Original Post
I am using the Hardware Shadow Mapping using NVidia graphics card on my project, and it works fine. However, I am using quite a large shadow map, and I am forced to create a render target texture in addition to the depth texture in order to render to the depth buffer. The memory for this texture is just wasted, as I am only rendering the depth (calling SetRenderState(D3DRS_COLORWRITEENABLE, 0)) before rendering to the shadow map). Is there any way around this, or do you have to create a render target texture that you do not use in order for the hardware shadow mapping to work?