Point of View 9800GTX : Intro /Review

[Paracelsus]

Because they want to try see what card can get the extra FPS out it, most play at 1600's or 1920's today.

Doubt it, I reckon most still play at 1280x1024.

 

[LoadsaMoney]

We know one thing at least, they can't put the poor performance down to drivers.

 

[CBP]

I am so dissapointed with these results, but was expecting something like this

 

[helmutcheese]

Doubt it, I reckon most still play at 1280x1024.

Then there is no need for them to buy a high end 8800.

 

[jaykay]

Doubt it, I reckon most still play at 1280x1024.

He means most people who buy such an expensive card play at 1600 or 1920.

 

[LoadsaMoney]

Definately something a miss with those UT3 results, as with the 169.02 drivers, its only 1 frame slower than the GTX at 1600x1200, and its only 3 slower at 1920x1200 with no AA, so god knows how they get about 24fps slower than a GTX at only 1680x1050.

http://www.xbitlabs.com/articles/video/display/leadtek-px8800gt-512mb_12.html#sect1

 

[lay-z-boy]

These results are a bit weird imo.

And why test with *insert res* and then *insert res* with 16x af?

AF?, who doesn't have it at 16x anyway?

4-8X AA please.

 

[Paracelsus]

Then there is no need for them to buy a high end 8800.

I disagree, what if you want to play Crysis on very high with AA/AF? Even an 8800 won't do it smoothly at 1280x1024.

 

[addylad]

You sort of missed my point and that is playable frame rate, res then eye candy. So if say 1440x1050 = playable frame rate you can then give a fair represention of the differential between the cards in the group test.
I have a 24" dell 1920x1200 but to play a game at that res then it has to run at a playable frame rate (33fps) or it's unrealistic.
I dont have a problem with pushing a card to unrealistic settings as a part of the test but to draw any conlusions (or all in this case) on such a flaky test procedure is just plain wrong, its quite clear that the cards are not ment to play this game at such high settings and working the differance out of how poorly they performe on a % basis is... well worst than wrong, its misleading.

+1

 

[Makunouchi]

LOL, the old tech Ultra still the fastest Single card ? (hard to tell there)

Dont drag this up again, Gx2 is a single card.

 

[tbh]

A pointless upgrade for GTX and ultra users then.

I think you mean a pointless downgrade for GTX and Ultra users?

I'm looking for a new card, was very tempted to buy a GTS at the price it is going for but decided to wait a few extra days to see how the 9800GTX shapes up. If these results are accurate I think I'm just going to buy a 8800GTX and be done with it.

 

[jambus]

I dont understand why they test crysis at high res when its so clear that the card hasn't the grunt to run the game at high settings. whats wrong with testing with playable( realistic ) settings instead of quoting unplayable frame rates and giving a % of the speed differance last time i checked 50% of **** all is errm...still **** all!.


Crysis

1680x1050- 0X AA/AF @18 fps

1680x1050- 4X AA/16X AF @12 fps

1920x1200- 0X AA/AF @14 fps

Amen to that one brother!

 

[helmutcheese]

I disagree, what if you want to play Crysis on very high with AA/AF? Even an 8800 won't do it smoothly at 1280x1024.

You are mad enough to buy a new GPU to play 1 buggy game

Oh, and dont forget 1 or 2 badly coded console ports.

You wont be playing Crysis are your above mentioned settings on any card till they bring out Patch 1.99999999.