Use instancing for sprite rendering #8872
Closed
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
alice-i-cecile
added
C-Performance
|
there are FPS regressions for me on both
|
|
This is (surprisingly) faster on my system. (for bevymark, I'm seeing slightly better minimums, and ~112fps average vs ~90fps average for our current batching, for many_sprites perf is the same). The old wisdom was that you want to use batching for sprites because you're drawing quads, and GPUs run wavefronts of 32-64 vertices (so you end up with very low utilization per wavefront). However, these days (some? AMD, maybe Nvidia, others idk) GPU drivers can merge instances into wavefronts. (but sometimes it merges and it's still slow? idk) Basically, need people to test this on a bunch of different hardware and check wavefront occupancy to figure out who does merging and who doesn't and if we're okay with the performance tradeoffs. (or maybe even have both paths, but have it be an option?) |
github-merge-queue bot
pushed a commit
that referenced
this pull request
Sep 2, 2023
# Objective - Supercedes #8872 - Improve sprite rendering performance after the regression in #9236 ## Solution - Use an instance-rate vertex buffer to store per-instance data. - Store color, UV offset and scale, and a transform per instance. - Convert Sprite rect, custom_size, anchor, and flip_x/_y to an affine 3x4 matrix and store the transpose of that in the per-instance data. This is similar to how MeshUniform uses transpose affine matrices. - Use a special index buffer that has batches of 6 indices referencing 4 vertices. The lower 2 bits indicate the x and y of a quad such that the corners are: ``` 10 11 00 01 ``` UVs are implicit but get modified by UV offset and scale The remaining upper bits contain the instance index. ## Benchmarks I will compare versus `main` before #9236 because the results should be as good as or faster than that. Running `bevymark -- 10000 16` on an M1 Max with `main` at `e8b38925` in yellow, this PR in red:  Looking at the median frame times, that's a 37% reduction from before. --- ## Changelog - Changed: Improved sprite rendering performance by leveraging an instance-rate vertex buffer. --------- Co-authored-by: Giacomo Stevanato <giaco.stevanato@gmail.com>
|
Thanks for the contribution. Closing this as #9597 was merged. |
rdrpenguin04
pushed a commit
to rdrpenguin04/bevy
that referenced
this pull request
Jan 9, 2024
# Objective - Supercedes bevyengine#8872 - Improve sprite rendering performance after the regression in bevyengine#9236 ## Solution - Use an instance-rate vertex buffer to store per-instance data. - Store color, UV offset and scale, and a transform per instance. - Convert Sprite rect, custom_size, anchor, and flip_x/_y to an affine 3x4 matrix and store the transpose of that in the per-instance data. This is similar to how MeshUniform uses transpose affine matrices. - Use a special index buffer that has batches of 6 indices referencing 4 vertices. The lower 2 bits indicate the x and y of a quad such that the corners are: ``` 10 11 00 01 ``` UVs are implicit but get modified by UV offset and scale The remaining upper bits contain the instance index. ## Benchmarks I will compare versus `main` before bevyengine#9236 because the results should be as good as or faster than that. Running `bevymark -- 10000 16` on an M1 Max with `main` at `e8b38925` in yellow, this PR in red:  Looking at the median frame times, that's a 37% reduction from before. --- ## Changelog - Changed: Improved sprite rendering performance by leveraging an instance-rate vertex buffer. --------- Co-authored-by: Giacomo Stevanato <giaco.stevanato@gmail.com>
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.
Objective
Solution
Store per-instance data instead of 6 vertices to the vertex buffer and change the
VertexStepModetoInstance.This allows for less data being transferred.
Originally, 1728 bits (216 bytes) had to be transferred to the GPU for each sprite.
Now, only 768 bits (96 bytes) are needed per-sprite.
Performance
(Sprite size set to 0.01 to minimize sampler performance impact, buffer size captured from Intel GPA)
bevymarkOriginal: 2150408 bytes transferred per frame, 145-147 fps.
Instanced: 960004 bytes transferred per frame, 158-161 fps.