Path: chuka.playstation.co.uk!news From: "Colin Hughes" Newsgroups: scee.yaroze.programming.2d_graphics Subject: Re: Caches Date: Tue, 18 Aug 1998 11:55:06 +0100 Organization: PlayStation Net Yaroze (SCEE) Lines: 32 Message-ID: <6rbn5f$ea917@chuka.playstation.co.uk> References: <35C81A30.4ACB@club-internet.fr> <35C82DDB.3E12E37C@scee.sony.co.uk> <01bdc061$2c186980$f30b0a0a@Angela1.intelligent-group.com> <35D2E01D.5CA1@mdx.ac.uk> <6r0vps$n7i6@chuka.playstation.co.uk> <35D43404.282A@mdx.ac.uk> <35D438D2.908C3316@ndirect.co.uk> <35D43F08.36BA@mdx.ac.uk> <35D44CBE.1FA95769@ndirect.co.uk> <35D46843.5DF5@mdx.ac.uk> <35D83E4D.32BA4E05@scee.sony.co.uk> <35D93FB4.6345@mdx.ac.uk> NNTP-Posting-Host: mailgate.scee.sony.co.uk X-Newsreader: Microsoft Outlook Express 4.72.2106.4 X-MimeOLE: Produced By Microsoft MimeOLE V4.72.2106.4 More about caches.... On the GPU a cache line is 4 words long..... This has a big hit on textures that don't fit on cache. If you draw a 256x256 texture at 90 degrees to normal, Every time a pixel is needed a new cache line will be read, ( 16 pixels worth for a 4 bit texture ) Subdividing the poly into 64x64 sections will speed up this case to the point where 1 cache line is read for every 16 pixel writes ( For 4 bits ) Cheers, Colin. Robert Swan wrote in message <35D93FB4.6345@mdx.ac.uk>... >James Russell wrote: > >> The GPU has an internal CLUT as well, so that's > >Really? sounds cool, as we've been wondering how the cluts got >processed. Now you say dont change cluts... how do the rgb scalars >affect this? is it quicker to change cluts (which dont require scaling) >or to scale the clut cache?? > >cool, something else I can get bewildered about > >Rob