Gamespot benchmark review BIOSHOCK

If 2900xt can beat 8800gtx in XP it can surely beat it with Vista if the drivers are optimized.
 
Wow, Ati Spanks every equivalent nVidia in these test. The 1950pro spanks the shit out of the 7900gs and 8600GTS, hell the 1650 is almost beside the 7900GS. Hell the 1900xtx is right behind the 8800GTS.. Hope they do a MOH:A set as well.

~4 FPS difference = squash?

Isn't the 512 2900 more of a competitor to the GTS 640, then the GTX?
 
Wow, Ati Spanks every equivalent nVidia in these test. The 1950pro spanks the shit out of the 7900gs and 8600GTS, hell the 1650 is almost beside the 7900GS. Hell the 1900xtx is right behind the 8800GTS.. Hope they do a MOH:A set as well.



Isn't the 512 2900 more of a competitor to the GTS 640, then the GTX?

Now I regret trading my 1900gt for a 8600gts. :( 10 fps is quite a bit of difference in high quality.
 
it looks legit what firing squad wrote since the bottom shows vista scores. Needless to say the 2900 xt scores drop like a rock.
 
these seem like bullshit to me, how does a 8800GTS 320mb beat out a 640mb of the exact same card by 7fps?
 
Wow, Ati Spanks every equivalent nVidia in these test. The 1950pro spanks the shit out of the 7900gs and 8600GTS, hell the 1650 is almost beside the 7900GS. Hell the 1900xtx is right behind the 8800GTS.. Hope they do a MOH:A set as well.



Isn't the 512 2900 more of a competitor to the GTS 640, then the GTX?

No mate, thats what gts user have ALLEDGED 2900xt is up against. Everywhere u go before 8800gts user think that it kicks 2900xt ass.
But the truth is mate, 2900xt was designed for dx10 and future titles. 2900xt users have been waiting to fight back for a while now becuase there wern't much real dx10 games out.
Now the tides have turned. With more dx10 titles like Assassins Creed, Gear of war, Unreal t 3 and crysis on the way, i can almost assure you that the 2900xt will match or beat the GTX.
Remeber the price of 2900xt is match with the gts however on paper the 2900xt is better and faster than GTX.
For the past few months i've been following almost every 8800gts vs 2900xt thread and as more new gen games are released those 8800gts users are probably banging thier heads in right now becuase they were stupid enough to thing that benchmarks on dx9 games meant 8800gts kicked 2900xt ass.

Oh and if your a 8800gts n gtx owner, i'm not starting anything becuase i don't need to. Come back in a month when crysis demo comes out, and let the FPS do the talking :p

And to all those no brainers that listened to gts user complain about the noise and power comsumption of the 2900xt here are my thoughts.
noise=turn up ur speakers like every gamer should or get headphones!
power comsumption=60watts more, turn a damn light bulb off lol

cheers guys
 
overclocked edition should maybe get 1 fps more, no way would it change 7 fps difference.

A 8800gts is 24 texture per clock pipe card with 320bit memory bus. A little overclock gives you dramatic results.

I suggest you read some articles.
 
Lol, gamespot...hahaha. Those benches are bullshit.
Don't trust gamespot, just wait for tomshardware or anantech.
 
Lol, gamespot...hahaha. Those benches are bullshit.
Don't trust gamespot, just wait for tomshardware or anantech.

Damn you would be stupid to count out gamespot and what if they were right? You would look like a dumb ass wouldn't you? what makes toms or anand better? Why because they're better fitted for mindless zombies like you that dictate for the masses? :eek:

Benchmarks are done by humans. Anyone can make a mistake. Tom's been a joke lately. So has anand. They haven't put out a quality review in ages.

If you are trying to justify your 8800gts purchase I suggest you go back to Tom's.
 
Lol, gamespot...hahaha. Those benches are bullshit.
Don't trust gamespot, just wait for tomshardware or anantech.

lol urself those benches near match the ones from http://firingsquad.com/hardware/bioshock_directx10_performance/page6.asp

lol lol no brainer

VERY CLEAR 2900XT has vista issues and need to be fixed

2900xt 68 fps at 1600x1200 max dx9 win xp
2900xt 48 fps at 1600x1200 max dx9 vista (hello does this look funny to u??VISTA BUG?)

8800gtx 65 fps at 1600x1200 max dx9 winxp
8800gtx 62 fps at 1600x1200 max dx9 vista
8800gtx 60 fps at 1600x1200 max dx10 vista
 
Lol look at all the benches from gamespot and you'll understand that they are completely nonsense, plus, gamespot is biased.
And no, I don't need to justify my 8800 purchase. Besides, with my x800xt-pe, I've never had so much problems, had to rma it twice and it stills has problems. (Asus version mind you)
 
These benchmarks from what I can make out from a quick flick through the site seem to not have AA turned on which we all know is a massive FPS killer for the new 2600/2900 series. AA is perfectly capable in DX9 mode and hopefuly be available in DX10 mode soon yet is left out.

