A "Generalized k-Tree-Based Model to Sub-system Allocation" for Partitionable Multi-dimensional Mesh-Connected Architectures

Authors:
Jeeraporn Srisawat;Nikitas A. Alexandridis
Affiliations:
-;-
Venue:
ISHPC '00 Proceedings of the Third International Symposium on High Performance Computing
Year:
2000

Citing 7
Cited 0

Submesh allocation in mesh multicomputers using busy-list: a best-fit approach with complete recognition capability

Journal of Parallel and Distributed Computing
Processor allocation using partitioning in mesh connected parallel computers

Journal of Parallel and Distributed Computing
An Efficient Task Allocation Scheme for 2D Mesh Architectures

IEEE Transactions on Parallel and Distributed Systems
On Submesh Allocation for Mesh Multicomputers: A Best-Fit Allocation and a Virtual Submesh Allocation for Faulty Meshes

IEEE Transactions on Parallel and Distributed Systems
A new “quad-tree-based” sub-system allocation technique for mesh-connected parallel machines

ICS '99 Proceedings of the 13th international conference on Supercomputing
Job Scheduling in a Partitionable Mesh Using a Two-Dimensional Buddy System Partitioning Scheme

IEEE Transactions on Parallel and Distributed Systems
An Adaptive Submesh Allocation Strategy for Two-Dimensional Mesh Connected Systems

ICPP '93 Proceedings of the 1993 International Conference on Parallel Processing - Volume 02

Quantified Score

Hi-index	0.00

Visualization

Abstract

This paper presents a new processor allocation approach called "a generalized k-Tree-based model" to perform dynamic sub-system allocation/ deallocation decision for partitionable multi-dimensional mesh-connected architectures. Time complexity of our generalized k-tree-based sub-system allocation algorithm is O(k42k(NA+NF)+k222k) for the partitionable k-D meshes and O(NA+NF) for the partitionable 2-D meshes, where NA is the maximum number of allocated tasks, NF is the corresponding number of free sub-meshes, N is the system size, and NA+NF ≤ N. Most existing processor allocation strategies have been proposed for the partitionable 2-D meshes with various degrees of time complexity and system performance. In order to evaluate the system performance, the generalized k-Tree-based model was developed and by simulation studies the results of applying our k-Tree-based approach for the partitionable 2-D meshes were presented and compared to existing 2-D mesh-based strategies. Our results showed that the k-Tree-based approach (when it was applied for the partitionable 2-D meshes) yielded the comparable system performance to those recently 2-D mesh-based strategies.