Extended overhead analysis for OpenMP performance tuning

Authors:
Chen Yongjian;Wang Dingxing;Zheng Weimin
Affiliations:
Institute of HPC, Tsinghua University, China;Institute of HPC, Tsinghua University, China;Institute of HPC, Tsinghua University, China
Venue:
WOMPAT'03 Proceedings of the OpenMP applications and tools 2003 international conference on OpenMP shared memory parallel programming
Year:
2003

Citing 7
Cited 0

ATExpert

Journal of Parallel and Distributed Computing - Special issue on tools and methods for visualization of parallel systems and computations
Finding bottlenecks in large-scale parallel programs

Finding bottlenecks in large-scale parallel programs
Performance improvement through overhead analysis: a case study in molecular dynamics

ICS '97 Proceedings of the 11th international conference on Supercomputing
Parallel performance prediction using lost cycles analysis

Proceedings of the 1994 ACM/IEEE conference on Supercomputing
Extended Overhead Analysis for OpenMP (Research Note)

Euro-Par '02 Proceedings of the 8th International Euro-Par Conference on Parallel Processing
SCALEA: A Performance Analysis Tool for Distributed and Parallel Programs

Euro-Par '02 Proceedings of the 8th International Euro-Par Conference on Parallel Processing
A hierarchical classification of overheads in parallel programs

Proceedings of the First IFIP TC10 International Workshop on Software Engineering for Parallel and Distributed Systems

Quantified Score

Hi-index	0.00

Visualization

Abstract

Overhead analysis was developed as a performance tuning approach for parallel programming and were adopted by several performance analysis systems for OpenMP programs. In this paper, an extended overhead analysis scheme based on layered model is proposed for OpenMP programming, to further enhance the capability of overhead analysis and thus make the OpenMP performance tuning easier. An example case called ILP/TLP overlap is studied in detail to show the idea of layered overhead model, and a new way to organize the overhead hierarchically is also presented based on the layered overhead model.