Exploring thread and memory placement on NUMA architectures: Solaris and Linux, UltraSPARC/FirePlane and Opteron/HyperTransport

Joseph Antony, Pete P. Janes, Alistair P. Rendell

Research output: Chapter in Book/Report/Conference proceedingConference contribution

26 Citations (Scopus)

Abstract

Modern shared memory multiprocessor systems commonly have non-uniform memory access (NUMA) with asymmetric memory bandwidth and latency characteristics. Operating systems now provide application programmer interfaces allowing the user to perform specific thread and memory placement. To date, however, there have been relatively few detailed assessments of the importance of memory/thread placement for complex applications. This paper outlines a framework for performing memory and thread placement experiments on Solaris and Linux. Thread binding and location specific memory allocation and its verification is discussed and contrasted. Using the framework, the performance characteristics of serial versions of lmbench, Stream and various BLAS libraries (ATLAS, GOTO, ACML on Opteron/Linux and Sunperf on Opteron, UltraSPARC/Solaris) are measured on two different hardware platforms (UltraSPARC/FirePlane and Opteron/HyperTransport). A simple model describing performance as a function of memory distribution is proposed and assessed for both the Opteron and UltraSPARC.

Original languageEnglish
Title of host publicationHigh Performance Computing - HiPC 2006 - 13th International Conference Proceedings
Pages338-352
Number of pages15
Volume4297
ISBN (Electronic)9783540680406
DOIs
Publication statusPublished - Dec 2006
Externally publishedYes
Event13th International Conference on High Performance Computing, HiPC 2006 - Bangalore, India
Duration: 18 Dec 200621 Dec 2006

Publication series

NameLecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)
Volume4297 LNCS
ISSN (Print)0302-9743
ISSN (Electronic)1611-3349

Conference

Conference13th International Conference on High Performance Computing, HiPC 2006
CountryIndia
CityBangalore
Period18/12/0621/12/06

Fingerprint Dive into the research topics of 'Exploring thread and memory placement on NUMA architectures: Solaris and Linux, UltraSPARC/FirePlane and Opteron/HyperTransport'. Together they form a unique fingerprint.

  • Cite this

    Antony, J., Janes, P. P., & Rendell, A. P. (2006). Exploring thread and memory placement on NUMA architectures: Solaris and Linux, UltraSPARC/FirePlane and Opteron/HyperTransport. In High Performance Computing - HiPC 2006 - 13th International Conference Proceedings (Vol. 4297, pp. 338-352). (Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics); Vol. 4297 LNCS). https://doi.org/10.1007/11945918_35