The original paper is in English. Non-English content has been machine-translated and may contain typographical errors or mistranslations. ex. Some numerals are expressed as "XNUMX".
Copyrights notice
The original paper is in English. Non-English content has been machine-translated and may contain typographical errors or mistranslations. Copyrights notice
Cet article décrit la conception et la mise en œuvre d'un système de mémoire partagée distribuée (DSM) multithread, appelé Cohesion, qui offre une flexibilité de programmation et un masquage de latence élevés, et prend en charge l'équilibrage de charge. Cohesion offre un environnement de programmation parallèle très similaire à celui d'un système multiprocesseurs. Les threads peuvent être créés de manière récursive dans cet environnement et les utilisateurs ne sont pas tenus de gérer les emplacements des threads. Au lieu de prendre en charge un modèle de variable partagée, Cohesion fournit un espace d'adressage global partagé entre tous les nœuds du système. L'espace est ensuite divisé en trois régions, à savoir la mémoire de version, la mémoire conventionnelle et la mémoire basée sur les objets, chacune étant appliquée avec un protocole de cohérence différent. Dans cet article, les problèmes de conception d'un système de threads ordinaire, tels que la gestion des threads, l'équilibrage de charge et la synchronisation, ont été reconsidérés avec la gestion de la mémoire fournie par le système DSM. Plusieurs applications réelles ont été utilisées pour évaluer les performances du système. Les résultats montrent que le multithreading offre généralement de meilleures performances que le monothreading, car la latence du réseau peut être masquée par le chevauchement des communications et des calculs. Cependant, le gain dépend du comportement du programme et du nombre de threads exécutés sur chaque nœud du système.
The copyright of the original papers published on this site belongs to IEICE. Unauthorized use of the original or translated papers is prohibited. See IEICE Provisions on Copyright for details.
Copier
Jyh-Chang UENG, Ce-Kuen SHIEH, Su-Cheong MAC, An-Chow LAI, Tyng-Yue LIANG, "Multi-Threaded Design for a Software Distributed Shared Memory Systems" in IEICE TRANSACTIONS on Information,
vol. E82-D, no. 12, pp. 1512-1523, December 1999, doi: .
Abstract: This paper describes the design and implementation of a multi-threaded Distributed Shared Memory (DSM) system, called Cohesion, which provides high programming flexibility and latency masking, and supports load balancing. Cohesion offers a parallel programming environment which is very similar to that on a multiprocessors system. Threads could be created recursively in this environment, and users are not required to handle the locations of the threads. Instead of supporting a shared variable model, Cohesion provides a global shared address space among all nodes in the system. The space is further divided into three regions, i. e. , release, conventional, and object-based memory, each is applied with different consistency protocol. In this paper, the design issues in an ordinary thread system, such as thread management, load balancing, and synchronization, have been reconsidered with the memory management provided by the DSM system. Several real applications have been used to evaluate the performance of the system. The results show that multi-threading usually has better performance than single-threading because the network latency can be masked by overlapping communication and computation. However, the gain depends on program behavior and the number of threads executed on each node in the system.
URL: https://global.ieice.org/en_transactions/information/10.1587/e82-d_12_1512/_p
Copier
@ARTICLE{e82-d_12_1512,
author={Jyh-Chang UENG, Ce-Kuen SHIEH, Su-Cheong MAC, An-Chow LAI, Tyng-Yue LIANG, },
journal={IEICE TRANSACTIONS on Information},
title={Multi-Threaded Design for a Software Distributed Shared Memory Systems},
year={1999},
volume={E82-D},
number={12},
pages={1512-1523},
abstract={This paper describes the design and implementation of a multi-threaded Distributed Shared Memory (DSM) system, called Cohesion, which provides high programming flexibility and latency masking, and supports load balancing. Cohesion offers a parallel programming environment which is very similar to that on a multiprocessors system. Threads could be created recursively in this environment, and users are not required to handle the locations of the threads. Instead of supporting a shared variable model, Cohesion provides a global shared address space among all nodes in the system. The space is further divided into three regions, i. e. , release, conventional, and object-based memory, each is applied with different consistency protocol. In this paper, the design issues in an ordinary thread system, such as thread management, load balancing, and synchronization, have been reconsidered with the memory management provided by the DSM system. Several real applications have been used to evaluate the performance of the system. The results show that multi-threading usually has better performance than single-threading because the network latency can be masked by overlapping communication and computation. However, the gain depends on program behavior and the number of threads executed on each node in the system.},
keywords={},
doi={},
ISSN={},
month={December},}
Copier
TY - JOUR
TI - Multi-Threaded Design for a Software Distributed Shared Memory Systems
T2 - IEICE TRANSACTIONS on Information
SP - 1512
EP - 1523
AU - Jyh-Chang UENG
AU - Ce-Kuen SHIEH
AU - Su-Cheong MAC
AU - An-Chow LAI
AU - Tyng-Yue LIANG
PY - 1999
DO -
JO - IEICE TRANSACTIONS on Information
SN -
VL - E82-D
IS - 12
JA - IEICE TRANSACTIONS on Information
Y1 - December 1999
AB - This paper describes the design and implementation of a multi-threaded Distributed Shared Memory (DSM) system, called Cohesion, which provides high programming flexibility and latency masking, and supports load balancing. Cohesion offers a parallel programming environment which is very similar to that on a multiprocessors system. Threads could be created recursively in this environment, and users are not required to handle the locations of the threads. Instead of supporting a shared variable model, Cohesion provides a global shared address space among all nodes in the system. The space is further divided into three regions, i. e. , release, conventional, and object-based memory, each is applied with different consistency protocol. In this paper, the design issues in an ordinary thread system, such as thread management, load balancing, and synchronization, have been reconsidered with the memory management provided by the DSM system. Several real applications have been used to evaluate the performance of the system. The results show that multi-threading usually has better performance than single-threading because the network latency can be masked by overlapping communication and computation. However, the gain depends on program behavior and the number of threads executed on each node in the system.
ER -