Precalc some splat data each frame #6
Merged
Conversation
This file contains 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
Instead of computing it for each vertex of the quad we're drawing. On M1 Mac/Metal gets down from 218ms (increased a lot since chunks/textures) to 58ms. Downside is 40 bytes/splat of runtime GPU memory
dylanebert
pushed a commit
to dylanebert/UnityGaussianSplatting
that referenced
this pull request
Oct 13, 2023
Precalc some splat data each frame
HIHHIYAYAYOO
pushed a commit
to HIHHIYAYAYOO/UnityGaussianSplatting
that referenced
this pull request
Nov 30, 2023
Precalc some splat data each frame
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Outdated suggestions cannot be applied.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.
...instead of computing it over and over again for each vertex of each quad that is being rendered.
This does not help on PC (RTX 3080Ti) much (13.4ms -> 12.6ms), but massively helps on Mac (M1), which had previously regressed in perf due to using all the data chunking & textures (from initial 80ms it had went up to 220ms). But now it's down to 55ms, so yay progress! And a lesson that vastly different GPU architectures have vastly different performance characteristics, who would have thunk.
Downside is that this now needs 40 bytes/splat of extra GPU memory to store the view data. Might be able to get that smaller, maybe.