integrate mx dim1 triton kernel into MXLinear

[vkuzo]

Summary:

Adds a config option to MXLinear to use a triton kernel for mxfp8 cast
across dim1.

Test Plan:

Reviewers:

Subscribers:

Tasks:

Tags:

[vkuzo]

Stack from ghstack (oldest at bottom):

• mx_formats: move inference to the quantize_ API #1971
• mx_formats: move training to the quantize_ API #1970
• use torch.float8_e8m0fnu in mx_formats #1966
• delete mxfp8 torch._scaled_mm wrapper #1965
• mx roofline: adjust mxfp8 formulas #1953
• print MX config when printing MXLinear and MXInferenceLinear #1947
• -> integrate mx dim1 triton kernel into MXLinear #1943

[pytorch-bot]

🔗 Helpful Links

🧪 See artifacts and rendered test results at hud.pytorch.org/pr/pytorch/ao/1943

• 📄 Preview Python docs built from this PR

Note: Links to docs will display an error until the docs builds have been completed.

❌ 1 New Failure

As of commit f245d64 with merge base d32afef ():

NEW FAILURE - The following job has failed:

• .github/workflows/float8nocompile_test.yaml (gh)

This comment was automatically generated by Dr. CI and updates every 15 minutes.