A dynamic processor management strategy on the reconfigurable meshes
by Kyung-Hee Seo, Sung-Chun Kim
International Journal of High Performance Computing and Networking (IJHPCN), Vol. 3, No. 1, 2005

Abstract: A reconfigurable mesh consists of a mesh connected topology augmented by a dynamically reconfigurable bus system. The reconfigurable mesh combines two attractive features of massively parallel system, namely constant diameter and the flexibility of the bus system. Efficient utilisation of processing resources in a large, multi user parallel system depends on the reliable processor management scheme. This paper presents a dynamic and reliable processor allocation strategy to increase the performance of mesh connected parallel systems with faulty processors. The basic idea is to reconfigure a faulty mesh system into a maximum convex system using the fault free upper or lower boundary nodes to compensate for the nonboundary faulty nodes. To utilise the nonrectangular shaped system parts, our strategy tries to allocate L shaped submeshes instead of signalling the allocation failure. Extensive simulations show that the strategy performs more efficiently than other strategies in terms of the completion time, the job response time and the system utilisation.

Online publication date: Wed, 28-Sep-2005

The full text of this article is only available to individual subscribers or to users at subscribing institutions.

 
Existing subscribers:
Go to Inderscience Online Journals to access the Full Text of this article.

Pay per view:
If you are not a subscriber and you just want to read the full contents of this article, buy online access here.

Complimentary Subscribers, Editors or Members of the Editorial Board of the International Journal of High Performance Computing and Networking (IJHPCN):
Login with your Inderscience username and password:

Username:        Password:          Forgotten your password?

Want to subscribe?
A subscription gives you complete access to all articles in the current issue, as well as to all articles in the previous three years (where applicable). See our Orders page to subscribe.

If you still need assistance, please email subs@inderscience.com