PDA

View Full Version : 52.14 Benchmarks...


Pages : [1] 2

dnottis
10-01-03, 12:56 PM
I was checking out some of the new 9800 XT previews and noticed the 5900 ultra kicking some ass in the Anandtech review. After checking the driver revision I found this...

"We used ATI’s publicly available Catalyst 3.7 drivers and in order to support the NV38 we used NVIDIA’s forthcoming 52.14 drivers. The 52.14 drivers apparently have issues in two games, neither of which are featured in our test suite (Half Life 2 & Gunmetal)."

And regarding Aquamark 3 testing...
"Image quality appears to have improved for NVIDIA in this benchmark over what has been reported of previous drivers, and the NV38 handled the massive overdraw portion of the test the smoothest of all the cards."

Now we just have to wait for these drivers to "leak" or for nvidia to finish them..gah!

http://www.anandtech.com/video/show...?i=1890&p=5

Nutty
10-01-03, 02:41 PM
Sounding good to me soo far.. provided those problems are fixed..

theultimo
10-01-03, 02:45 PM
I dont know, but I'm hoping IQ is better than 51.75's
EDIT Link down?

Nutty
10-01-03, 02:46 PM
And regarding Aquamark 3 testing...
"Image quality appears to have improved for NVIDIA in this benchmark over what has been reported of previous drivers,

Did you forget to read this bit?

theultimo
10-01-03, 02:48 PM
Whoops, misread post bah :\

dnottis
10-01-03, 02:58 PM
The IQ improvement was encouraging. Scores were very good and IQ is improved...sounds like nVidia may pull this one off....just leak us a beta already!! :angel:

StealthHawk
10-01-03, 05:36 PM
Originally posted by dnottis
"We used ATI’s publicly available Catalyst 3.7 drivers and in order to support the NV38 we used NVIDIA’s forthcoming 52.14 drivers. The 52.14 drivers apparently have issues in two games, neither of which are featured in our test suite (Half Life 2 & Gunmetal)."

Why is he beating around the bush? Either they have issues or they don't. Notice how he doesn't say what the issues are, but at the conclusion talks about the speed of HL2 being greatly improved by the drivers. Why didn't Anand reiterate that the drivers "apparently" have problems in HL2 at the same time instead of hiding it in the front?

And regarding Aquamark 3 testing...
"Image quality appears to have improved for NVIDIA in this benchmark over what has been reported of previous drivers, and the NV38 handled the massive overdraw portion of the test the smoothest of all the cards."

Reported by who? Did Anandtech look into the issue with previous drivers? What does he mean it "appears" to have improved. Why didn't he look into this?


If you're going to use unreleased drivers you better damn well investigate how they affect your benchmark suite :rolleyes: Looks like Anand cares more about having an exclusive than about providing correct, factual information to his readers. Then again, I'm not sure I trust Anand's eyes. This is the guy who didn't realize FSAA caused visual problems in Splinter Cell and said it worked fine on NVIDIA cards(drivers hardcoded to not enalbe FSAA), did he even play the game :rolleyes:

Tsunami
10-02-03, 06:58 AM
Originally posted by dnottis
The IQ improvement was encouraging. Scores were very good and IQ is improved...sounds like nVidia may pull this one off....just leak us a beta already!! :angel:
I hope so but my NV35 performs great anyway so with the det 50 boost........ should be interesting:eek: as long as Gunmetal and hl2 problems are both fixed

windwaver
10-05-03, 11:15 AM
Sounds good!

dexiter
10-05-03, 08:25 PM
Originally posted by StealthHawk
Why is he beating around the bush? Either they have issues or they don't. Notice how he doesn't say what the issues are, but at the conclusion talks about the speed of HL2 being greatly improved by the drivers. Why didn't Anand reiterate that the drivers "apparently" have problems in HL2 at the same time instead of hiding it in the front?



Reported by who? Did Anandtech look into the issue with previous drivers? What does he mean it "appears" to have improved. Why didn't he look into this?


If you're going to use unreleased drivers you better damn well investigate how they affect your benchmark suite :rolleyes: Looks like Anand cares more about having an exclusive than about providing correct, factual information to his readers. Then again, I'm not sure I trust Anand's eyes. This is the guy who didn't realize FSAA caused visual problems in Splinter Cell and said it worked fine on NVIDIA cards(drivers hardcoded to not enalbe FSAA), did he even play the game :rolleyes:

