A Performance Prediction Methodology for Data-dependent Parallel Applications

The increase in the use of parallel distributed architectures in order to solve large-scale scientific problems has generated the need for performance prediction for both deterministic applications and non-deterministic applications. In particular, the performance prediction of datadependent programs is an extremely challenging problem because for a specific issue the input datasets may cause different execution times. Generally, a parallel application is characterized as a collection of tasks and their interrelations. If the application is time-critical it is not enough to work with only one value per task, and consequently knowledge of the distribution of task execution times is crucial. The development of a new prediction methodology to estimate the performance of data-dependent parallel applications is the primary target of this study. This approach makes it possible to evaluate the parallel performance of an application without the need of implementation. A real data-dependent ...

• Type: Conference
• Year: 2006
• Where: CLUSTER
• Authors: Paula Cecilia Fritzsche, Concepció Roig, Ana Ripoll, Emilio Luque, Aura Hernandez
• Keywords: CLUSTER 2006, Cluster Computing, Execution Times, Parallel Applications, Performance Prediction