2900XT > 8800GTS 640MB

[drakore]

Check this out:

http://www.fudzilla.com/index.php?op...744&Ite mid=1

Pretty impressive

 

[mazeroth]

Very interesting...

A side note, drakore. You're only getting 2.79 ghz. out of your e4300 on water? I was able to hit 3.25 without upping my voltages at all and upped them to 1.38 to hit 3.3ghz with the stock cooling. You should definately be able to get higher than that, I would think.

 

[Kromis]

Originally posted by: mazeroth
Very interesting...

A side note, drakore. You're only getting 2.79 ghz. out of your e4300 on water? I was able to hit 3.25 without upping my voltages at all and upped them to 1.38 to hit 3.3ghz with the stock cooling. You should definately be able to get higher than that, I would think.

Heh, I can't even achieve that. Only 2.8 Ghz @ 1.34

 

[apoppin]

Pretty impressive

not really ... depends how they tested ... and how well Cats run with Vista vs FW in Vista
-and i believe the results are reversed with XP ...

AMD has a lead here by over 3,000 points in the default test in 3DMark 06

that is simply ridiculous ... AMD is optimized better for 3DMark ... my HD2900xt scores over 1000 more 3dMarks in 06 than my GTS ... it doesn't perform 10% better overall with the latest cats [period]

the ONLY thing i look at in 3DMark06 is to track changes in my rig and to roughly compare with other GTSes to make sure i ma in the 'ballpark' and that there is no serious inconsitency.

--and the next 7.7 drivers ... well, they are still future for the regular user

 

[Extelleron]

Originally posted by: apoppin
<div class="FTQUOTE"><begin quote>Pretty impressive</end quote></div>
not really ... depends how they tested ... and how well Cats run with Vista vs FW in Vista
-and i believe the results are reversed with XP ...
<div class="FTQUOTE"><begin quote>AMD has a lead here by over 3,000 points in the default test in 3DMark 06</end quote></div>that is simply ridiculous ... AMD is optimized better for 3DMark ... my HD2900xt scores over 1000 more 3dMarks in 06 than my GTS ... it doesn't perform 10% better overall with the latest cats [period]

the ONLY thing i look at in 3DMark06 is to track changes in my rig and to roughly compare with other GTSes to make sure i ma in the 'ballpark' and that there is no serious inconsitency.

--and the next 7.7 drivers ... well, they are still future for the regular user

I don't care about 3D Mark 06 either, it's this that matters:

CoH, 2560x1600: 84 FPS vs 32 FPS
Prey, 2560x1600: 107 FPS vs 65 FPS

 

[apoppin]

what you have to realise is that this is benchmarking and may have nothing to do with real world results or "play-ability"

a single 'hitch' ... a tiny stumble usually caused by immature drivers at this point - for just an almost imperceptible instant will drop the Minimum to a ridiculous low ... by that same token - since the architectures are SO different in the WAY they render, one GPU may render a brief 1/2 sec at 3,000 FPS ... guess what happens to your average and your maximum ?

i am also getting these inconsistencies in the benchmarking and UNLESS you are going to pull the FPS graph of the demo, you'd never even notice it [conveniently, if you want to 'prove' something] ... i believe this is partly why the HW sites are getting such variability in their results.

here's how Derek Wilson put it [better] ... from Key's 'in house' thread:

If I have a bench that's 100 seconds long and a /single/ frame gets rendered at 5000 fps (say, cause you're looking at a sky that can render at 5000 fps for one frame), it's not going to matter and I don't care. Lets say that arbitrarily this demo ended up getting an average of 40 fps. (generating a total of 4000 frames in 100 seconds).

But ...

what if, instead of one instantaneous frame that no one cares about, you pause on the sky for even 1 second. That will insert 5000 more frames into your benchmark, giving you an average of 90 fps rather than 40. And all this boost comes from 1/100th of the time you spend doing the benchmark.

yes my numbers are exaggerated for demonstration (though the oblivion pause screen can render at 3k+ fps on some cards) ...

but while in the real world we don't see instances where the difference is >2x performance, we do see real skew in the averages because of things like this.

he also says he is working to show "something additional" in their upcoming AT articles ... and i am certainly looking forward to it

 

[ArchAngel777]

Yeah, i am with Apop.

Even minimum framerates don't tell us much, as he said if the framerates drops for a tiny millsecond and then is blazing fast through the rest of the second, then it reports the average frames for that particular second. The problem is, it looks like hell, even though min framerate is 40fps --

With that said, IMO if I had to chose between what three results Min, Avg, Max, I would chose min as prefered, since you want to know how a card performs in the word case scenarios. It isn't perfect to use min, but it is as close as we can get at this point.

 

[apoppin]

It isn't perfect to use min, but it is as close as we can get at this point.

You also have to remember that the "minimum FPS" - especially in STALKER is dependent also on the CPU as Keys and i are finding out
... the best way - if you can't get two and do it yourself - is to get as many reliable benches together and try to put together a picture of the respective GPUs

i don't think we will ever hear the "last word" on G80 vs r600 ... any more than someone can pick audio components on benchmarks, reviews or tests

 

[yacoub]

Good maybe this will make the 640MB GTS come down in price to where it should be - $300. Because for $300 I'd totally take the 640MB GTS and get a quieter, less-of-my-PSU using GPU with 140MB more memory.

 

[Keysplayr]

What am I missing here? Anything in particular I should be looking for when I click the link in the first post? It just leads to a thousand articles of varying topic.

 

[apoppin]

http://www.fudzilla.com/index....=view&id=1759&Itemid=1

the OP is a lazy linker