Roerdink, Jos B.T.M.
AffiliationUniversity of Groningen
University of Bedfordshire
Super Computing Solutions
MetadataShow full item record
Other TitlesParallel Processing and Applied Mathematics: 10th International Conference, PPAM 2013, Warsaw, Poland, September 8-11, 2013, Revised Selected Papers, Part I
AbstractIn 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.
CitationWilliams D, Codreanu V, Yang P, Liu B, Dong F, Yasar B, Mahdian B, Chiarini A, Zhao X, Roerdink J (2014) 'Evaluation of autoparallelization toolkits for commodity GPUs', in Wyrzykowski R, Dongarra J, Karczewski K, Waśniewski J (ed(s).). Parallel Processing and Applied Mathematics: 10th International Conference, PPAM 2013, Warsaw, Poland, September 8-11, 2013, Revised Selected Papers, Part I, Springer Verlag pp.447-457.