Attaining higher performance in collective communication

E. Chan; M. Heimlich; A. Purkayastha; R. van de Geijn

doi:10.1109/CLUSTR.2004.1392650

2004 IEEE International Conference on Cluster Computing (IEEE Cat. No.04EX935)

Attaining higher performance in collective communication

Year: 2004, Pages: 484

DOI Bookmark: 10.1109/CLUSTR.2004.1392650

Authors

E. Chan, Texas Univ., USA
M. Heimlich, Texas Univ., USA
A. Purkayastha, Texas Univ., USA
R. van de Geijn, Texas Univ., USA

Abstract

Summary form only given. It has long been thought that research into collective communication algorithms on distributed-memory parallel computers has been exhausted. This project demonstrates that the implementations available as part of widely-used libraries are suboptimal. We demonstrate this through the implementation of the "reduce-scatter" collective communication and comparison with the MPICH implementation of MPI. Performance on a large cluster is reported.

Like what you’re reading?

Already a member?

Get this article FREE with a new membership!

On optimizing collective communication
2004 IEEE International Conference on Cluster Computing (IEEE Cat. No.04EX935)
Auto-tuning Non-blocking Collective Communication Operations
2015 IEEE International Parallel and Distributed Processing Symposium Workshop (IPDPSW)
Communication Modeling of Heterogeneous Networks of Workstations for Performance Characterization of Collective Operations
Heterogeneous Computing Workshop
Efficient Collective Communication on Heterogeneous Networks of Workstations
Proceedings. 1998 International Conference on Parallel Processing (Cat. No.98EX205)
Exploiting Hierarchy in Parallel Computer Networks to Optimize Collective Operation Performance
Parallel and Distributed Processing Symposium, International
Accelerating MPI Collective Communications through Hierarchical Algorithms Without Sacrificing Inter-Node Communication Flexibility
2014 IEEE International Parallel & Distributed Processing Symposium (IPDPS)
Analysis of Algorithm Selection for Optimizing Collective Communication with MPICH for Ethernet and Myrinet Networks
2007 8th International Conference on Parallel and Distributed Computing, Applications and Technologies, PDCAT 2007
Network Performance Aware MPI Collective Communication Operations in the Cloud
IEEE Transactions on Parallel & Distributed Systems
Framework for Scalable Intra-Node Collective Operations using Shared Memory
2018 SC18: The International Conference for High Performance Computing, Networking, Storage, and Analysis (SC)
Predicting MPI Collective Communication Performance Using Machine Learning
2020 IEEE International Conference on Cluster Computing (CLUSTER)

Attaining higher performance in collective communication

Authors

Abstract

Related Articles