Profiler Training Series, Part 3: What Tools to Use When Interactive
Event Details
Date and Time:
Thursday, February 19, 2026 from 1:00 PM to 3:00 PM Eastern Time
Presenters:
AMD Performance Tools Team in collaboration with Oak Ridge Leadership Computing Facility staff
About the Series:
This session is part of the Profiler Training Series, a multi-part training designed to help users build practical performance profiling skills on Frontier. View the full series schedule and upcoming sessions on the Profiler Training Series page.
Hackathon Connection:
Participation in this training series is strongly encouraged for teams planning to apply to the 2026 Spring Frontier Hackathon. The skills covered will help teams prepare competitive abstracts and optimize applications ahead of the hackathon. Visit the Frontier Hackathon page for event details and application information.
Event Overview
OLCF will host a two-hour interactive discussion focused on selecting profiling tools at different stages of the performance analysis process. This session is designed to help users understand when and why to apply specific tools, and how to move efficiently from high-level insight to targeted optimization.
The event combines a tool-agnostic overview, brief vendor introductions, a moderated panel discussion, and open Q&A to support practical decision-making across a wide range of application types.
Session Agenda
| Time (ET) | Session | Description |
|---|---|---|
| 1:00 PM – 1:30 PM | Profiling Strategy Overview | Tool-agnostic guidance on selecting profiling tools for light, medium, and heavy profiling, presented by Bob Robey (AMD). |
| 1:30 PM – 2:00 PM | Tool Introductions | Brief, single-slide overviews from participating tool vendors highlighting available profiling capabilities. |
| 2:00 PM – 2:30 PM | Moderated Panel Discussion | Guided discussion on when and why to use specific tools at different stages of performance analysis. Questions will be collected in Slack and addressed live by the panel. |
| 2:30 PM – 3:00 PM | Open Q&A | Interactive discussion where participants can bring their own profiling challenges and receive targeted guidance. |
Who Should Attend
This session is open to all users. Teams planning to participate in the 2026 Spring Frontier Hackathon are strongly encouraged to attend. Teams intending to apply but unable to attend should consider scheduling OLCF office hours either before or after the session. The hackathon call for proposals closes on March 4.
We are particularly interested in participation from teams running applications that are both AI- and HPC-intensive, regardless of whether you plan to apply to the Spring Frontier Hackathon.
Example Self-Assessment Questions
-
Do I know how my application actually performs on the system?
-
How do I profile PyTorch workloads, and why should I?
-
I use Fortran with OpenMP. How can I improve performance?
-
I need visibility inside my kernels. Which tools can help?
-
My I/O is slow. How can profiling help identify and address the bottleneck?
These examples are not exhaustive. If you have performance-related questions, this session is designed to help you think through them.
Presentation Materials
Slides | Recording
