Français
2. Efficiency of asynchronous parallel algorithms
In this section, we analyze, with reference to applications and associated parallel simulations, when synchronous or asynchronous methods are worth using. Beforehand, we give some indicators for measuring the performance of parallel algorithms.
2.1 Cluster-based performance measurement
To measure performance in terms of reduced job rendering time on multi-processor machines, a number of indicators can be used to measure the gain (or otherwise) obtained when using a parallel calculator. The ideal case is when the ratio of sequential time to parallel time (i.e. acceleration) is equal to the number of processors. However, except in special cases, this cannot practically be the case, as parallelizing a calculation code requires the activation of processors, which...
The Ultimate Scientific and Technical Reference
This article is included in
Mathematics
This offer includes:
Knowledge Base
Updated and enriched with articles validated by our scientific committees
Services
A set of exclusive tools to complement the resources
Practical Path
Operational and didactic, to guarantee the acquisition of transversal skills
Doc & Quiz
Interactive articles with quizzes, for constructive reading
Efficiency of asynchronous parallel algorithms
Conclusion
Bibliography
The Ultimate Scientific and Technical Reference
