Benchmarks on average....I got bored....

[JPB]

X1800XT 625/1500
7800 GTX 430/1200




DOOM 3 1024x768 / 1280x1024 / 1600x1200 / Average

X1800 XT 116.2 / 100.6 / 80.4 / 99.06
7800 GTX 114.9 / 106.9 / 91.7 / 104.5

Winner.....TIE
_________________________________________________________________
Far Cry 1.3 1024x768 / 1280x1024 / 1600x1200 / Average

X1800 XT 74.94 / 77.33 / 76.29 / 76.18
7800 GTX 76.55 / 76.93 / 75.82 / 76.43

Winner.....TIE
_________________________________________________________________
Half Life 2 1024x768 / 1280x1024 / 1600x1200 / Average

X1800 XT 50.84 / 53.59 / 52.56 / 52.33
7800 GTX 51.49 / 52.03 / 51.29 / 51.60

Winner.....TIE
_________________________________________________________________
DOOM 3 FSAA 4X + Aniso 8X
1024x768 / 1280x1024 / 1600x1200 / Average

X1800 XT 82.6 / 58.3 / 41.7 / 60.86
7800 GTX 94.7 / 69.6 / 51.8 / 72.03

Winner.....7800 GTX
__________________________________________________________________
Half Life 2 FSAA 4X + Aniso 8X
1024x768 / 1280x1024 / 1600x1200 / Average

X1800 XT 52.82 / 53.40 / 52.12 / 52.78
7800 GTX 51.12 / 51.93 / 51.15 / 51.40

Winner.....TIE
__________________________________________________________________
Composite Figures 3Dmark03 / Game 2 / Game 3 / Nature / Average

X1800 XT 16648 / 138.9 / 102.8 / 106.1 / 115.93
7800 GTX 16066 / 130.7 / 108.4 / 92.3 / 110.46

Winner..... TIE
___________________________________________________________________
Composite Figures 3Dmark05 / Game 1 / Game 2 / Game 3 / Average

X1800 XT 8967 / 37.4 / 27.3 / 45.2 / 36.63
7800 GTX 7618 / 32.3 / 23.4 / 37.4 / 31.03

Winner.....X1800 Xt

________________________

7800GTX............1 WIN
X1800 XT............1 WIN
5 TIES

 

[fierydemise]

I'd call Doom 3 (first set) and 3Dmark03 are ties also when you have near 100fps 5fps either way isn't that special but thats a judgement call, nice work anyway x80064

 

[JPB]

Well, 3Dmark 1st set the XT got a total score of 16648 score...While the GTX got 16066 which would be a tie....In the 2nd set 3D Mark the XT total score was 8967...While the GTX scored a total of 7618. Which is a difference of 1349. A tie in benches...But a XT win in total score.

 

[tuteja1986]

well after reading 5 reviews ;(
I come to conclusion that ATI strenght lies in games with shader 3.0. Also ATI card do good with next gen game like : COD 2/FEAR but it does horrible on Quake 3 engine/Doom 3 engine. Well i want to see a detail review of R520 vs G70 on next gen stuff not test of Doom 3 , farcry or half life 2.

COD 2
Quake 4
Age of empire 3
Elder scroll 4
Fear

Image quality comparison
Power consumption comparison
Overclocking ability

Also 512MB of R520 vs 256MB R520 in fear and call of duty 2

But its not going to happen untill next month when these game come out.

 

[JPB]

Also 512MB of R520 vs 256MB R520 in fear and call of duty 2


I'd like to see this test also..

 

[fierydemise]

I think we'd all like to see the extensive battery of tests tuteja mentioned but considering X1800XT won't be out for another month we may have to wait

 

[JPB]

Power consumption comparison.... http://graphics.tomshardware.com/graphi...enters_the_x1000_promised_land-13.html

 

[nRollo]

Originally posted by: tuteja1986
well after reading 5 reviews ;(
I come to conclusion that ATI strenght lies in games with shader 3.0. Also ATI card do good with next gen game like : COD 2/FEAR but it does horrible on Quake 3 engine/Doom 3 engine. Well i want to see a detail review of R520 vs G70 on next gen stuff not test of Doom 3 , farcry or half life 2.

