A methodology for dimensioning SPMD distributed systems applied to industrial applications
The aim of this study is to develop a method to determine the minimum number of processors which are necessary to satisfy the time constraints imposed by an application. For this purpose, a mathematical algorithm based on simplex optimisation technique has been developed. The method described may be used to estimate the computation time of an industrial application parallelised in SPMD mode and simulated on MIMD or SIMD machines communicating in rendezvous mode. The architecture of the network of processors may be a pipeline or a tree depending upon the number of communication links used. The distribution of elementary work tasks is proportional with respect to all the network processors.
Bibliographic Reference: Paper presented: The International Conference and Exhibition on High Performance Computing Networking, Münich (DE), April 18-20, 1994
Availability: Available from (1) as Paper EN 38156 ORA
Record Number: 199410628 / Last updated on: 1994-11-28
Original language: en
Available languages: en