Enhance metrics parsing and add CUDA stencil example with profiling#13
Merged
Alessandro624 merged 9 commits intodevfrom Dec 30, 2025
Merged
Enhance metrics parsing and add CUDA stencil example with profiling#13Alessandro624 merged 9 commits intodevfrom
Alessandro624 merged 9 commits intodevfrom
Conversation
…plots for better precision
Alessandro624
added
documentation
This file contains hidden or bidirectional Unicode text that may be interpreted or compiled differently than what appears below. To review, open the file in an editor that reveals hidden Unicode characters.
Learn more about bidirectional Unicode characters
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
1 participant
Add this suggestion to a batch that can be applied as a single commit.This suggestion is invalid because no changes were made to the code.Suggestions cannot be applied while the pull request is closed.Suggestions cannot be applied while viewing a subset of changes.Only one suggestion per line can be applied in a batch.Add this suggestion to a batch that can be applied as a single commit.Applying suggestions on deleted lines is not supported.You must change the existing code in this line in order to create a valid suggestion.Outdated suggestions cannot be applied.This suggestion has been applied or marked resolved.Suggestions cannot be applied from pending reviews.Suggestions cannot be applied on multi-line comments.Suggestions cannot be applied while the pull request is queued to merge.Suggestion cannot be applied right now. Please check back later.
Description
This PR improves the robustness of the metrics parsing pipeline and introduces a new CUDA stencil module, extending the set of GPU computation patterns covered by the project.
Key changes
Enhanced CSV parsing in
parse_metrics.py, improving kernel name extraction and overall robustness when processing profiling outputs.Updated histogram and occupancy handling in the metrics analysis workflow.
Added a new
stencil/module including:Makefilefor standardized builds.run.shfor execution.profile_nvprof.shfor profiling support.READMEdocumenting usage and performance analysis steps.Updated
.gitignoreto include artifacts generated by the stencil module.Updated all
README.mdfiles to ensure consistency and reflect the latest project structure and tooling.Impact
The improved metrics parser increases the reliability of performance analysis across all CUDA examples. The stencil workload adds another memory-bound computation pattern, commonly used in scientific computing, enabling deeper exploration of cache behavior, occupancy, and bandwidth limitations.