Implementation and Evaluation of Recurrence Equation Solvers on GPGPU systems using Rearrangement of Array Configurations
Faculty of Intelligence and Informatics, Konan University
The Fourth International Conference on Future Computational Technologies and Applications (FUTURE COMPUTING), 2012
@inproceedings{wakatani2012implementation,
title={Implementation and Evaluation of Recurrence Equation Solvers on GPGPU Systems Using Rearrangement of Array Configurations},
author={Wakatani, A.},
booktitle={FUTURE COMPUTING 2012, The Fourth International Conference on Future Computational Technologies and Applications},
pages={32–38},
year={2012}
}
The recurrence equation solver is used in many numerical applications and other general-purpose applications, but it is inherently a sequential algorithm, so it is difficult to implement the parallel program for it. Recently, GPGPU (General Purpose computing on Graphic Processing Unit) attracts a great deal of attention, which is used for generalpurpose computations like numerical calculations as well as graphic processing. In this paper, we implement a parallel and scalable algorithm for solving recurrence equations on GPUs by using CUDA (Compute Unified Device Architecture) and evaluate its effectiveness. The algorithm was originally implemented for MIMD parallel computers like a PC cluster and an SMP system by the authors and we modify the algorithm suitable for the GPGPU system by rearranging arrays configurations.
July 29, 2012 by hgpu
