Anandtech Benchmarks Out!

[Tido14]

http://www.anandtech.com/video/showdoc.html?i=1863

 

[d8cam]

SWEET THANKS MAN!

 

[Tork]

cool , they're so lucky. they can at least get a taste of hl2

edit: Reading now!

 

[d8cam]

Again it seems ATI wins in their tests. Nvidia didn't fare bad...but ATI just did better.

 

[ValVed RaY]

looks like nvidia is the official "teh suck"

sorry nvidia owners, nothing personal.

/me whistles a chipper tune

 

[=)PoLo(=]

That's one of the best reviews I've seen yet. Also interesting to see the ti4600 performing so well. It's a directX 8.0 card and you'll miss out on some eyecandy, but still that's not bad..

 

[Apos]

This is just yet another recounting of the same shader day event. I wish valve would just release the benchmarking software to the big review sites, even if they cant release it to us.

 

[Apos]

I take that back: they have real benchmarks!

 

[Anwar]

nvidia us faster in doom 3 its on that same link

 

[shapeshifter]

WHO THE FU*K cares ABOUT DOOM FING 3!?! were talking about HALF LIFE 2 Benchmarks what DON'T YOU UNDERSTAND?

 

[undertoe]

im impressed with the performance of the 9600 PRO, great value indeed

 

[Javert]

If ATI raises the price of their 9600 because of this (not saying they would, but who knows), then there will be serious violence on my part...:flame:
Nvidia are still resolving driver issues, they have yet to play their final card, but most likely, only the 5900 Ultra will outperform the 9600.

 

[Detharin]

Curious, i noticed that in Nvidias letter they stated that the press has had 50.xx drivers for over two weeks, but in this article we see it stated that valve INSISTED they use only commercially available drivers. My question is this, why would valve care if they did a seperate test using nvidias newest drivers which have been available to the people doing the benchmarks? Would it really matter if there was something along the lines of "with 45.33 drivers" "with 50.xx drivers*" "* fog disabled by drivers."

 

[Yatta]

Originally posted by shapeshifter
WHO THE FU*K cares ABOUT DOOM FING 3!?! were talking about HALF LIFE 2 Benchmarks what DON'T YOU UNDERSTAND?

Hold your pennises!

 

[shapeshifter]

duh if they use drivers that "cheat" or as nvidia likes to call them "remove the nice looking things to get more fps" aka optimazations, they will get more fps, which isnt fair becuase well they would be CHEATING.

 

[Tyl3n0L]

Very nice preview by Anand.
I wonder why he don't have any FSAA/AF results by the way?

 

[shapeshifter]

(from rage3d) owned thread.

 

[Banana]

Originally posted by Detharin
Curious, i noticed that in Nvidias letter they stated that the press has had 50.xx drivers for over two weeks, but in this article we see it stated that valve INSISTED they use only commercially available drivers. My question is this, why would valve care if they did a seperate test using nvidias newest drivers which have been available to the people doing the benchmarks? Would it really matter if there was something along the lines of "with 45.33 drivers" "with 50.xx drivers*" "* fog disabled by drivers."

My guess is Valve wants the benchmarks to run the way they will run for end users. End users dont currently have access to any of the rel 50 drivers so they shouldnt be tested. I think i read somewhere (i'm sure it was on one of these tech sites but i cant remember which at the moment) that someone stated that the rel 50 drivers that these websites have were never going to be released to the public, they were only produced to make NVIDIA hardware pass the benchmarks. I dunno if that is true or not.

I still hold that NVIDIA should just cut their losses and push their next card out as fast as possible. Though i am worried that it wont matter. It seems like NVIDIA is trying to promote a certain way of engineering their cards and it doesnt seem to be working. What if their next card is plagued by the same crap.

 

[dscowboy]

Originally posted by Banana
I think i read somewhere (i'm sure it was on one of these tech sites but i cant remember which at the moment) that someone stated that the rel 50 drivers that these websites have were never going to be released to the public, they were only produced to make NVIDIA hardware pass the benchmarks. I dunno if that is true or not.

This is similar to what Gabe's slide said in the presentation: That NV had provided Valve with a build of the detonator drivers that was never intended to ship, just so Valve would use it for benchmarks and inflate NV's numbers.

EDIT: spelling