[Mesa-dev] [RFC] [BRANCH] Floating point textures and rendering for Mesa, softpipe and llvmpipe

[Ian Romanick]

-----BEGIN PGP SIGNED MESSAGE-----
Hash: SHA1

Luca Barbieri wrote:

>> But we've been seeing some results which point that the whole color
>> buffer swizzling idea might be overrated: it increases memory bandwidth
>> usage substantially,
> 
> Why?
> It should decrease it due to a lower number of cache misses, thanks to
> having a 2D instead of a 1D neighborhood of a pixel in the cache.

For scenes with lots of small triangles (each triangle only covers one
fragment of a quad) you can get piles and piles of overdraw.  It's
turning out to be a major problem for hardware implementations as well.

See:

Fatahalian, K., Boulos, S., Hegarty, J., Akeley, K., Mark, W. R.,
Moreton, H., and Hanrahan, P. 2010. Reducing shading on GPUs using
quad-fragment merging. In ACM SIGGRAPH 2010 Papers (Los Angeles,
California, July 26 - 30, 2010). H. Hoppe, Ed. SIGGRAPH '10. ACM, New
York, NY, 1-8.  http://graphics.stanford.edu/papers/fragmerging/

-----BEGIN PGP SIGNATURE-----
Version: GnuPG v1.4.10 (GNU/Linux)
Comment: Using GnuPG with Mozilla - http://enigmail.mozdev.org/

iEYEARECAAYFAkx+twkACgkQX1gOwKyEAw8xCwCgk4dLoHhDAvsp505R1GnPbj+m
kecAn0XoQ+lPG152VIkpJRWYtiP0hYrU
=kVQ4
-----END PGP SIGNATURE-----