SLI Optimizations

Hi,

I’m trying to get my Engine to work well with SLI. I have a Geforce 690 GTX for testing purposes and I was able to verify that my engine could run at twice the FPS as it does right now (renaming the executable to AFR-FriendlyD3D.exe). I was trying to follow all the recommendations in the SLI best practices, but even after clearing all render targets before setting them, creating the swap chain with the DXGI_USAGE_DISCARD_ON_PRESENT flag and disable all use of stream output buffers I still run on half of the potential frame rate. Is there any way to find out, which resources are copied between the gpu cores? Any profiling tool or anything like that? How is anyone supposed to optimize their stuff for this, if there are no tools available?

Bump.