Skip to content

Conversation

@IMbackK
Copy link
Collaborator

@IMbackK IMbackK commented Jul 28, 2025

This is useful for testing for regressions on GCN with CDNA hardware.

With GGML_HIP_MMQ_MFMA=Off and GGML_CUDA_FORCE_MMQ=On we can conveniently test the GCN code path on CDNA. As CDNA is just GCN renamed with MFMA added and limited use ACC registers, this provides a good alternative for regression testing when GCN hardware is not available.

…This is usefull for testing for regressions on GCN with CDNA hardware.
@github-actions github-actions bot added Nvidia GPU Issues specific to Nvidia GPUs ggml changes relating to the ggml tensor library for machine learning labels Jul 28, 2025
@IMbackK IMbackK requested a review from JohannesGaessler July 28, 2025 21:01
@IMbackK
Copy link
Collaborator Author

IMbackK commented Jul 28, 2025

@deepsek this one is particularly for your benefit.

@deepsek
Copy link
Contributor

deepsek commented Jul 28, 2025

@IMbackK I will try to get a setup with other architectures with wavefront 64 size to test for regressions based on work done on newer CDNA arch. This might take a while.
This will be extremely useful in the meantime!

@IMbackK IMbackK merged commit b77d111 into ggml-org:master Jul 29, 2025
47 checks passed
Nexesenex added a commit to Nexesenex/croco.cpp that referenced this pull request Aug 7, 2025
Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment

Labels

ggml changes relating to the ggml tensor library for machine learning Nvidia GPU Issues specific to Nvidia GPUs

Projects

None yet

Development

Successfully merging this pull request may close these issues.

3 participants