Conversation
njroussel
force-pushed
the
ser
branch
2 times, most recently
from
7673107 to
3467fc7
Compare
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
1 participant
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Outdated suggestions cannot be applied.This suggestion has been applied or marked resolved.Suggestions cannot be applied from pending reviews.Suggestions cannot be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 adds
dr.reorder_threads(): an explicit way for users to trigger the SER functionality of their GPU without it being tied to a ray-intersection.Related: mitsuba-renderer/drjit-core#145
The function takes a
keyargument that acts as the sorting key for the reordering as well as avalueargument. The latter solely serves a a way for us to tie the operation into the JIT graph.Because the shuffling happens under the hood, there really isn't a way to test this properly other than a smoke test. I've added some basic ones in a new `test_reorder.py file.