Nsight Analysis System: Build Custom Python Analysis Scripts to Summarize Performance and Reveal Bottlenecks With Single- and Multi-Node Applications

In this hands-on lab, we'll show you how to write custom analysis scripts, called recipes, to gain deeper insights into the performance of your application using the Analysis System of the Nsight Systems profiler. We'll walk through multiple examples to show the Nsight Analysis System functions that can process, filter, correlate, and visualize the various hardware and software events captured by Nsight Systems. We'll also show how to upgrade a single-report recipe to a multi-report recipe in order to identify performance issues in large-scale HPC and cloud applications.
Prerequisite(s):

Familiarity with Nsight Systems, Python, Pandas, and Jupyter Notebooks.