fix faulty assumption about remote key sizes #137
Merged
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Outdated suggestions cannot be applied.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.
I had put an assumption that the memory required to store UCX remote keys for GPU memory would always be larger than for Host memory.
This assumption is incorrect. The size of buffer for the keys is dependent on which transports are enable, the number of NICs, the number of GPUs and a variety of other runtime variables used to initialize UCX.
I've left a comment in the tests to remind developers and users of that fact.