Skip to content

Add NVLink profiling fallback#21

Open
hdq66666 wants to merge 1 commit into
KempnerInstitute:mainfrom
hdq66666:main
Open

Add NVLink profiling fallback#21
hdq66666 wants to merge 1 commit into
KempnerInstitute:mainfrom
hdq66666:main

Conversation

@hdq66666

Copy link
Copy Markdown

Summary

Add a fallback for NVLink bandwidth reporting when DCGM_FI_DEV_NVLINK_BANDWIDTH_TOTAL is unavailable.

I observed that field 449 reports N/A on Tesla V100 systems, while the NVLink TX/RX profiling fields still provide valid values. This change has been tested successfully on V100.

Changes

  • Collect NVLink TX/RX profiling fields 1011/1012.
  • Prefer field 449 when available.
  • Fall back to TX + RX when field 449 is N/A.
  • Add tests for parsing and fallback behavior.

Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment

Labels

None yet

Projects

None yet

Development

Successfully merging this pull request may close these issues.

1 participant