Any review that leaves out AA settings isn't really worth reading IMO, it's a fairly standard feature for the API which almost all gamers with high end kit use.

*edit*

Correction AA is possible in XP and not Vista, either way AA should have been enabled where possible.
 
A 8800gts is 24 texture per clock pipe card with 384bit memory bus. A little overclock gives you dramatic results.

I suggest you read some articles.

I give you the same suggestion. It's 320 bit, for both 8800 GTS. 384 bit is for the GTX.
 
No mate, thats what gts user have ALLEDGED 2900xt is up against. Everywhere u go before 8800gts user think that it kicks 2900xt ass.
But the truth is mate, 2900xt was designed for dx10 and future titles. 2900xt users have been waiting to fight back for a while now becuase there wern't much real dx10 games out.
Now the tides have turned. With more dx10 titles like Assassins Creed, Gear of war, Unreal t 3 and crysis on the way, i can almost assure you that the 2900xt will match or beat the GTX.
Remeber the price of 2900xt is match with the gts however on paper the 2900xt is better and faster than GTX.
For the past few months i've been following almost every 8800gts vs 2900xt thread and as more new gen games are released those 8800gts users are probably banging thier heads in right now becuase they were stupid enough to thing that benchmarks on dx9 games meant 8800gts kicked 2900xt ass.

Oh and if your a 8800gts n gtx owner, i'm not starting anything becuase i don't need to. Come back in a month when crysis demo comes out, and let the FPS do the talking :p

And to all those no brainers that listened to gts user complain about the noise and power comsumption of the 2900xt here are my thoughts.
noise=turn up ur speakers like every gamer should or get headphones!
power comsumption=60watts more, turn a damn light bulb off lol

cheers guys

Here we go again...
You seem to be a part of the group that "wants" to believe that the 2900 XT will be faster in the future, than it is now, but that's not the case. There is no magic in hardware and cards that have lackbuster performance now, will not perform better in future titles. Of course, if a game is developed, with the architecture of some graphics chip in mind (i.e. Radeon 9xxx and Half Life 2 or Doom 3 and GeForce 6xxx) it will run better on said hardware, but this happens in one or two titles and not the majority.
 
Here we go again...
You seem to be a part of the group that "wants" to believe that the 2900 XT will be faster in the future, than it is now, but that's not the case. There is no magic in hardware and cards that have lackbuster performance now, will not perform better in future titles. Of course, if a game is developed, with the architecture of some graphics chip in mind (i.e. Radeon 9xxx and Half Life 2 or Doom 3 and GeForce 6xxx) it will run better on said hardware, but this happens in one or two titles and not the majority.

one or two titles? lol ever heard of unreal 3 engine?
every game that has used unreal 3 2900xt rocks gts n gtx sox. Rainbow six Vegas and now Bioshock 2900xt wins gts and beats if not matches gtx.
Future titles lik Gears of war? Unreal t 3? and a whole heap of games which will use unreal 3 will kick ass on the 2900xt.
So for ur information not just 2 titles lol.

And for all those that think gamespot is biased lol why don't i post some fraps pics or even vids for u to show they are real? gtx users can do the same if they're not scared! ahaha enough of this gamespot is biased thing.
Do ur own benches and post up PROOF!

i will post mine after i get my new hhd in 3 days

Cheers all
 
one or two titles? lol ever heard of unreal 3 engine?
every game that has used unreal 3 2900xt rocks gts n gtx sox. Rainbow six Vegas and now Bioshock 2900xt wins gts and beats if not matches gtx.
Future titles lik Gears of war? Unreal t 3? and a whole heap of games which will use unreal 3 will kick ass on the 2900xt.
So for ur information not just 2 titles lol.

And for all those that think gamespot is biased lol why don't i post some fraps pics or even vids for u to show they are real? gtx users can do the same if they're not scared! ahaha enough of this gamespot is biased thing.
Do ur own benches and post up PROOF!

i will post mine after i get my new hhd in 3 days

Cheers all

Yup UT engine is one of those engines that is going to be used a lot for future games.
 
Yup UT engine is one of those engines that is going to be used a lot for future games.

Almost certainly.

But if Bioshock is evidence of anything is that they dont have to use deffered rendering which kills off the ability to use MSAA in DX9, as this is confirmed as working in XP.

In which case MSAA is a very real possibility for all Unreal engine 3.0 based games, and certainly those running Vista with DX10 (which does support MSAA with deffered rendering according to Mark Rein)

We all know the 2900 cards take a massive performance hit with MSAA turned on, if people want to compare high end video cards it should be done with the settings cranked like we're used to, MSAA is very much a part of that.

