CCRG OpenMP compiler: experiments and improvements

Authors:
Huang Chun;Yang Xuejun
Affiliations:
National Laboratory for Parallel and Distributed Processing, P.R. China;National Laboratory for Parallel and Distributed Processing, P.R. China
Venue:
IWOMP'05/IWOMP'06 Proceedings of the 2005 and 2006 international conference on OpenMP shared memory parallel programming
Year:
2005

Citing 3
Cited 0

JIAJIA: A Software DSM System Based on a New Cache Coherence Protocol

HPCN Europe '99 Proceedings of the 7th International Conference on High-Performance Computing and Networking
Portable Compilers for OpenMP

WOMPAT '01 Proceedings of the International Workshop on OpenMP Applications and Tools: OpenMP Shared Memory Parallel Programming
SPEComp: A New Benchmark Suite for Measuring Parallel Computer Performance

WOMPAT '01 Proceedings of the International Workshop on OpenMP Applications and Tools: OpenMP Shared Memory Parallel Programming

Quantified Score

Hi-index	0.00

Visualization

Abstract

In this paper, we present the design and experiments of a practical OpenMP compiler for SMP, called CCRG OpenMP Compiler, with the focus on its performance comparison with commercial Intel Fortran Compiler 8.0 using SPEC OMPM2001 benchmarks. The preliminary experiments showed that CCRG OpenMP is a quite robust and efficient compiler for most of the benchmarks except mgrid and wupwise. Then, further performance analysis of mgrid and wupwise are provided through gprof tool and Intel optimization report respectively. Based on the performance analysis, we present the optimized static schedule implementation and inter-procedural constant propagation techniques to improve the performance of CCRG OpenMP Compiler. After optimization, all of the SPEC OMPM2001 Fortran benchmarks can be executed on SMP systems efficiently as expected.