• 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.

ATi to get performance boost in Doom3

eraser_16 said:
umm...wtf? "You 9800pro"
anway, i'm going to call the BS card on this one... the 9800 is not based on the .15 micron manufacturing process. it uses the .13micron man. process... you need to get your facts right. the 9500 and 9700 are .15micron, 9600 and 9800 are .13 micron. that's why the 9800 pro doesn't have a humungous cooler on it, it's just a little piece of aluminum with a fan on it.

9800 uses .15, it is a fact, you cannot change that fact no matter how much you delude yourself.
 
eraser_16 said:
umm...wtf? "You 9800pro"
anway, i'm going to call the BS card on this one... the 9800 is not based on the .15 micron manufacturing process. it uses the .13micron man. process... you need to get your facts right. the 9500 and 9700 are .15micron, 9600 and 9800 are .13 micron. that's why the 9800 pro doesn't have a humungous cooler on it, it's just a little piece of aluminum with a fan on it.


Actually although I don't like to get in these pissing matches, the 9800 pro is .15 micron technology.

http://www.firingsquad.com/hardware/hercules_3d_prophet_9800_pro_review/default.asp

http://www.hardocp.com/article.html?art=NDM5LDI=

2 enough? I can get more :D
 
ok, ok, you get 1 fact wrong and everybody thinks you're a dumbass. well, to be honest i deserve to be thought of as a dumbass right now... i have to apologize for my own arrogance and for being such an asshole. i was wrong this time, but i hope everybody doesn't concider me a dumbass from now on, or i'll just have to avoid you guys in the rest of the forum, lol...j/k. i'm really not usually this stupid. and cmon gimme a break, i'm allowed 1 mistake, i'm only 16. ;) oh and i'd especially like to apologize to Met-AL, and to Xtasy0 and CyberCRAP for helping me see my mistake.
 
FragMagnet said:
I have to agree about OGL. It's too bad really, as I have always felt that DX was a bloated API. Glide was a a "true" direct to hardware, OGL a little less, but still very efficient. You'd think it was easier to program for without the bloat, not sure why everyone else went to DX ??

You have to code seperate paths for each architecture. That's why you read so much about Carmack explaining things in his .plan about the ARB1, ARB2, NV20, etc. paths. With DirectX, you code it once.
 
Neurofreeze said:
You have to code seperate paths for each architecture. That's why you read so much about Carmack explaining things in his .plan about the ARB1, ARB2, NV20, etc. paths. With DirectX, you code it once.

Great for the programmer, but it sucks that all the end users have to upgrade/turn down features earlier to cover the code bloat. Me spending $ to cover a rich programmers laziness, doesn't sit well with me. But that's just me :D
 
We don't know about image quality yet, ATI could be the overall winner, we will have to see.
 
eraser_16 said:
ok, ok, you get 1 fact wrong and everybody thinks you're a dumbass. well, to be honest i deserve to be thought of as a dumbass right now... i have to apologize for my own arrogance and for being such an asshole. i was wrong this time, but i hope everybody doesn't concider me a dumbass from now on, or i'll just have to avoid you guys in the rest of the forum, lol...j/k. i'm really not usually this stupid. and cmon gimme a break, i'm allowed 1 mistake, i'm only 16. ;) oh and i'd especially like to apologize to Met-AL, and to Xtasy0 and CyberCRAP for helping me see my mistake.

dont sweat it. you havent even made a small dent in your mistake pile yet...

everyone goes through a know it all phase in their youth, then you get older and realize you dont know diddly....
 
Its very obvious to me that since Ati's opengl drivers suck so bad there is plenty room for improvement. I think they will even the score soon.
 
eraser_16 said:
i have to apologize for my own arrogance and for being such an asshole.

That statement right there is what makes you diffrerent from an asshole. Don't sweat it. You will have plenty of chances to point out flaws and mistakes in my posts to return the "favor."
 
Blad3 said:
We don't know about image quality yet, ATI could be the overall winner, we will have to see.

Are you talking about DOOM 3? How could ATI be the winner in overall image quality for D3? lol

The game was built using nVidia hardware. The nVidia cards support UltraShadow II Technology and full Trilinear filtering with very little performance hit. The X800's only support Trilinear with Brilinear. Fudging was reported on the ATI cards for their filtering quality in D3.

I dont think its going to be any competition which has better IQ. In nearly every review i've read the 6800's have had as good or better IQ then the X800's. Its just a matter of where you look and what you want to try and point out.

The majority of IQ issues pointed out on the 6800's are just game and driver bugs that havn't been fixed yet. All the Far Cry issues were addressed in the 1.2 patch.
 
burningrave101 said:
Thats not true at all. The X800XT PE and 6800u are tied for the performance crown in D3D. The settings you choose to use and the part of the level you test determines which is faster.

