9600GT rumored release date Feb 21 - any Linux driver news?

[deanjo]

Really odd that nobody on nvnews forums are reporting the same findings.

[Michael]

Really odd that nobody on nvnews forums are reporting the same findings.

Well, how many 9600GT owners do you think actually own Radeon HD 3850/3870 cards?

[deanjo]

Well, how many 9600GT owners do you think actually own Radeon HD 3850/3870 cards?

What does that have to do with it? You don't need another card to compare too when it comes to poor performance in games. People scream in there when glxgears drops by 5 frames.

[Michael]

What does that have to do with it? You don't need another card to compare too when it comes to poor performance in games.

By having something to compare it to quantitatively.

[deanjo]

Most of the owners of the 9600 would be on their 3/4/5th card already. They would know how the performance is comparing them. As they have for every generation of card that nvidia has put out. A new card comes out and it doesn't perform well they always holler. What would make the 9600 immune to that?

[Thetargos]

Well, these results may indicate one thing: That the AMD drivers are better optimized for their high-end graphics than nVidia's. Just like you said in the article, Michael, nvidia had similar problems with the 8 series. So probably what actually is showing in the results are optimizations in the fglrx that are not yet completely present in the nvidia driver.

Now, as why wouldn't these findings be all over nvnews, there are a lot of reasons, for example not many users having the cards yet. I do, however, think that a comparison in Linux of the 9600GT against nvidia hardware would be indeed interesting. For example, have it tested against the 8600GTS and 8800GT (and maybe throw in the 3870 for reference).