• Some users have recently had their accounts hijacked. It seems that the now defunct EVGA forums might have compromised your password there and seems many are using the same PW here. We would suggest you UPDATE YOUR PASSWORD and TURN ON 2FA for your account here to further secure it. None of the compromised accounts had 2FA turned on.
    Once you have enabled 2FA, your account will be updated soon to show a badge, letting other members know that you use 2FA to protect your account. This should be beneficial for everyone that uses FSFT.

HD 2900XT Oblivion Performance

B

Blackstone

Supreme [H]ardness
2FA
Joined
Mar 8, 2007
Messages
4,098
I'm looking for feedback from users who have purchased a 2900XT as to how well it performs with Oblivion.

Please tell us what settings you are using, what resolution, ect.

I'd really like to know how this card handles 1680 x 1050 4xaa with Oblivion, outdoors, hdr, max everything.
 
Blackstone said:
I'm looking for feedback from users who have purchased a 2900XT as to how well it performs with Oblivion.

Please tell us what settings you are using, what resolution, ect.

I'd really like to know how this card handles 1680 x 1050 4xaa with Oblivion, outdoors, hdr, max everything.
Click to expand...

Check out the hardocp article on the 2900xt. I believe it has oblivion as a benchmark.
 
Tzzird said:
Check out the hardocp article on the 2900xt. I believe it has oblivion as a benchmark.
Click to expand...

I know I want to hear from actual users though as well. I'm trying to figure out why it beats the GTS on Tom's Hardware but not on [H].
 
Blackstone said:
I know I want to hear from actual users though as well. I'm trying to figure out why it beats the GTS on Tom's Hardware but not on [H].
Click to expand...

Where's this link? I hope you're not talking about the toms hardware charts.
 
Blackstone said:
http://www.tomshardware.com/2007/05...ardware_from_ati/page21.html#oblivion__xp_pro
Click to expand...

The difference is in quality settings. In the hardocp review, they ran oblivion at 8xAA with the 8800gtx and 2xAA with both the 8800gts 640mb and the 2900xt. For the two lower cards, the 8800GTS outperformed the 2900xt, allowing slightly higher detail settings for a slightly higher fps average. For the specific game of oblivion, the ATI card seems to beat the 8800gts 640mb when AA is turned OFF.

The guys at Hard do a different comparison model than other sites. You kinda have to get used to it, but it does provide very useful information.
 
Blackstone said:
I'm looking for feedback from users who have purchased a 2900XT as to how well it performs with Oblivion.

Please tell us what settings you are using, what resolution, ect.

I'd really like to know how this card handles 1680 x 1050 4xaa with Oblivion, outdoors, hdr, max everything.
Click to expand...

Based off what you said in your original post and also the results of the benchmarks, you're probably better off going with the 8800gts 640mb, and if you can afford it, the 8800gtx. The 8800gtx is around $150 more than the 2900xt right now, but it can offer you 8xAA at 50fps with everything at full.

Both the gts and the 2900 will have to run 2x just to get around 40fps, not to mention the lowering of quality settings.
 
Tzzird said:
The difference is in quality settings. In the hardocp review, they ran oblivion at 8xAA with the 8800gtx and 2xAA with both the 8800gts 640mb and the 2900xt. For the two lower cards, the 8800GTS outperformed the 2900xt, allowing slightly higher detail settings for a slightly higher fps average. For the specific game of oblivion, the ATI card seems to beat the 8800gts 640mb when AA is turned OFF.

The guys at Hard do a different comparison model than other sites. You kinda have to get used to it, but it does provide very useful information.
Click to expand...

well is it not obvious something extremely bizarre is afoot?
basically H is saying that the 2900xt at lower settings performs worse than the gts 640 at higher settings and considerably worse than the gtx at full max settings...
meanwhile the two sites other people just linked [tomshardware and techreport] using apples to apples comparisons show the 2900 winning against the 640 and right below the gtx.
what's going on?
 
ninjaspa said:
well is it not obvious something extremely bizarre is afoot?
basically H is saying that the 2900xt at lower settings performs worse than the gts 640 at higher settings and considerably worse than the gtx at full max settings...
meanwhile the two sites other people just linked [tomshardware and techreport] using apples to apples comparisons show the 2900 winning against the 640 and right below the gtx.
what's going on?
Click to expand...

