A look at application performance sensitivity to the bandwidth and latency of infiniband networks

Authors:
Darren J. Kerbyson
Affiliations:
Performance and Architecture Lab, Los Alamos National Laboratory, NM
Venue:
IPDPS'06 Proceedings of the 20th international conference on Parallel and distributed processing
Year:
2006

Citing 7
Cited 0

Predictive performance and scalability modeling of a large-scale application

Proceedings of the 2001 ACM/IEEE conference on Supercomputing
Exploring Advanced Architectures Using Performance Prediction

IWIA '02 Proceedings of the International Workshop on Innovative Architecture for Future Generation High-Performance Processors and Systems (IWIA'02)
The Case of the Missing Supercomputer Performance: Achieving Optimal Performance on the 8,192 Processors of ASCI Q

Proceedings of the 2003 ACM/IEEE conference on Supercomputing
Evaluating InfiniBand Performance with PCI Express

IEEE Micro
A performance comparison between the Earth Simulator and other terascale systems on a characteristic ASCI workload: Research Articles

Concurrency and Computation: Practice & Experience - The High Performance Architectural Challenge: Mass Market versus Proprietary Components?
Performance and Scalability Analysis of Teraflop-Scale Parallel Architectures Using Multidimensional Wavefront Applications

International Journal of High Performance Computing Applications
Microbenchmark Performance Comparison of High-Speed Cluster Interconnects

IEEE Micro

Quantified Score

Hi-index	0.00

Visualization

Abstract

This work explores the expected performance of three applications on a High Performance Computing cluster interconnected using Infiniband. In particular, the expected performance across a range of configurations is analyzed notably Infiniband 4x, 8x and 12x representing link-speeds of 10Gb/s, 20Gb/s, and 30Gb/s respectively as well as near-neighbor MPI message latencies of 4µs and 1.5µs. In addition we also consider the impact of node size, from one to eight processors that share a single network connection. The performance analysis is based on the use of detailed performance models of the three applications developed at Los Alamos. The results of the analysis show that the application performance can range by as much as 60% from best to worst. The relative importance of bandwidth, latency and node size differs between the applications.