COD 2
Quake 4
Age of empire 3
Elder scroll 4
Fear

Image quality comparison
Power consumption comparison
Overclocking ability

Also 512MB of R520 vs 256MB R520 in fear and call of duty 2

But its not going to happen untill next month when these game come out.

Errrr Quake 4will be in stores on 10/18 (in 12 days) and FEAR will be in stores 10/11. (in 5 days)

Don't know about the others because I don't care, but Q4 and FEAR are gold with firm release dates.

 

[tuteja1986]

Originally posted by: x80064
Power consumption comparison.... http://graphics.tomshardware.com/graphi...enters_the_x1000_promised_land-13.html

i never trust anandtech testing anymore. they seem to have some benchmark result that are sometime way off to other website benchmarking ;*(
But G70 has less power Power consumption for sure but i want to really know by how much ?

 

[Pabster]

Originally posted by: x80064
Well, 3Dmark 1st set the XT got a total score of 16648 score...While the GTX got 16066 which would be a tie....In the 2nd set 3D Mark the XT total score was 8967...While the GTX scored a total of 7618. Which is a difference of 1349. A tie in benches...But a XT win in total score.

Um, I hate to break it to ya, but actual game benches are superior to 3DMark scores.

 

[boshuter]

Do you own both cards and run these tests yourself? Or is this just more linking to reviews you found online?

You can find a review to back up any claim you want to make..... until someone "respected" actually has both cards and runs some benches it's all crap.

 

[JPB]

Originally posted by: boshuter
Do you own both cards and run these tests yourself? Or is this just more linking to reviews you found online?

You can find a review to back up any claim you want to make..... until someone "respected" actually has both cards and runs some benches it's all crap.

Nope, I dont have BOTH cards. Nobody does. The scores are from different sources ( sites ) on average. I'm trying to figure out what card to go with. I also did not imply that these scores were gonna be real world scores when individuals run their tests. Since there is no X18 XT on shelves yet...I cannot find out what the real world scores are. So until then, this is what I did to HOPEFULLY get a clue about which to purchase. So they may be crap to you, but their not to me. And you say respected? Sure..I very well may not be.....But how respectful is it to crap on someone else's thread?

 

[ZobarStyl]

All those scores appear to be without AA/AF, which is pretty absurd for cards this powerful. When you move to 20x15 resolution or 16x12 full AA/AF, the numbers actual start to show up. The x1800XT does better with AA/AF at these rezzes (it has a lot more memory bandwidth) but still can't run OGL games worth a damn. I'd look up those kind of numbers for comparison, instead of using CPU-limited numbers.

 

[JPB]

DOOM 3 FSAA 4X + Aniso 8X
1024x768 / 1280x1024 / 1600x1200 / Average

X1800 XT 82.6 / 58.3 / 41.7 / 60.86
7800 GTX 94.7 / 69.6 / 51.8 / 72.03

Winner.....7800 GTX
__________________________________________________________________
Half Life 2 FSAA 4X + Aniso 8X
1024x768 / 1280x1024 / 1600x1200 / Average

X1800 XT 52.82 / 53.40 / 52.12 / 52.78
7800 GTX 51.12 / 51.93 / 51.15 / 51.40

Winner.....TIE

 

[Insomniak]

Thank God someone put this up. Finally people will get the picture.

 

[tuteja1986]

Originally posted by: Insomniak
Thank God someone put this up. Finally people will get the picture.

give rollo a break. too much rollo bashing these days "lol". Anyways i saw few of his post today they weren't bashing ATI and i was srupprised :Q

But rollo doesn't seem to be hardcore gamer since he didn't care about AOE 3 and elderscroll 4. elderscroll should be best looking game of this year i think. it looks better than fear. aoe 3 , cod 2 and even quake 4. well i think that game will really put G70 and r520 to the limit and also be the best RPG of this year.

