Averages, distributions and scalability of MPI communication times for Ethernet and Myrinet networks

Authors:
Nor Asilah Wati Abdul Hamid;Paul Coddington
Affiliations:
School of Computer Science, University of Adelaide, Adelaide, SA, Australia;School of Computer Science, University of Adelaide, Adelaide, SA, Australia
Venue:
PDCN'07 Proceedings of the 25th conference on Proceedings of the 25th IASTED International Multi-Conference: parallel and distributed computing and networks
Year:
2007

Citing 6
Cited 1

Performance Evaluation of the Quadrics Interconnection Network

Cluster Computing
Performance Evaluation of Fast Ethernet, Giganet, and Myrinet on a Cluster

ICCS '02 Proceedings of the International Conference on Computational Science-Part I
Communication performance issues for two cluster computers

ACSC '03 Proceedings of the 26th Australasian computer science conference - Volume 16
Performance and scalability of MPI on PC clusters: Performances

Concurrency and Computation: Practice & Experience
Comparing Ethernet and Myrinet for MPI communication

LCR '04 Proceedings of the 7th workshop on Workshop on languages, compilers, and run-time support for scalable systems
Analytical models of probability distributions for MPI point-to-point communication times on distributed memory parallel computers

ICA3PP'05 Proceedings of the 6th international conference on Algorithms and Architectures for Parallel Processing

Packet prediction for speculative cut-through switching

Proceedings of the 4th ACM/IEEE Symposium on Architectures for Networking and Communications Systems

Quantified Score

Hi-index	0.00

Visualization

Abstract

Most modern parallel computers are clusters using Myrinet or Ethernet communication networks. Several studies have been published comparing the performance of these two networks for parallel computing, however these focus on average performance, and do not address the distributions of communication times, which can have long tails due to contention effects. In the case of Ethernet with TCP, retransmit timeouts (RTOs) can also occur. Slow communication events may have significant impact, particularly for applications requiring frequent synchronization, where the performance is determined by the slowest process. We have analysed the distributions of communication times for standard MPI routines on Ethernet with TCP and Myrinet with GM communications networks on the same cluster, and studied the scalability of the distributions as the number of communicating processes is increased, and the effect of RTOs for Ethernet with TCP.