12,000

We have over 12,000 students, from over 100 countries, within one of the safest campuses in the UK

93%

93% of Lancaster students go into work or further study within six months of graduating

Home > Research > Publications & Outputs > Portable mapping of data parallel programs to O...
View graph of relations

« Back

Portable mapping of data parallel programs to OpenCL for heterogeneous systems

Research output: Contribution in Book/Report/ProceedingsPaper

Published

Publication date2013
Host publicationCode Generation and Optimization (CGO), 2013 IEEE/ACM International Symposium on
PublisherIEEE
Pages1-10
Number of pages10
ISBN (Print)978-1-4673-5524-7
Original languageEnglish

Abstract

General purpose GPU based systems are highly attractive as they give potentially massive performance at little cost. Realizing such potential is challenging due to the complexity of programming. This paper presents a compiler based approach to automatically generate optimized OpenCL code from data-parallel OpenMP programs for GPUs. Such an approach brings together the benefits of a clear high level-language (OpenMP) and an emerging standard (OpenCL) for heterogeneous multi-cores. A key feature of our scheme is that it leverages existing transformations, especially data transformations, to improve performance on GPU architectures and uses predictive modeling to automatically determine if it is worthwhile running the OpenCL code on the GPU or OpenMP code on the multi-core host. We applied our approach to the entire NAS parallel benchmark suite and evaluated it on two distinct GPU based systems: Core i7/NVIDIA GeForce GTX 580 and Core 17/AMD Radeon 7970. We achieved average (up to) speedups of 4.51× and 4.20× (143× and 67×) respectively over a sequential baseline. This is, on average, a factor 1.63 and 1.56 times faster than a hand-coded, GPU-specific OpenCL implementation developed by independent expert programmers.