So yes I agree, post benchmark results, do so with AA turned on and we'll see how the cards measure up then!
 
one or two titles? lol ever heard of unreal 3 engine?
every game that has used unreal 3 2900xt rocks gts n gtx sox. Rainbow six Vegas and now Bioshock 2900xt wins gts and beats if not matches gtx.
Future titles lik Gears of war? Unreal t 3? and a whole heap of games which will use unreal 3 will kick ass on the 2900xt.
So for ur information not just 2 titles lol.

And for all those that think gamespot is biased lol why don't i post some fraps pics or even vids for u to show they are real? gtx users can do the same if they're not scared! ahaha enough of this gamespot is biased thing.
Do ur own benches and post up PROOF!

i will post mine after i get my new hhd in 3 days

Cheers all

We already know that UT3 is fillrate dependant. And those are not good news for the HD 2900 XT, which severely lacks capabilities in that area. Gears of War, being done by Epic, will probably follow the same logic. Get your facts straight and don't get your hopes up in your quest to "think" the HD 2900 XT will get better, as new games come along, that may use UE3.

Also, we already know that the HD 2900 XT holds the 3DMark record and usually beats almost anything without AA. But who exactly "plays" 3DMark or buys a $400 card, to play with no AA ? Only if that's a limitation of the game itself, which in Bioshock's case and AFAIK, only NVIDIA cards can do AA in DX9.
I prefer to wait for [H]'s real world gameplay evaluations, to know which of these early benchmarks, are true. If Firingsquad's, which favor NVIDIA's hardware or Gamespot's, which favor the HD 2900 XT.
 
Last month: No! You can't test the 2900XT in XP. It is a card of the future and only the DX10 Vista results matter.

Post Bioshock: No! You can't test the 2900XT in Vista. The Vista drivers are buggy and only the DX9 XP results matter.

Larf. What is the next major release and will that pop us back to Vista results again?
 
yeah im really glad i got the 2900 xt. performs great, ATI brand, first ATI dx10, first 512 bus, and even though its not the fastest card, i didnt pay anywhere near $500 for it. $370 actually.
 
We already know that UT3 is fillrate dependant. And those are not good news for the HD 2900 XT, which severely lacks capabilities in that area. Gears of War, being done by Epic, will probably follow the same logic. Get your facts straight and don't get your hopes up in your quest to "think" the HD 2900 XT will get better, as new games come along, that may use UE3.

Also, we already know that the HD 2900 XT holds the 3DMark record and usually beats almost anything without AA. But who exactly "plays" 3DMark or buys a $400 card, to play with no AA ? Only if that's a limitation of the game itself, which in Bioshock's case and AFAIK, only NVIDIA cards can do AA in DX9.
I prefer to wait for [H]'s real world gameplay evaluations, to know which of these early benchmarks, are true. If Firingsquad's, which favor NVIDIA's hardware or Gamespot's, which favor the HD 2900 XT.

Can't you accept that 2900xt can perform. Jeez.. Some people
 
Can't you accept that 2900xt can perform. Jeez.. Some people

No, it's absolutely not that. And you of all people, that have talked about this with me, on several occasions, should know.
It's not that HD 2900 XT can't perform decently, even if not at the levels of competition, but people keep insisting on magic numbers, which the XT will provide, in future titles and that's BS. Yes, it performs very well in some titles (mostly without AA), but in no way, that's an indication of future performance. Graphics cards performance does NOT get better as game's tech evolves, otherwise ATI or NVIDIA would have found the Holy Grail of graphics processing. A GPU that would last you 10 years...
 
I have a really noobish question.....:confused:

It seems that benchmarking is a science and as a science, you should get repeatable results. So, why are the results so very different, depending on who does the testing? Is this dependant solely on the hardware used or are other factors involved? How does bias come into play?

I guess I want things cut and dry, but it seems that rhetoric plays a big part in evaluating PC components.... even when numbers are involved.
 
i just can not see a 2900xt coming out of no where and beating a gtx,if i read these post correctly.
 
Frosteh and Silus are thinking in right way. All 8800 series cards + HD2900 XT should be tested with AA turned on. Only when we'll see real performance of the cards under XP.
As it's already known that UT3 engine is very demanding on texture fillrate power (2900 XT is weak on this point) I take priority to Firingsquad tests, not to mention, that test was done in very graphical intense area. Anyway, we will find out the true soon, as another one HW site will do Bioshock perf. review. Is XT beating GTX or not under XP...
 
I have a really noobish question.....:confused:

It seems that benchmarking is a science and as a science, you should get repeatable results. So, why are the results so very different, depending on who does the testing? Is this dependant solely on the hardware used or are other factors involved? How does bias come into play?

I guess I want things cut and dry, but it seems that rhetoric plays a big part in evaluating PC components.... even when numbers are involved.

