Computers & Geosciences (10 November 2010)
@article{shi2010computing,
title={Computing Prestack Kirchhoff Time Migration on General Purpose GPU},
author={Shi, X. and Li, C. and Wang, S. and Wang, X.},
journal={Computers & Geosciences},
issn={0098-3004},
year={2010},
publisher={Elsevier}
}
This paper introduces how to optimize a practical Prestack Kirchhoff Time Migration program by Compute Unified Device Architecture (CUDA) on General Purpose GPU (GPGPU). A few useful optimization methods on GPGPU are demostrated, such as how to increase the kernel thread numbers on GPU cores, and how to utilize the memory streams to overlap GPU kernel execution time, etc. The floating-point errors on CUDA and NVidia’s GPUs are discussed in detail. Some effective methods that can be used to reduce the floating-point errors are introduced. The images generated by the practical Prestack Kirchhoff Time Migration programs for the same real-world seismic data inputs on CPU and GPU are demonstrated. The final GPGPU approach on NVidia GTX 260 is more than 17 times faster than its original CPU version on Intel’s P4 3.0G.
November 26, 2010 by hgpu
