To read all comments associated with this story, please click here.
Yeah, the radeon rivers need to be rethought and reimplemented. Much of the code in the radeon drivers are based off of what ati had given the oss community a long time ago. I think starting from scratch and doing clean room drivers would be the better solution. This article doesn't speak on performance though.
Are you kidding? You're really comparing a real-world working driver which supports pretty much every application on a large gamut of video cards released since a least 2000 with a new incomplete driver which has problem even switching resolution in a fairly common card and can't really do more than display glxgears?
Why should one throw out years of development for a "clean room" work which couldn't possibly give comparable results in years? The nouveau project is a good thing and deserves attention but, please, keep your feet on the ground and don't buy all the hype.
Actually, there *is* a project similar to this for AMD / ATI graphics cards. It is called avivo and it is located here:
http://gitweb.freedesktop.org/?p=avivo/xf86-video-avivo.git;a=summa...
They are currently trying to get acceleration for the r500 based cards working like they do with the r300 based cards. Long live open source!
Edit:
Adding link talking about this driver a bit. Not everyone wants to just read through code:
http://www.phoronix.com/scan.php?page=article&item=753&num=1
Edited 2007-07-01 05:48







Member since:
2006-01-03
As I remember, when this project started, i had a strange feeling about it's sucess. But as i saw later work, Noveau looks very promising. Now we need similar project with AMD.
Browser: Opera/9.50 (J2ME/MIDP; Opera Mini/4.0.8462/20; U; en)