aboutsummaryrefslogtreecommitdiff
path: root/README.md
diff options
context:
space:
mode:
authormakamys <makamys@outlook.com>2023-12-01 13:18:19 +0100
committerGitHub <noreply@github.com>2023-12-01 13:18:19 +0100
commita4ec946f5931f8fe6e6271b22c553f99b9d5afb8 (patch)
tree1faeaff10a98cdeac2ba2cfa6d91af2931b807bd /README.md
parent0fa977243a06dfe656f9798110ccaa80bb745dfb (diff)
downloadNeodymium-a4ec946f5931f8fe6e6271b22c553f99b9d5afb8.tar.gz
Neodymium-a4ec946f5931f8fe6e6271b22c553f99b9d5afb8.tar.bz2
Neodymium-a4ec946f5931f8fe6e6271b22c553f99b9d5afb8.zip
[skip ci] Remove mesh simplification from readme as well
Diffstat (limited to 'README.md')
-rw-r--r--README.md1
1 files changed, 0 insertions, 1 deletions
diff --git a/README.md b/README.md
index b348825..ab17927 100644
--- a/README.md
+++ b/README.md
@@ -49,7 +49,6 @@ The mod injects callbacks that run when world renderers (16x16x16 sections of th
There are also some additional optimizations:
* Face culling: Faces that aren't facing the camera won't be submitted for rendering. This reduces GPU workload, and will increase the framerate if the GPU was choking. Inspired by a similar optimization in Sodium.
-* Mesh simplification: Experimental feature. Identical adjacent quads are merged together into a single large quad with a wrapped texture. This reduces the amount of quads that have to be rendered without affecting the visuals. However, it also slows down the fragment shader. I found it to reduce performance overall, so it's disabled by default.
The mod increases memory usage, since the chunk meshes have to be stored somewhere. On Normal render distance in a vanilla world, it uses ~70-150 MB, both in RAM and VRAM.