Yes, it depends mostly on two factors. The Hardware itself and the person that does it.
First the hardware. When evaluating a graphics card, the reviewer should use a system, which takes everything out of the equation, except the graphics card itself. This means, using the fastest possible CPU available, so that it doesn't bottleneck the graphics card being tested. The amount of RAM is also important and a fast HDD.
Second, the person that does it. He/She is responsible for the methodology used. Some like to use built-in benchmarks, which are an indication of performance, but may or may not be the same as real-world performance i.e. playing the games themselves.
Others, like [H] do play the game and show us the results. Usually tests are repeated a few times, when results seem skewed, but we usually only see the final results in the evaluation itself: Min/Avg/Max fps and quality settings, in a given game.

That said, results will always be different, even if not by much, but it's usual to see some tests, that totally inflate scores and cause some confusion. You should not trust only one source. I trust [H] more than others, due to their testing methodology, but I do check Tech-Report's numbers for comparison. The results may differ sometimes, but usually their conclusions are very similar, even if worded differently.
 
always thought that this card had more than what it has shown so far.of course the jury is still out.maybe this thing will finally live up to so of the hype.who knows.
 
We already know that UT3 is fillrate dependant. And those are not good news for the HD 2900 XT, which severely lacks capabilities in that area. Gears of War, being done by Epic, will probably follow the same logic. Get your facts straight and don't get your hopes up in your quest to "think" the HD 2900 XT will get better, as new games come along, that may use UE3.

Also, we already know that the HD 2900 XT holds the 3DMark record and usually beats almost anything without AA. But who exactly "plays" 3DMark or buys a $400 card, to play with no AA ? Only if that's a limitation of the game itself, which in Bioshock's case and AFAIK, only NVIDIA cards can do AA in DX9.
I prefer to wait for [H]'s real world gameplay evaluations, to know which of these early benchmarks, are true. If Firingsquad's, which favor NVIDIA's hardware or Gamespot's, which favor the HD 2900 XT.

dude
http://www.techarp.com/showarticle.aspx?artno=358

Fillrate
• 36.8 Gigapixels / sec. GTX
• 24.0 Gigapixels / sec. GTS

http://www.3dgameman.com/content/view/8349/47/
2900XT pixel fill rate is
47.5GPixels/s and the memory bandwidth is 106GB/s

SO HOWS THAT NOT GOOD NEWS FOR 2900XT ???????? sorry if i don't sound nice but you just rockedd urself hardddd, next time you get UR facts right. just so u know higher is better? and again i did not bring this issue about fill rates up, he did, i'm just correcting the NO BRAINER :p
 
dude
http://www.techarp.com/showarticle.aspx?artno=358

Fillrate
• 36.8 Gigapixels / sec. GTX
• 24.0 Gigapixels / sec. GTS

http://www.3dgameman.com/content/view/8349/47/
2900XT pixel fill rate is
47.5GPixels/s and the memory bandwidth is 106GB/s

SO HOWS THAT NOT GOOD NEWS FOR 2900XT ???????? sorry if i don't sound nice but you just rockedd urself hardddd, next time you get UR facts right. just so u know higher is better? and again i did not bring this issue about fill rates up, he did, i'm just correcting the NO BRAINER :p

No. 2900xt fillrate is only equivalent to 8800gts. 3dgameman is a joke. I'm sorry.
 
i believe the lack of pipes kills the fill rate unless shear clock speed makes up the difference
 
But it is in windows xp. :rolleyes:

As stated below, depending on who is reviewing the given piece of hardware the results can be skewed to either make one look better or the other look worse. In this case not using AA makes ATI look better... simple as that, same reason why specific hardware makers 'optimize' for specific things, to make their hardware look better in a given bench. Doesn't make it law... just makes it yet another viewpoint.
 
As stated below, depending on who is reviewing the given piece of hardware the results can be skewed to either make one look better or the other look worse. In this case not using AA makes ATI look better... simple as that, same reason why specific hardware makers 'optimize' for specific things, to make their hardware look better in a given bench. Doesn't make it law... just makes it yet another viewpoint.

You might want to read Firingsquad article as well. Dx10 drivers for ATI cards in this game is lacking. With proper optimization it's a serious contender even to the GTX.

The benchmark don't lie. Either way it beats a GTX without AA. If AA was used in this game it can surely beat out the GTS in this game by looking at results how 2900xt behave with other games.

Based on the results we’ve seen up to this point, it looks like NVIDIA’s DX10 driver for BioShock is giving their cards the performance advantage under the game’s DX10 path, but clearly the Radeon HD 2600 XT is no slouch in this game once it is running under the DX9 codepath. As we noted earlier in this article, performance improves by a factor of up to 1.5X under DX9. This is simply an enormous performance improvement and we’re hoping AMD can eventually bring the performance of their DX10 driver closer to this point in the game.

 
Back
Top