Robustness analysis for the method of assignment MATEHa

Abstract

The TTIGHa model has been developed to model and predict the performance of parallel applications run over heterogeneous architectures. In addition, the task assignment algorithm was implemented to MATEHa processors based on the TTIGHa model. This paper analyzes the assignment algorithm robustness before different variations which the model parameters may undergo (basically, communication and processing times).

Keywords

• parallel systems
• cluster and multi-cluster architectures
• performance prediction models
• tasks to processors mapping
• heterogeneous processors
• robustness