Evaluation of autoparallelization toolkits for commodity graphics hardware
University of Groningen, The Netherlands
10th International Conference on Parallel Processing and Applied Mathematics, 2013
@article{williams2013evaluation,
title={Evaluation of autoparallelization toolkits for commodity graphics hardware},
author={Williams, David and Codreanu, Valeriu and Yang, Po and Liu, Baoquan and Dong, Feng and Yasar, Burhan and Mahdian, Babak and Chiarini, Alessandro and Zhao, Xia and Roerdink, Jos BTM},
year={2013}
}
In this paper we evaluate the performance of the OpenACC and Mint toolkits against C and CUDA implementations of the standard PolyBench test suite. Our analysis reveals that performance is similar in many cases, but that a certain set of code constructs impede the ability of Mint to generate optimal code. We then present some small improvements which we integrate into our own GPSME toolkit (which is derived from Mint) and show that our toolkit now out-performs OpenACC in the majority of tests.
September 24, 2013 by hgpu
