Repository navigation
Tune CUDA gaph sizes on B200 and H100 - #3077
Merged
Merged
Conversation
Member
Author
|
CC @jessegross seems like the values on the consumer GPU work well here also. I tried a range on both machines. and this was more or less optimal. |
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
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Outdated suggestions cannot be applied.This suggestion has been applied or marked resolved.Suggestions cannot be applied from pending reviews.Suggestions cannot be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.
Benchmarks for inference. I also ran a training benchmark (1B Transformer) and didn't notice much change.
H100
Pre: prompt_tps=27330.184, generation_tps=184.120, peak_memory=9.545
Post: prompt_tps=30322.676, generation_tps=209.213, peak_memory=10.007
B200
Pre: prompt_tps=45910.345, generation_tps=228.303, peak_memory=9.332
Post: prompt_tps=45875.251, generation_tps=239.479, peak_memory=9.787
H100
Pre: prompt_tps=6885.463, generation_tps=87.226, peak_memory=62.217
Post: prompt_tps=6910.998, generation_tps=96.125, peak_memory=64.219
B200
Pre: prompt_tps=7885.334, generation_tps=114.522, peak_memory=62.715
Post: prompt_tps=7905.644, generation_tps=118.695, peak_memory=64.219