The International Conference for High Performance Computing, Networking, Storage, and Analysis
The International Conference for High Performance Computing, Networking, Storage, and Analysis
Authors: Suraj P. P. Kesavan (University of California, Davis); Harsh Bhatia (Lawrence Livermore National Laboratory); Abhinav Bhatele (University of Maryland); Stephanie Brink, Olga Pearce, Todd Gamblin, and Peer-Timo Bremer (Lawrence Livermore National Laboratory); and Kwan-Liu Ma (University of California, Davis)
Abstract: Optimizing the performance of large-scale parallel codes is critical for efficient utilization of computing resources. Code developers often explore multiple execution parameters and are interested in detecting and understanding bottlenecks in different executions. They usually collect hierarchical performance profiles represented as call graphs, which combine performance metrics with their execution contexts. The crucial task of exploring multiple call graphs together is challenging because of the structural differences in the execution contexts and significant variability among performance metrics. We present an enhanced version of CallFlow to support the exploration of ensembles of call graphs. The visual design combines the Sankey diagram's strengths to describe the resource flow of the call graphs and box-plot visualization techniques to convey the performance variability within the ensemble. Our interactive visual interface provides linked views to explore the ensembles of call graphs by facilitating the analysis of structural differences and identifying similar or distinct call graphs.
Best Poster Finalist (BP): yes
Poster: PDF
Poster summary: PDF
Back to Poster Archive Listing