 

[deadseasquirrel]

Originally posted by: tuteja1986

Originally posted by: Insomniak
Thank God someone put this up. Finally people will get the picture.

give rollo a break. too much rollo bashing these days "lol". Anyways i saw few of his post today they weren't bashing ATI and i was srupprised :Q

But rollo doesn't seem to be hardcore gamer since he didn't care about AOE 3 and elderscroll 4. elderscroll should be best looking game of this year i think. it looks better than fear. aoe 3 , cod 2 and even quake 4. well i think that game will really put G70 and r520 to the limit and also be the best RPG of this year.

????
How was anyone bashing Rollo?
And to say someone isn't a hardcore gamer because they have no interest in 2 games is silly. Some people just aren't into RTS or RPG.
I was very interested in AOE3 until I played the demo. So I guess I'm not much of a gamer either since the only game that sounds good to me of the games you listed is Oblivion. In the mean time, Civ 4 will be out this month (I hope) and that will keep me plenty busy until Oblivion arrives.

 

[milomnderbnder21]

Err, while I appreciate the collection of scores and whatnot, and dually recognize that the GTX does surpass the X1800 in some places and is close in most, these scores on average don't seem in line with what I have been reading. Doom 3, GTX is a clear winner, but in most games, especially with AF and AA, the X1800 is a fairly clear winner, by varying margins.

Basically, I want to know what sources you used.

 

[Pete]

Originally posted by: x80064
Also 512MB of R520 vs 256MB R520 in fear and call of duty 2

I'd like to see this test also..

Given how the XL compares to the GT, it would seem the extra RAM helps. I'm looking forward to a reviewer downclocking the XT to XL speeds to see exactly how much the extra RAM helps.

 

[Ronin]

If I receive a card anytime soon, I'll be happy to run the same tests.

I wish I could tell you guys I have one, but I don't.

 

[Suspicious-Teach8788]

These are mostly low res benchies. We need 1280, 1600 and 2048 as references. Then AA on and HDR on too. Bench with those on and off and diff combos and see how it turns out. I think there should be more variation between teh cards...

 

[Ackmed]

Sorry, but 3 games and one synthetic bench doesnt make a good numbers to compare to. Especially when there is no link to verify anything...

 

[DidlySquat]

Originally posted by: ZobarStyl
All those scores appear to be without AA/AF, which is pretty absurd for cards this powerful. When you move to 20x15 resolution or 16x12 full AA/AF, the numbers actual start to show up. The x1800XT does better with AA/AF at these rezzes (it has a lot more memory bandwidth) but still can't run OGL games worth a damn. I'd look up those kind of numbers for comparison, instead of using CPU-limited numbers.

agreed. The lower resolution 1024x768 and even 1280x1024 are not relevant because people that buy these cards for $500 are likely to own high end monitors and run their games at 1600x1200 or above. I know I'd never even get my 7800GTX if I planned to run my games at 1280x1024 or below.

 

[lifeguard1999]

Personally, I would like to see a lot more benchmarks. Xbitlabs tends to do a good job producing tons of them. One thing I would like to see is NVidia's 80.xx WHQL drivers (not yet available) with a dual core system vs. ATI's WHQL drivers (not yet available) for x1800 series.

 

[monster64]

Nothing really changed with the X1800XT. Same thing as 6800 Ultra vs X850XT. ATI has its strong games and so does nvidia. Also, you can't really compare a 512MB card vs a 256MB card, as it's just not fair, especially in games like FEAR and COD, and at resolutions above 1600x1200 in general. Of course a 512MB card will win those as the games will use well over 256MB of vram. Why do you think the 512MB x1800XT was about 100% better in FEAR and COD2? Because its super special and extremely good? No, because of the 512MB. When we see a 512MB 7800 GTX, which will be soon, it will be a fair benchmark.