Developer Blog: How to Migrate to the Latest Versions of the Nsight Suite

Discuss (0)

We recently published three developer blogs to help you migrate from NVIDIA Visual Profiler, and NVProf to the new generation of Nsight Tools and Nsight systems. If you missed the posts, here is a summary:

The NVIDIA Nsight suite provides a powerful, fast, and feature-rich set of tools, enabling you to more effectively analyze and profile your applications.

This new generation of Nsight tools are architected to better address the challenges developers face and to provide:

  • Performance analysis for longer captures and larger data files
  • Better insights into both system and kernel profiling via specialized Nsight Systems and Nsight Compute tools
  • More open and broad architecture for supporting future enhancements and features

Migrating to the Nsight Tool Suite for Analysis and Optimization

This series of articles describes the new NVIDIA Nsight tools in depth, shows the functionality through examples, and explains how to move your development to the new tools.

Migrating to NVIDIA Nsight Tools from NVVP and Nvprof

This first article explains how to transition from the earlier generation NVIDIA Visual Profiler or nvprof, to the more efficient and expansive NVIDIA Nsight tools. The article describes the benefits of migrating your profiling and analysis tools from NVVP and nvprof. It lists feature support tables and examples of both Nsight Systems and Nsight Compute usage to detail the different focus areas.

Read more>

Transitioning to Nsight Systems from NVIDIA Visual Profiler / nvprof

This next blog focuses on Nsight Systems usage, profiling overarching system interactions, including code samples, examples of how you would have profiled in the past, and how you can more effectively profile your application now with Nsight Systems GUI and/or CLI operation options. The story also lists other capabilities and benefits of the new Nsight tools. 

Read more>

Using Nsight Compute to Inspect your Kernels

The third blog focuses on Nsight Compute to help you better implement and optimize the use of innovative new GPU features that emerge with each new architecture generation. This content will leverage Nsight Compute is to provide access to kernel-level analysis using GPU performance metrics.

The article includes code samples, a discussion of which metrics to use to analyze kernel operation, both GUI and CLI use, and lists new capabilities included with Nsight Compute above and beyond past generations of NVIDIA Visual Profiler and nvprof.

Read more>

If you have been a nvprof or NVIDIA Visual Profiler user before, you’ll want to review these blogs and leverage the benefits of moving to the NVIDIA Nsight tools.

Learn more about the latest Nsight Systems 2020.1 release.