I know GPU prefetches textures and that's why dependent texture reads are slower, but how does it work and at what point that happens?

EDIT: Split the content of this question into others as suggested by trichoplax Here's a link to other questions:

How does Texture Cache work considering multiple shader units

How does Texture Cache work in Tile Based Rendering GPU

Is using many texture maps bad for caching?

  • $\begingroup$ This seems like several interesting questions in one. At least (4), (5) and (6) sound like independent questions, and maybe (2) and (3) could be a single question. I'd upvote any of those as separate questions, but all in one it seems too broad. $\endgroup$
    – trichoplax
    Aug 26 '15 at 22:31
  • $\begingroup$ Ok! Will do separates and then delete this. thx! $\endgroup$ Aug 26 '15 at 22:34
  • $\begingroup$ Actually, I just edited this and kept link to others. $\endgroup$ Aug 26 '15 at 22:54