Understand the Workflow
Follow this workflow to use Intel® VTune™ Profiler to identify and analyze performance bottlenecks in your serial or parallel application. This tutorial uses a sample application called matrix.
Prerequisites
Download these Intel software tools to your Linux system:
Intel® VTune™ Profiler version 2025.0 or newer
You can get both tools in the Intel® oneAPI Base Toolkit.
NOTE:
This tutorial uses the Intel® oneAPI DPC++/C++ Compiler to establish a common baseline for analysis and to track performance gain. Your choice of a different compiler may change your results in this workflow.
Workflow
To find and fix performance issues in the matrix sample application,
Establish a baseline for application performance.
Identify any bottlenecks in the matrix application. Run the Memory Access analysis and interpret your result.
Eliminate memory access bottlenecks, if any. Recompile the application.
Assess the performance improvement. Run the HPC Performance Characterization analysis and interpret your result.
See the performance gain. Compare results before and after optimization.
GUI and Command Line Use
This tutorial explains how to run an analysis from the VTune Profiler graphical user interface (GUI). You can also use the VTune Profiler command-line interface (vtune command) to run an analysis. For any analysis type that you run from the GUI, get the equivalent command line syntax by clicking the Command Line button at the bottom of the window. The Command Line Interface section has more information.