Better GPU memory monitoring, and works with more GPUs #1
Conversation
This file contains 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
added 2 commits
November 22, 2021 15:25
The code now has fewer assumptions about the data which nvidia_smi can deliver. On some GPUs (probably some cheap models) it many parameters may have a value of "N/A", and those are now ignored. Also, the GPU memory is better read from the gpu_fb_memory_usage_used value, at least on the GPUs I've tested. So the number of memory related performance data has been expanded. Maybe the memory_util monitoring point should be removed, unless we know for sure that it is really useful with some GPU/driver combinations?
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
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Outdated suggestions cannot be applied.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.
These changes have been running in several setups for a number of weeks, involving several different NVIDIA cards, including a very primitive one which can only report GPU temperature and GPU clock frequency.
The pull request does not contain an updated MKP; I can add an updated MKP, if needed.