I/O Placement Strategies and their Data Locality Sensitivities in kxk Toroidal Networks

Authors:
Bader Almohammad;Bella Bose;Jong-Hyun Kim
Affiliations:
-;-;-
Venue:
HPC-ASIA '97 Proceedings of the High-Performance Computing on the Information Superhighway, HPC-Asia '97
Year:
1997

Citing 5
Cited 1

Performance evaluation of a parallel I/O subsystem for hypercube multicomputers

Journal of Parallel and Distributed Computing - Special issue on parallel I/O systems
Lee Distance and Topological Properties of k-ary n-cubes

IEEE Transactions on Computers
Design, Analysis, and Simulation of I/O Architectures for Hypercube Multiprocessors

IEEE Transactions on Parallel and Distributed Systems
Resource Placement in Torus-Based Networks

IPPS '96 Proceedings of the 10th International Parallel Processing Symposium
PROTEUS: A HIGH-PERFORMANCE PARALLEL-ARCHITECTURE SIMULATOR

PROTEUS: A HIGH-PERFORMANCE PARALLEL-ARCHITECTURE SIMULATOR

On resource placements in 3D tori

Journal of Parallel and Distributed Computing - Special section best papers from the 2002 international parallel and distributed processing symposium

Quantified Score

Hi-index	0.00

Visualization

Abstract

Various I/O placement strategies have been proposed for a kxk toroidal distributed system. In previous studies, it has been shown that I/O placement strategy has a major impact on average network latency. In this study the sensitivity of different strategies toward I/O data locality has been investigated. An intensive simulation has been carried out for this purpose. The simulation considers both scientific and transaction computing environments. Simulation results show that the sensitivities of different strategies can vary significantly. In addition, it has been noticed that the sensitivity of an I/O placement strategy is influenced by two other factors: network size, and network contention.