-
Notifications
You must be signed in to change notification settings - Fork 1.1k
New issue
Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.
By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.
Already on GitHub? Sign in to your account
DRAFT: Vectorize ForUtil encoding for the 9.0 codec (same format) #12412
base: main
Are you sure you want to change the base?
Conversation
…9.0 codec (only encode!). It should show as example how it works, this is not ready for productions, although all tests pass
Important: I don't intend to merge this, it just shows WIP. |
lucene/core/src/java/org/apache/lucene/internal/vectorization/VectorizationProvider.java
Show resolved
Hide resolved
# Conflicts: # lucene/core/src/java/org/apache/lucene/internal/vectorization/VectorizationProvider.java
What's also missing in the code: Some randomized test that feeds both implementations with random data and verifies that both results are identical. |
to run this branch with vectorization enabled: $ export CI=true
$ export RUNTIME_JAVA_HOME=/path/to/jdk20or21
$ gradlew :lucene:core:test |
Thanks for putting this together @uschindler. I'm going to move my experiments over to a clone of this branch. |
Now that we are removing patching for the doc block encoding maybe vectorizing decode of these blocks is more palatable? |
DRAFT: This uses code from #12396 to vectorize ForUtil for the 9.0 codec (only encode!). It should show as example how it works, this is not ready for productions, although all tests pass.