The 6800's only weakness is AF. nVidia is suppost to be working on new optimizations in the 62.xx series drivers and if they are pretty good it will most likely put the 6800's ahead of the X800's in the majority of D3D benches.

If you disable AF on the X800XT PE the 6800u wins almost every time in D3D games. The X800XT PE shines in the fact it has better AF optimizations and can run high AF without losing any performance. The 6800's do much the same thing with AA performance.

The 62.01 drivers gave a really nice Far Cry performance boost. The 61.77 drivers also fixed performance issues in Far Cry for the 6800's. All the reviews i've seen by hard are using older 61.45 drivers. The 6800u could very well be beating the X800XT PE with the 62.01 drivers w/ SM 3.0 enabled.

And if you've seen the SM 3.0 Far Cry performance results you would know that the 6800u wins about half the time and the PE wins about half the time.

http://techreport.com/etc/2004q3/farcry/index.x?pg=1

You need to get out and read some more reviews if you think the X800XT PE beats the 6800u in the majority of D3D games.

http://www.gamepc.com/labs/view_con...D95DDC4A8F963FC2C3E5C22B472046C051310D91DDA7A

Hard tests in-game with FRAPS which alot of the time isn't truly accurate because it adds extra variables such as AI interaction which changes with each run. Its impossible to get the same results twice. You just have to take averages and hope you followed the same path and about the same actions happen over and over again which isn't likely. FRAPS can also be optimized for just like any other benchmark. And we all know how ATI likes to optimize :).

And Hard only uses a handful of D3D games like Far Cry, BF: Vietnam, Painkiller, and one or two others. How does that equate for D3D performance when those are the games most highly optimized for by ATI and nVidia? lol

ATI is getting the majority of their performance from Brilinear filtering. Briliner filtering is lower quality then Trilinear. The 6800's will run full Trilinear and dont lose alot of performance. The X800's can't run Full Trilinear except through a registry hack i believe and then i dont know for sure how its done and if it can only be done in certain games. They also lose alot of performance with their Tri and AF ops disabled.


That's not true at all. The X800XT PE and 6800u are NOT tied for the performance crown in D3D. The settings you choose to use and the part of the level you test determines which is faster. Translation: If you want the faster product and don't care about having the best image quality possible, which means giving up high levels of AA and AF, go with Nvidia's high-end product. If you want the faster product and do care about having the best image quality possible, which means NOT giving up high levels of AA and AF, go with ATI's high-end product.

I really wish review sites would quit utilizing underpowered CPUs' and focus on Intel's Pentium Extreme series and AMD's Athlon 3800+ or FX-53(939&940) series, which are as good as it gets now for the respective platforms and gives the true capabilities of the respective video cards and yes, where ATI has an advantage due to its level of CPU scaling.

ATI is getting most of their performance from a more refined architecture with a higher core and memory speeds. Also, ATI's "bri-linear" filtering technique is still on par with Nvidia's "trilinear" filtering, which means Nvidia still has some work to do. As far as OpenGL performance between the two platforms, perhaps we should call it ID'S OPENGL performance to be more specific. Check out the results with IL2 Forgotten Battles running OpenGL and that isn't the only example. As far as all of the other OpenGL games out there such as Call of Duty, and the Jedi series, yes, the Nvidia platform blows ATI away on those, but once again, we are talking about ID'S OPENGL, which also have a number of other factors involved that favor the Nvidia platform besides OpenGL. Another thing is that while Quake Arena, Call of Duty, the Jedi series, etc. may smoke the ATI platform, does it really matter when that doesn't have an impact on the future and ATI already runs those games at 100+ frames per second anyways? I think not.

Quit being a lemming of review sites and YOU need to get out and perform real world tests as opposed to reading more reviews if you think the 6800u beats the X800XT in the majority of D3D games, much less tie them. When benchmarking different storage devices, review sites don't even try multiple controllers, which can make a HUGE difference in benchmark results. That's just another example of short-sighted reviewing techniques.

Doom 3 can beat ATI by 12fps with everything on, because I will take these instead along with still getting over what will be an estimated 45 fps or more at 1600x1200x0aax16af on the platform mentioned below

Here some SCOREBOARD for you:


FX-53 (socket 940)
Asus SK8V motherboard
Corsair XMS3200PTL (1 gig)
ATI X800 XT (4.8 beta Catalyst)
Nvidia 6800 Ultra (61.77)
No overclocking for the system or video card is used


C&C Generals: 1600x1200x4aax16af:

ATI: 60
Nvidia: 36



Colin McRae Rally 4 1600x1200x4xaa:

ATI: 94
Nvidia: 80


Ground Control 2: 1600x1200x4xaax8af

ATI: 73
Nvidia: 76


F1 Challenge: 1600x1200x4aax16af:

