GenialPP / MinimalCompute

Geek Repo:Geek Repo

Github PK Tool:Github PK Tool

Minimal Compute Shader Examples

Minimal test scenes contains compute shaders, compute buffers etc Playing with the transport between CPU <-> GPU

Unity version : 2022.1+, contains both BuiltinRP and Universal RP (URP) scenes
See branches for older Unity versions
Tested with : Win DX11

Scene Image Description
ComputeUAVTexture The most basic one, edit texture with compute shader
AsyncGPUReadbackTex Similar to ComputeUAVTexture, but use AsyncGPUReadback to get texture data back to CPU
ComputeUAVTexFlow Example of using compute to animate texture pixels
ComputePaintTexture Paint the texture by sending object positions to compute shader
ComputePaintTexture_DFT Similar to above but drawing with Epicycles using Discrete Fourier Transform. Ref to The Coding Train's youtube video
StructuredBufferWithCompute Another basic one, use compute to calculate some data and send back to CPU
AsyncGPUReadback Similar to StructuredBufferWithCompute, but use AsyncGPUReadback to get array data back to CPU
StructuredBufferNoCompute ComputeBuffer doesn't always need to stick with ComputeShader
IndirectCompute Simple indirect compute (indirect dispatch) and CopyCount
IndirectReflectedStar Draw stars on the screen only if the pixels are bright enough
ComputeSketch Draw quads on the screen with color filled by compute shader
ComputeParticlesDirect GPU Particle, drawing fixed no. of particles
ComputeParticlesIndirect GPU Particle, drawing dynamic no. of particles, no need to read back to CPU!
ComputeVertex Replace vertex buffer with StructuredBuffer and drive vertex displacement by compute
SkinnedMeshBuffer Blend the vertex data from 2 SkinnedMeshRenderer vertex buffer and render it with MeshRenderer
SkinnedMeshBuffer_DiffMesh Similar to above but blending 2 different SkinnedMeshes. The blended triangles are drawn with DrawMeshInstancedIndirect()
ComputeVertexLit A usecase of ComputeVertex so that different shader passes share same vertex data
UAVInShader Read some data back to CPU from fragment shader
AsyncGPUReadbackMesh It is much faster to update mesh vertices with compute + AsyncGPUReadback to get the vertex data back to CPU for physics

Disclaimer: The stuff here might not be the best practice / optimized :'(. But at least they works. Play them for fun.

About


Languages

Language:C# 62.8%Language:ShaderLab 37.2%