[bugfix] preserve DRAM buffer lifetime to restore inference accuracy #322
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
ygwpz
approved these changes
Oct 29, 2025
flesher0813
pushed a commit
that referenced
this pull request
Oct 30, 2025
…322) * linear buffer for device * check data consistency after embedding
you-seesee-you
pushed a commit
to you-seesee-you/unified-cache-management
that referenced
this pull request
Nov 5, 2025
…odelEngine-Group#322) * linear buffer for device * check data consistency after embedding
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
2 participants
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Outdated suggestions cannot be applied.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.
Problem
When KVCache is read concurrently, the DRAM buffer is freed and reused before the asynchronous H2D copy finishes. As a result, the device reads to the stale buffer and inference accuracy is corrupted.
Fix
Keep the buffer alive until the device stream signals completion.
Verification
Added read-after-write checks to the E2E preload script to guarantee data consistency.