Could have something to do with the fact that techreport claims to be usings 8.37.4 drivers while other sites were using 8.37.2.
 
ah, so drivers are still helping eh? so many people have been shrugging that off though

and the other thing i wondered was if any publications got a hold of the 8.38 driver which people who got the cards ahead of the time claim to have nabbed somehow or is this actually the 'alpha'?
 
just buy it from somewhere with a return policy. it wont be the end of the world if you dont like and have to return it you know.
 
the 8.38 drivers haven't been benched yet . The newest drivers or alpha drivers are the 8.37.4 (or whatever)

If your looking to buy a video card and can wait 4-6 weeks I personaly would . At the rate ati is releasing beta drivers they will be much further ahead and we should have a much better view of the 2900xt's performance

http://www.hardware.info/nl-NL/articles/amdnZGpqZGia/ATI_Radeon_HD_2900_XT_test/1

in other games even simply switching to an intel mobo results in the radeon kicking the gts's back side.

Hopefully hardocp tests on two platforms next time with newer drivers. Hopefully we can see one at the end of the month or at the very least when the rest of the product lineup launches
 
gljvd said:
the 8.38 drivers haven't been benched yet . The newest drivers or alpha drivers are the 8.37.4 (or whatever)

If your looking to buy a video card and can wait 4-6 weeks I personaly would . At the rate ati is releasing beta drivers they will be much further ahead and we should have a much better view of the 2900xt's performance

http://www.hardware.info/nl-NL/articles/amdnZGpqZGia/ATI_Radeon_HD_2900_XT_test/1

in other games even simply switching to an intel mobo results in the radeon kicking the gts's back side.

Hopefully hardocp tests on two platforms next time with newer drivers. Hopefully we can see one at the end of the month or at the very least when the rest of the product lineup launches
Click to expand...

Im getting the impression non USA review sites are getting different drivers. Based on the above review and others from european sites, the 2900 seems awesome! :eek: :confused: :)
 
Tzzird said:
Based off what you said in your original post and also the results of the benchmarks, you're probably better off going with the 8800gts 640mb, and if you can afford it, the 8800gtx.
Click to expand...
If you look closely at the TechReport site you'll see that in the Oblivion 1920x1200 4AA/16AF test it is the XT edging out an OC GTS640. Well not really edging out, basically matching it. As I suggested to Blackstone in the main Video forum; time to wait on the next major driver release to see exactly WTF is going on.

P.S. Please spare the "they test things different" at HardOCP parroting drivel because if that is the reason for the difference you might as well flush this "different" test down the toilet....which I happen to think would silly. It looks a lot more like just a case of the drivers distributed a couple of days before the actual release date have some really bad performance issues.
 
The R600 as it seems get a large performance hit with AF settings while competition and former gen does not.

Is this a driver problem and will it get solved in future Drivers.
 
SuperGee said:
The R600 as it seems get a large performance hit with AF settings while competition and former gen does not.
Click to expand...
From what I've gathered:
AF doesn't have much of an effect on performance. However AA has been the problem. With the 2900XT AA is now fully programmable (this'll have to come for NVidia in the future sometime too, ATI is just ahead of the curve here). To do the calculations for those AA algorithms the 2900XT GPU uses the same stream processors (those 320 shaders) that are also used for the initial rendering. So if your algorithm (which can be set from the driver, and in the future even by the application) is inefficent or you just pour on the AA the rendering (AKA FPS) ends up fighting for use of the stream processors.
 
http://www.hardware.info/





benchmark


benchmark


benchmark


benchmark
 
Dave Orton should personally be whipping those poor driver coders in their dungeon. With all the odd benchmarks where X1950 cards are beating 2900XT, it just screams for driver failure. From what I have hear, the drivers are stable but the performance as we all see, is somewhat a rollercoaster ride.
 
sam0t said:
Dave Orton should personally be whipping those poor driver coders in their dungeon. With all the odd benchmarks where X1950 cards are beating 2900XT, it just screams for driver failure. From what I have hear, the drivers are stable but the performance as we all see, is somewhat a rollercoaster ride.
Click to expand...
Stable >> fast. For two reasons. First the FPS of a BSOD sucks. :p Second generally speaking if software is stable you have a much better grasp of what it is doing so it is easier to work with and go forward from. A solid base to build off of. It is notoriously difficult to retrofit stability into software, especially something as complicated a system as a driver.
 
You must log in or register to reply here.
Back
Top