A distributed heterogeneous supercomputing management system
- 1 June 1993
- journal article
- Published by Institute of Electrical and Electronics Engineers (IEEE) in Computer
- Vol. 26 (6) , 78-86
- https://doi.org/10.1109/2.214443
Abstract
A general management framework for distributed heterogenous supercomputing systems (DHSSs) that is based on an application-characterization technique is presented. The technique uses code profiling and analytical benchmarking of supercomputers. An optimal scheduling of tasks in these systems is an NP-complete problem. The use of network caching to reduce the complexity associated with the scheduling decisions is discussed. An experimental prototype of a DHSS management system is described.Keywords
This publication has 9 references indexed in Scilit:
- Network Supercomputing: Experiments with a Cray-2 to Cm-2 Hippi ConnectionPublished by Institute of Electrical and Electronics Engineers (IEEE) ,2005
- Augmenting the Optimal Selection Theory for SuperconcurrencyPublished by Institute of Electrical and Electronics Engineers (IEEE) ,2005
- Estimation of Execution times on Heterogeneous Supercomputer ArchitecturesPublished by Institute of Electrical and Electronics Engineers (IEEE) ,1993
- On the assignment problem of arbitrary process systems to heterogeneous distributed computer systemsIEEE Transactions on Computers, 1992
- Benchmark characterizationComputer, 1991
- PAWS: a performance evaluation tool for parallel computing systemsComputer, 1991
- Gigabit network testbedsComputer, 1990
- Computer benchmarking: Paths and pitfalls: The most popular way of rating computer performance can confuse as well as inform; avoid misunderstanding by asking just what the benchmark is measuringIEEE Spectrum, 1987
- The Effects of Problem Partitioning, Allocation, and Granularity on the Performance of Multiple-Processor SystemsIEEE Transactions on Computers, 1987