Vulkan: Default to 1GB allocations instead of 4GB to avoid fragmentation and driver issues #12434
+1
−5
Conversation
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
…ion and driver issues
github-actions
bot
added
Vulkan
jeffbolznv
approved these changes
Mar 17, 2025
arthw
pushed a commit
to arthw/llama.cpp
that referenced
this pull request
Mar 19, 2025
…ion and driver issues (ggml-org#12434)
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Outdated suggestions cannot be applied.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.
Extend the changes done in #11551 to all cases to avoid driver allocation issues that keep coming up. On some drivers allocations fail for other reasons than fragmentation, despite being below the size limit. Some examples in #5441. This should fix that.
I couldn't reproduce performance differences I saw in the previous PR, so it might just have been noise. Sorry about that @jeffbolznv