Previous | Next --- Slide 73 of 79
Back to Lecture Thumbnails

How is the GPU's special hardware to perform point-in-triangle tests better than how a CPU would do it? Is there some hardware optimization? Or just more processors?


Recalling that we've discussed cache optimization in classes before, are there "blocking" techniques that could help us optimize which triangle bounding boxes to point-test first? My first guess would be points that are near each other, but I'm not sure if that's exactly correct...