Hello, I've a texture which I fill each frame and then display using a quad.
When the texture is kept in VRam it takes 3ms but when I keep it in normal ram it takes 2.3ms.
I am filling the texture from arrays using writepixelfast and locking the buffer.
I always thought keeping it in VRam was faster. I have a shadow texture which is much (10x) faster in VRam.
I think it must be because the processor is doing a lot of the filling of the texture?
Is there a rule to follow?
Marg
When the texture is kept in VRam it takes 3ms but when I keep it in normal ram it takes 2.3ms.
I am filling the texture from arrays using writepixelfast and locking the buffer.
I always thought keeping it in VRam was faster. I have a shadow texture which is much (10x) faster in VRam.
I think it must be because the processor is doing a lot of the filling of the texture?
Is there a rule to follow?
Marg