If you check out that review, it's titled "Preview of.. Part I" He will supposedly answer all those IQ questions and issues with 52.14 in Part II, which should be coming out in the next couple of days. Why did he leave all those unsubstantiated comments in his article? I don't know, maybe Anand wanted people to look forward to Part II, maybe he didn't realize he should've thought twice about writing anything positive about NVIDIA, maybe NVIDIA paid him off to write those comments so that.. I dunno, keep NV fanboys excited and ATI fanboys enraged for a few days? Patience.. Much ado about nada at this point. :baaa::

nfhq|matrix
10-06-03, 05:51 AM
so far I haven't been able to find any serious visual differences between the two using the CAT3.7s and the 52.14 drivers but we'll be taking screenshots to make sure.

from anandtechs latest weblog

http://www.anandtech.com/weblog/index.html

Tsunami
10-06-03, 09:56 AM
Wow this is really great news looks like NV problems were software related. I know people are gonna come in here saying NV payed them off and they are cheating so i though i'd add quickly a well done to Nvidia:eek: :D

mrsabidji
10-06-03, 10:14 AM
Hmmm... I'm glad it can be fixed too. But to my understanding, although they can be fixed via drivers, nv DirectX9 related problems are due to the fact nv3x GPUs don't have a fully DX9 compliant design. So I think it's still hardware related.

mrsabidji

halduemilauno
10-06-03, 10:42 AM
I've sent 12 post. now there is only one. Where are the others???
thanks, goodbye, good luck at all!

Tsunami
10-06-03, 11:02 AM
Originally posted by mrsabidji
Hmmm... I'm glad it can be fixed too. But to my understanding, although they can be fixed via drivers, nv DirectX9 related problems are due to the fact nv3x GPUs don't have a fully DX9 compliant design. So I think it's still hardware related.

mrsabidji
yes but i thought (driver enabling) that if they could create greater performance through driver updates then they could make the card more dx9 compliant (although not fully. Maybe pixel shader 2.0 will be on by default but who knows:confused:

Tsunami
10-06-03, 11:06 AM
and i don't think that nvidia with all their experience would have made the card knowing that it would never perform DX9 well. I have the box here in my hand and it sais t supports DX9 so i think they must have a fall back method:angel:

The Baron
10-06-03, 11:12 AM
Originally posted by Tsunami
and i don't think that nvidia with all their experience would have made the card knowing that it would never perform DX9 well. I have the box here in my hand and it sais t supports DX9 so i think they must have a fall back method:angel:
um, GF3 runs PS1.1 shaders like ass, but it's DX8-compliant. supporting something does not mean performing well.

Paul
10-06-03, 11:19 AM
Originally posted by The Baron
um, GF3 runs PS1.1 shaders like ass, but it's DX8-compliant. supporting something does not mean performing well.

Hush now.

digitalwanderer
10-06-03, 11:25 AM
Originally posted by Tsunami
and i don't think that nvidia with all their experience would have made the card knowing that it would never perform DX9 well. I have the box here in my hand and it sais t supports DX9 so i think they must have a fall back method
You would think it's just an incredibly stupid move that no graphic card company in their right mind could possibly make, wouldn't you? I sure do.

But alass, they did. :(

ragejg
10-06-03, 11:35 AM
Originally posted by The Baron
um, GF3 runs PS1.1 shaders like ass, but it's DX8-compliant. supporting something does not mean performing well.

IMO it didn't/doesn't run PS 1.1 bad for what generation of hardware it is...

StealthHawk
10-06-03, 03:33 PM
Originally posted by The Baron
um, GF3 runs PS1.1 shaders like ass, but it's DX8-compliant. supporting something does not mean performing well. Stop using logic, it's unfair.

The Baron
10-06-03, 03:45 PM
Originally posted by StealthHawk
Stop using logic, it's unfair.
What? I'm sorry I cheated and took the easy way out. :)

Hellbinder
10-06-03, 06:52 PM
ill believe it when i see publically released drivers benched and examined by multiple Sites.

AS for this guy "not seeing any serious iq differences" Speaks for itself. They are obviously there he just "doesnt think they are serious"... Not suprising at all comming from these people.

Well see what Techreport, B3D and even NVnews has to say.

Hellbinder
10-06-03, 06:54 PM
BTW.. It is already known that Nvidia Drivers fudge Screen shots to make them look better than the screen output.

I say one more time he is clearly saying that he sees "diffferences" they are just not "Serious" to him.

:rolleyes:

The Baron
10-06-03, 07:17 PM
Originally posted by Hellbinder
iWell see what Techreport, B3D and even NVnews has to say.
Ho, ho, ho, it's magic, you know.

Originally posted by Hellbinder
BTW.. It is already known that Nvidia Drivers fudge Screen shots to make them look better than the screen output.

Gotta call you on this one. It's NOT known that NVIDIA drivers contain a screenshot hack--it was inferred by Gabe at the Shader Day presentation, but it has not been confirmed as far as I am aware.