Tutorial: Analyzing MPI Applications with Intel® Trace Analyzer and Collector and Intel® VTune™ Amplifier XE
Intel®
Trace Analyzer and Collector enables you to understand MPI application
behavior and quickly find bottlenecks to achieve high performance for
parallel cluster applications. Use the Intel Trace Analyzer and Collector
to evaluate profiling statistics and load balancing, identify communication
hotspots, and increase application efficiency.
Intel®
VTune™ Amplifier XE enables you to find serial and parallel code bottlenecks
and speed execution. Use this tool to analyze the algorithm choices, understand
where and how your application can benefit from available hardware resources,
and identify code sections that may cause unnecessary power consumption.
Both products are installed as part of Intel® Parallel Studio XE Cluster Edition.
To improve performance of some complex applications, it is necessary to analyze their cross-process behavior as well as their single process performance. Intel Trace Analyzer and Collector enables you to analyze communications between processes, while Intel VTune Amplifier helps you find single process performance issues.
About This Tutorial |
This tutorial demonstrates a workflow applied to a sample program. The source code is available at <install-dir>/examples/poisson, where <install-dir> is the Intel Trace Analyzer and Collector installation directory. You can ultimately apply the same workflow to your own applications:
|
Estimated Duration |
20-25 minutes |
Learning Objectives |
After you complete this tutorial, you should be able to:
|
More Resources |
Learn more about the Intel Trace Analyzer and Collector and Intel VTune Amplifier in the guides and tutorials available at:
|
Submit Feedback |
You can submit your feedback on the documentation at http://www.intel.com/software/products/softwaredocs_feedback/. |