Initial study of multi-endpoint runtime for MPI+OpenMP hybrid programming model on multi-core systems

Authors:
Miao Luo;Xiaoyi Lu;Khaled Hamidouche;Krishna Kandalla;Dhabaleswar K. Panda
Affiliations:
The Ohio State University, Columbus, OH, USA;The Ohio State University, Columbus, OH, USA;The Ohio State University, Columbus, OH, USA;The Ohio State University, Columbus, OH, USA;The Ohio State University, Columbus, OH, USA
Venue:
Proceedings of the 19th ACM SIGPLAN symposium on Principles and practice of parallel programming
Year:
2014

Citing 4
Cited 0

Parallel programming with message passing and directives

Computing in Science and Engineering
Dynamic Load Balancing for Structured Adaptive Mesh Refinement Applications

ICPP '02 Proceedings of the 2001 International Conference on Parallel Processing
Hybrid MPI/OpenMP Parallel Programming on Clusters of Multi-Core SMP Nodes

PDP '09 Proceedings of the 2009 17th Euromicro International Conference on Parallel, Distributed and Network-based Processing
Fine-Grained Multithreading Support for Hybrid Threaded MPI Programming

International Journal of High Performance Computing Applications

Quantified Score

Hi-index	0.00

Visualization

Abstract

State-of-the-art MPI libraries rely on locks to guarantee thread-safety. This discourages application developers from using multiple threads to perform MPI operations. In this paper, we propose a high performance, lock-free multi-endpoint MPI runtime, which can achieve up to 40\% improvement for point-to-point operation and one representative collective operation with minimum or no modifications to the existing applications.