Sending the textures once to the graphics card. CUDA should be able to access it.

I am looking for a way to send Textures to the graphics card via CUDA or GL once and for CUDA threads to access it …basically I want to texture mapping in a dynamic scene…
In my current implementatio of texturing( referred simpleTextures example)…I bind textures but this happens every frame…which I think is poor.
I just want to send Textures to the card ONCE and have CUDA access it on thread-by-thread level.
I hope I made the problem statement clear.


Once a texture is bound, you don’t need to bind it again.