ggml-vulkan: limit the number of command buffers to transfer data to 4#19956
Closed
mklimenk wants to merge 1 commit intoggml-org:masterfrom
Closed
ggml-vulkan: limit the number of command buffers to transfer data to 4#19956mklimenk wants to merge 1 commit intoggml-org:masterfrom
mklimenk wants to merge 1 commit intoggml-org:masterfrom
Conversation
Collaborator
|
We've been discussing possible solutions in #19420. This is one of them, yes. |
Author
|
@0cc4m I checked PRs for duplicates, but didn't look into issues. Thank you for looking into it, closing this as a duplicate. |
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Outdated suggestions cannot be applied.This suggestion has been applied or marked resolved.Suggestions cannot be applied from pending reviews.Suggestions cannot be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.
This PR introduces a limit to the amount of command buffers to force reusing them instead of creating a new command buffer for each megabyte of data to transfer.
I stumbled upon this limitation while trying to launch GPT-OSS-120B MXFP4 via Vulkan backend, which failed during the model loading stage.
Adding the proposed fix allows for a command buffer reuse instead of creating a new one for every transfer, which result in a lot of command buffers for larger models and might hit GPU hardware limitations. A number 4 is taken from the number of staging buffers for async uploads, but can be adjusted. No performance or accuracy implications were observed.