SPEC OpenMP benchmarks on four generations of NEC SX parallel vector systems

Authors:
Matthias S. Müller
Affiliations:
HLRS, University of Stuttgart, Stuttgart, Germany
Venue:
IWOMP'05/IWOMP'06 Proceedings of the 2005 and 2006 international conference on OpenMP shared memory parallel programming
Year:
2005

Citing 4
Cited 0

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
Quantitative performance analysis of the SPEC OMPM2001 benchmarks

Scientific Programming - OpenMP
Large System Performance of SPEC OMP2001 Benchmarks

ISHPC '02 Proceedings of the 4th International Symposium on High Performance Computing
Performance Evaluation of the Hitachi SR8000 Using OpenMP Benchmarks

ISHPC '02 Proceedings of the 4th International Symposium on High Performance Computing

Quantified Score

Hi-index	0.00

Visualization

Abstract

We describe the performance characteristics of the SPEC OMP benchmarks on parallel vector supercomputers. Points of interest are vectorization, scalability and the comparison between different generations of the same family of NEC SX vector supercomputers. We relate the different performance development of the 11 different applications to different hardware properties of the machine and also to results of the EPCC microbenchmarks. Of special interest is the fact the the NEC SX parallel architecture is not cache consistent.