In the Frame Profiler tab I can monitor that the FB is the largest bottleneck in a pass, while ROP is low. As I have read FB means Frame Buffer Unit, what exactly is this? How is the functionality separated between FB and ROP? Are the same optimization techniques addressing ROP also addressing FB?
I also have a more common question. How is the bottlenecks messured? A lot of times a unit with low utilization is high bottleneck and vice versa.
I also have a more common question. How is the bottlenecks messured? A lot of times a unit with low utilization is high bottleneck and vice versa.
