Abstract: Scientific simulation frameworks are common to use on HPC systems. They contain parallelized algorithms and provide various solvers for a specific application domain. Usually, engineers execute multiple steps to solve a particular problem which are often distributed over multiple jobs. Finding performance bottlenecks and the causing step in such a complex system is very difficult. Therefore in this work, we present a top-down approach that provides summarized performance metrics for the workflow, jobs and job steps. These summaries guides the user to identify inefficiencies and determine the causing job step. Finally, Vampir can be used for a detailed analysis of the regarding execution in order to resolve the issue.
Best Poster Finalist (BP): no
Poster summary: PDF
Reproducibility Description Appendix: PDF
Back to Poster Archive Listing