Title: Investigations on InfiniBand: Efficient Network Buffer Utilization at Scale
Author(s): Galen M. Shipman, Ron Brightwell, Brian Barrett, Jeffrey
M. Squyres, Gil Bloch
Abstract:
The default messaging model for the OpenFabrics "Verbs"
API is to consume receive buffers in order -- regardless of the actual
incoming message size -- leading to inefficient registered memory
usage. For example, many small messages can consume large amounts of
registered memory. This paper introduces a new transport protocol in
Open MPI implemented using the existing OpenFabrics Verbs API that
exhibits efficient registered memory utilization. Several real-world
applications were run at scale with the new protocol; results show
that global network resource utilization efficiency increases,
allowing increased scalability -- and larger problem sizes -- on
clusters which can increase application performance in some cases.
Presented: EuroPVM/MPI '07, September 30th - October 3rd, 2007, Paris,
France.
Paper:
Bibtex reference:
@InProceedings{shipman07:_inves_infin,
author = {Galen M. Shipman and Ron Brightwell and Brian Barrett and Jeffrey M. Squyres and Gil Bloch},
title = {Investigations on InfiniBand: Efficient Network Buffer Utilization at Scale},
booktitle = {Proceedings, Euro PVM/MPI},
year = 2007,
address = {Paris, France},
month = {October}
}
|
