Browsing by Author "Hsieh, C. -W."

Now showing 1 - 2 of 2

Large-scale simulations on multiple Graphics Processing Units (GPUs) for the direct simulation Monte Carlo method
(2012-10-01)
Su, C. -C.
;
Smith, M. R.
;
Kuo, F. -A.
;
Wu, J. -S.
;
Hsieh, C. -W.
In this study, the application of the two-dimensional direct simulation Monte Carlo (DSMC) method using an MPI-CUDA parallelization paradigm on Graphics Processing Units (GPUs) clusters is presented. An all-device (i.e. GPU) computational approach is adopted where the entire computation is performed on the GPU device, leaving the CPU idle during all stages of the computation, including particle moving, indexing, particle collisions and state sampling. Communication between the GPU and host is only performed to enable multiple-GPU computation. Results show that the computational expense can be reduced by 15 and 185 times when using a single GPU and 16 GPUs respectively when compared to a single core of an Intel Xeon X5670 CPU. The demonstrated parallel efficiency is 75% when using 16 GPUs as compared to a single GPU for simulations using 30 million simulated particles. Finally, several very large-scale simulations in the near-continuum regime are employed to demonstrate the excellent capability of the current parallel DSMC method. (C) 2012 Elsevier Inc. All rights reserved.
Parallel Direct Simulation Monte Carlo Computation Using CUDA on GPUs
(2011)
Su, C. -C.
;
Hsieh, C. -W.
;
Smith, M. R.
;
Jermy, M. C.
;
Wu, J. -S.
In this study computations of the two-dimensional Direct Simulation Monte Carlo (DSMC) method using Graphics Processing Units (GPUs) are presented. An all-device (GPU) computational approach is adopted-where the entire computation is performed on the GPU device, leaving the CPU idle-which includes particle moving, indexing, collisions between particles and state sampling. The subsequent application to GPU computation requires various changes to the original DSMC method to ensure efficient performance on the GPU device. Communications between the host (CPU) and device (GPU) occur only during problem initialization and simulation conclusion when results are only copied from the device to the host. Several multi-dimensional benchmark tests are employed to demonstrate the correctness of the DSMC implementation. We demonstrate here the application of DSMC using a single-GPU, with speedups of 3 similar to 10 times as compared to a high-end Intel CPU (Intel Xeon X5472) depending upon the size and the level of rarefaction encountered in the simulation.