Abstract
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. © 2014 Springer-Verlag.
Author supplied keywords
Cite
CITATION STYLE
Williams, D., Codreanu, V., Yang, P., Liu, B., Dong, F., Yasar, B., … Roerdink, J. B. T. M. (2014). Evaluation of autoparallelization toolkits for commodity GPUs. In Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics) (Vol. 8384 LNCS, pp. 447–457). Springer Verlag. https://doi.org/10.1007/978-3-642-55224-3_42
Register to see more suggestions
Mendeley helps you to discover research relevant for your work.
