diff options
| author | Dan Goodliffe <dan@randomdan.homeip.net> | 2024-06-30 19:05:02 +0100 |
|---|---|---|
| committer | Dan Goodliffe <dan@randomdan.homeip.net> | 2024-06-30 19:05:05 +0100 |
| commit | f83d7cc0e2f0f6558783a5f3f146c10417e9225a (patch) | |
| tree | c25d723d8e9a3954b41f48c0e7e64d7a6ed8e1cd /gfx/gl/uiShader.cpp | |
| parent | Fix run rules for perfs (diff) | |
| download | ilt-f83d7cc0e2f0f6558783a5f3f146c10417e9225a.tar.bz2 ilt-f83d7cc0e2f0f6558783a5f3f146c10417e9225a.tar.xz ilt-f83d7cc0e2f0f6558783a5f3f146c10417e9225a.zip | |
Add an InstanceVertices partition perf test
Summary:
* Given a trivially simple condition, like a bounding box, over 1
million items can be partitioned in under 3ms.
* Parallel algorithms reduce that a little but are only effective with
volumes in excess ~200k, this might be better with a more complex
condition/predicate.
Diffstat (limited to 'gfx/gl/uiShader.cpp')
0 files changed, 0 insertions, 0 deletions