Message proxies for efficient, protected communication on SMP clusters

Beng Hong Lim, Philip Heidelberger, Pratap Pattnaik, Marc Snir

Research output: Contribution to conferencePaper

Abstract

This research addresses the problem of providing efficient, protected communication in an SMP cluster without incurring the overhead of system calls or the cost of custom hardware. It analyzes an approach that uses an idle SMP processor to run a message proxy, a communication process that provides protected access to the network. We implement message proxy based communication between a pair of IBM Model G30 SMPs and analyze the resulting overheads. We derive a performance model that shows that cache-miss latency within an SMP influences message proxy performance significantly. Simulations of a suite of ten parallel applications demonstrate that message proxies match the performance of custom hardware for three of the ten applications, and are between 10-30% slower for the other seven applications. A direct cache-update mechanism to reduce cache misses improves the performance of message proxies on communication-intensive programs by 7-25%. We conclude that message proxies provide a viable alternative to custom hardware for protected communication.

Original languageEnglish (US)
Pages116-127
Number of pages12
StatePublished - Jan 1 1997
Externally publishedYes
EventProceedings of the 1997 3rd International Symposium on High-Performance Computer Architecture, HPCA - San Antonio, TX, USA
Duration: Feb 1 1997Feb 5 1997

Other

OtherProceedings of the 1997 3rd International Symposium on High-Performance Computer Architecture, HPCA
CitySan Antonio, TX, USA
Period2/1/972/5/97

Fingerprint

Communication
Hardware
Costs

ASJC Scopus subject areas

  • Hardware and Architecture

Cite this

Lim, B. H., Heidelberger, P., Pattnaik, P., & Snir, M. (1997). Message proxies for efficient, protected communication on SMP clusters. 116-127. Paper presented at Proceedings of the 1997 3rd International Symposium on High-Performance Computer Architecture, HPCA, San Antonio, TX, USA, .

Message proxies for efficient, protected communication on SMP clusters. / Lim, Beng Hong; Heidelberger, Philip; Pattnaik, Pratap; Snir, Marc.

1997. 116-127 Paper presented at Proceedings of the 1997 3rd International Symposium on High-Performance Computer Architecture, HPCA, San Antonio, TX, USA, .

Research output: Contribution to conferencePaper

Lim, BH, Heidelberger, P, Pattnaik, P & Snir, M 1997, 'Message proxies for efficient, protected communication on SMP clusters' Paper presented at Proceedings of the 1997 3rd International Symposium on High-Performance Computer Architecture, HPCA, San Antonio, TX, USA, 2/1/97 - 2/5/97, pp. 116-127.
Lim BH, Heidelberger P, Pattnaik P, Snir M. Message proxies for efficient, protected communication on SMP clusters. 1997. Paper presented at Proceedings of the 1997 3rd International Symposium on High-Performance Computer Architecture, HPCA, San Antonio, TX, USA, .
Lim, Beng Hong ; Heidelberger, Philip ; Pattnaik, Pratap ; Snir, Marc. / Message proxies for efficient, protected communication on SMP clusters. Paper presented at Proceedings of the 1997 3rd International Symposium on High-Performance Computer Architecture, HPCA, San Antonio, TX, USA, .12 p.
@conference{100f638416204fc48dbf7e3dd652d82b,
title = "Message proxies for efficient, protected communication on SMP clusters",
abstract = "This research addresses the problem of providing efficient, protected communication in an SMP cluster without incurring the overhead of system calls or the cost of custom hardware. It analyzes an approach that uses an idle SMP processor to run a message proxy, a communication process that provides protected access to the network. We implement message proxy based communication between a pair of IBM Model G30 SMPs and analyze the resulting overheads. We derive a performance model that shows that cache-miss latency within an SMP influences message proxy performance significantly. Simulations of a suite of ten parallel applications demonstrate that message proxies match the performance of custom hardware for three of the ten applications, and are between 10-30{\%} slower for the other seven applications. A direct cache-update mechanism to reduce cache misses improves the performance of message proxies on communication-intensive programs by 7-25{\%}. We conclude that message proxies provide a viable alternative to custom hardware for protected communication.",
author = "Lim, {Beng Hong} and Philip Heidelberger and Pratap Pattnaik and Marc Snir",
year = "1997",
month = "1",
day = "1",
language = "English (US)",
pages = "116--127",
note = "Proceedings of the 1997 3rd International Symposium on High-Performance Computer Architecture, HPCA ; Conference date: 01-02-1997 Through 05-02-1997",

}

TY - CONF

T1 - Message proxies for efficient, protected communication on SMP clusters

AU - Lim, Beng Hong

AU - Heidelberger, Philip

AU - Pattnaik, Pratap

AU - Snir, Marc

PY - 1997/1/1

Y1 - 1997/1/1

N2 - This research addresses the problem of providing efficient, protected communication in an SMP cluster without incurring the overhead of system calls or the cost of custom hardware. It analyzes an approach that uses an idle SMP processor to run a message proxy, a communication process that provides protected access to the network. We implement message proxy based communication between a pair of IBM Model G30 SMPs and analyze the resulting overheads. We derive a performance model that shows that cache-miss latency within an SMP influences message proxy performance significantly. Simulations of a suite of ten parallel applications demonstrate that message proxies match the performance of custom hardware for three of the ten applications, and are between 10-30% slower for the other seven applications. A direct cache-update mechanism to reduce cache misses improves the performance of message proxies on communication-intensive programs by 7-25%. We conclude that message proxies provide a viable alternative to custom hardware for protected communication.

AB - This research addresses the problem of providing efficient, protected communication in an SMP cluster without incurring the overhead of system calls or the cost of custom hardware. It analyzes an approach that uses an idle SMP processor to run a message proxy, a communication process that provides protected access to the network. We implement message proxy based communication between a pair of IBM Model G30 SMPs and analyze the resulting overheads. We derive a performance model that shows that cache-miss latency within an SMP influences message proxy performance significantly. Simulations of a suite of ten parallel applications demonstrate that message proxies match the performance of custom hardware for three of the ten applications, and are between 10-30% slower for the other seven applications. A direct cache-update mechanism to reduce cache misses improves the performance of message proxies on communication-intensive programs by 7-25%. We conclude that message proxies provide a viable alternative to custom hardware for protected communication.

UR - http://www.scopus.com/inward/record.url?scp=0030823599&partnerID=8YFLogxK

UR - http://www.scopus.com/inward/citedby.url?scp=0030823599&partnerID=8YFLogxK

M3 - Paper

AN - SCOPUS:0030823599

SP - 116

EP - 127

ER -