This website requires JavaScript.
Explore
Help
Sign In
izzy
/
UnrealEngineUWP
Watch
0
Star
0
Fork
0
You've already forked UnrealEngineUWP
mirror of
https://github.com/izzy2lost/UnrealEngineUWP.git
synced
2026-03-26 18:15:20 -07:00
Code
Issues
Packages
Projects
Releases
Wiki
Activity
Files
0e5ef4fa337d831ceb9977da2a374aaa8dd24c19
UnrealEngineUWP
/
Engine
/
Shaders
History
tim doerries
19f9cbffa0
Implemented support for SparseVolumeTextures falling back to coarser mip levels at page level. This is the first step in moving towards streaming at page granularity instead of mip level granularity. The main problem to solve here is to update the descendands of every streamed in/out page so they point to the proper fallback mip data. In this CL, this is done on the CPU using a new always loaded octree representing the sparse page table topology as well as bitsets to keep track of changes in page GPU residency. In the future, we could maybe move this step to the GPU: For every valid mip0 page a CS thread would walk up the mip chain to find the finest resident mip level to fall back to and then write it to all the relevant pages along the path to the parent.
...
#rb Patrick.Kelly #jira UE-197907 #rnx [CL 30803748 by tim doerries in ue5-main branch]
2024-01-23 09:55:23 -05:00
..
Private
Implemented support for SparseVolumeTextures falling back to coarser mip levels at page level. This is the first step in moving towards streaming at page granularity instead of mip level granularity. The main problem to solve here is to update the descendands of every streamed in/out page so they point to the proper fallback mip data. In this CL, this is done on the CPU using a new always loaded octree representing the sparse page table topology as well as bitsets to keep track of changes in page GPU residency. In the future, we could maybe move this step to the GPU: For every valid mip0 page a CS thread would walk up the mip chain to find the finest resident mip level to fall back to and then write it to all the relevant pages along the path to the parent.
2024-01-23 09:55:23 -05:00
Public
Use DoubleFloat instead of tile+offset in various systems on GPU to improve numerical precision.
2024-01-21 21:02:35 -05:00
Shared
Add hair position compression, which reduce position data by 30%.
2024-01-23 02:58:03 -05:00
StandaloneRenderer
PR
#11088
: Slate: Fix custom element shader for StandaloneRenderer
2023-12-02 18:37:25 -05:00
Shaders.Build.cs
Added shader.build.cs file.
2023-07-27 22:53:25 -04:00