I’m using CUDA 4.0 RC2 on Centos 5.4 Linux with a Quadro FX 5800 and a Tesla C2050.
I’ve been trying to manipulate OpenGL textures via CUDA. For speed, I’ve been attempting to access the OpenGL textures through CUDA.
Now, I can map the OpenGl 3D texture to a CUDA array and copy into it… but I can only copy the first slice! Like, my 3D texture is X x Y x Z, but I
can only copy X x Y elements into it. It displays that just fine.
Can anyone think of a fix besides using a bunch of 2D textures instead of one 3D texture?