Is it possible to use an InstancedMesh and keep matrix computation on the GPU? I’m trying to avoid a GPU->CPU->GPU data pass. I can see there is an InstancedMeshNode and this has an InstanceMatrixBuffer. But not sure how to use it.
1 Like
I don’t work with TSL, but it’s a shader this should be possible. You can compose a matrix from position rotation scale and then apply it to vertices.
1 Like