ARS: an adaptive runtime system for locality optimization

Shared memory programs running on Non-Uniform Memory Access (NUMA) machines usually face inherent performance problems stemming from excessive remote memory accesses. A solution, called the Adaptive Runtime System (ARS), is presented in this paper. ARS is designed to adjust the data distribution at...

Full description

Saved in:
Bibliographic Details
Published in:Future generation computer systems Vol. 19; no. 5; pp. 761 - 776
Main Authors: Tao, Jie, Schulz, Martin, Karl, Wolfgang
Format: Journal Article
Language:English
Published: Elsevier B.V 01.07.2003
Subjects:
Data locality optimization
Hardware monitor
Page migration
Shared memory programming on NUMA
Hardware monitor
Page migration
Shared memory programming on NUMA
Data locality optimization
ISSN:0167-739X, 1872-7115
Online Access:Get full text
Tags: Add Tag
No Tags, Be the first to tag this record!
Description
Summary:Shared memory programs running on Non-Uniform Memory Access (NUMA) machines usually face inherent performance problems stemming from excessive remote memory accesses. A solution, called the Adaptive Runtime System (ARS), is presented in this paper. ARS is designed to adjust the data distribution at runtime through automatic page migrations. It uses memory access histograms gathered by hardware monitors to find access hot spots and, based on this detection, to dynamically and transparently modify the data layout. In this way, incorrectly allocated data can be moved to the most appropriate node and hence data locality can be improved. Simulations show that this allows to achieve a performance gain of as high as 40%.
Bibliography:ObjectType-Article-2
SourceType-Scholarly Journals-1
ObjectType-Feature-1
content type line 23
ISSN:0167-739X
1872-7115
DOI:10.1016/S0167-739X(02)00183-8