Performance models for the processor farm paradigm

In this paper, we describe the design, implementation, and modeling of a runtime kernel to support the processor farm paradigm on multicomputers. We present a general topology-independent framework for obtaining performance models to predict the performance of the start-up, steady-state, and wind-do...

Ausführliche Beschreibung

Gespeichert in:
Bibliographische Detailangaben
Veröffentlicht in:IEEE transactions on parallel and distributed systems Jg. 8; H. 5; S. 475 - 489
Hauptverfasser: Wagner, A.S., Sreekantaswamy, H.V., Chanson, S.T.
Format: Journal Article
Sprache:Englisch
Veröffentlicht: IEEE 01.05.1997
Schlagworte:
Computer Society
Concurrent computing
Kernel
Multiprocessor interconnection networks
Network topology
Parallel programming
Predictive models
Runtime
Software performance
Steady-state
ISSN:1045-9219
Online-Zugang:Volltext
Tags: Tag hinzufügen
Keine Tags, Fügen Sie den ersten Tag hinzu!
Beschreibung
Zusammenfassung:In this paper, we describe the design, implementation, and modeling of a runtime kernel to support the processor farm paradigm on multicomputers. We present a general topology-independent framework for obtaining performance models to predict the performance of the start-up, steady-state, and wind-down phases of a processor farm. An algorithm is described, which for any interconnection network determines a tree-structured subnetwork that optimizes farm performance. The analysis technique is applied to the important case of k-ary tree topologies. The models are compared with the measured performance on a variety of topologies using both constant and varied task sizes.
Bibliographie:ObjectType-Article-2
SourceType-Scholarly Journals-1
ObjectType-Feature-1
content type line 23
ISSN:1045-9219
DOI:10.1109/71.598275