I'd like some clarification on this too. In an ideal multithreaded context with every single polygon being handed out to a different GPU core the draw call completing with the last polygon is rasterised it'd take as long as the biggest polygon does to draw. However, I doubt polygons are "truly" multithreaded like this…
Yes thank you. I guess I did my UVs like that because I was thinking to much about offsetting for some reason lol. So as a test I just baked my normal onto a simple 20 sided cylinder. It turned out pretty good with the exception of edge around the top, but I have a few questions. 1) Am I correct in assuming that the dark…