ATI: 47
Nvidia: 50


Far Cry (1.2 patch) 1600x1200x4aax16af

ATI: 49
Nvidia: 38


Thief 3: 1600x1200x4aax8xaf

ATI: 59
Nvidia: 33


Unreal Tournament 2004: 1600x1200x4aax16af
ATI: 78
Nvidia: 55


Tron 2.0: 1600x1200x4aax16af

ATI: 88
Nvidia: 54


Stalker: 1600x1200x4aax16af:

ATI: 127
Nvidia: 108


Painkiller: 1600x1200x4aax16aa:

ATI: 117
Nvidia: 85


Splinter Cell: 1600x1200x4aax16aa:

ATI 48
Nvidia: 44


Tomb Raider: Angel of Darkness 1600x1200x4aax16aa

ATI: 71
Nvidia: 55


Max Payne 2: 1600x1200x4aax16aa:

ATI: 145
Nvidia: 122


Kotor: 1600x1200x4aax16af:

ATI: 38
Nvidia: 33


IL2 Forgotten Battles(OpenGL): 1600x1200x4aax16af

ATI: 41
Nvidia: 27


Half-Life 2: 1600x1200 (no aa or af)

ATI: 107
Nvidia: 96


Lock-On: 1600x1200x4aax16af:

ATI: 19
Nvidia: 16


X2: 1600x1200x4aax16af:


ATI: 54
Nvidia: 52
 
If you're concerned about image quality...I would go with the card that actually allows you to turn off trilinear optimizations...if you like to wait long periods of time on pre-order and don't mind the somewhat questionable optimizations, the XTPE is almost unbeaten at 16x12 4xaa 8xaf, except for doom3 that is...;)

id OpenGL vs OpenGL...stop making excuses for ATi's shitty OpenGL drivers...even they admit they need to completely rewrite it...and not just for games, for thier workstation cards too...they admitted they suck at OpenGL, why can't you?
 
yep seems that ati can beat nvidia in OGL witth everything on. however i still think the 6800 is the better series of cards this round
 
If it ever happens, D3+3Dc = x800's ~even with same price 6800's.

If ATI also improves their drivers then they might actually beat the 6800's. Of course 3Dc will result in overall image quality being about 2or 3% lower on ATI than on Nvidia, probably just worse enough to be noticeable under a few rare circumstances.
 
tornadotsunamilife said:
yep seems that ati can beat nvidia in OGL witth everything on. however i still think the 6800 is the better series of cards this round

do what? even with everything on the x800s can't beat the 6800s in OpenGL...not that I've seen anyways...in fact the vanilla 6800nu beats the XTPE in Jedi Academy...clearly ATi has OpenGL driver issues...
 
arentol said:
If it ever happens, D3+3Dc = x800's ~even with same price 6800's.

If ATI also improves their drivers then they might actually beat the 6800's. Of course 3Dc will result in overall image quality being about 2or 3% lower on ATI than on Nvidia, probably just worse enough to be noticeable under a few rare circumstances.

I think a driver boost could conceivably put the XTPE over the Ultra...but you still can't buy the XTPE, so the GT remains the card to have in either circumstance...
 
I don't give a shit too much, looks like my card is fine just the way it is right now for Doom 3.(according to the Doom 3 hardware guide)
 
TheCrimsonCrusader said:
FX-53 (socket 940)
Asus SK8V motherboard
Corsair XMS3200PTL (1 gig)
ATI X800 XT (4.8 beta Catalyst)
Nvidia 6800 Ultra (61.77)
No overclocking for the system or video card is used

Kotor: 1600x1200x4aax16af:

ATI: 38
Nvidia: 33

although possibly wrong the nvidia is being beaten in an OGL game
 
^eMpTy^ said:
I think a driver boost could conceivably put the XTPE over the Ultra...but you still can't buy the XTPE, so the GT remains the card to have in either circumstance...

Well, maybe YOU can't buy one... ;)

MoBo -- ABIT KV8 Pro
CPU -- AMD 64 3000+ Newcastle @ 240mhz
Video Card -- ATI Radeon x800XT-Platinum Edition @ 552/1142
RAM -- 1GB Kingston DDR PC-3200 (single stick)
CPUHSF -- ThermalTake Tower112 w/2 90mm fans
GPUHSF -- ThermalTake Extreme Giant III (modded)
Hdrive1 -- Hitachi 160GB 7200RPM SATA (Raid 0)
Hdrive2 -- Hitachi 160GB 7200RPM SATA (Raid 0)
PSU -- 480watt ThermalTake Silent PurePower
FAN Controller -- Coolermaster Aerogate I
Sound Card -- Audigy2 ZS Gamer
Case -- ThermalTake Xaser II A6000A
Monitor -- Hitachi CM772B 19" - 1600x1200@85hz
Speakers -- 900mhz Sony Wireless Headphones
DVD Drive -- Sony 16x DVD-ROM
 
^eMpTy^ said:
If you're concerned about image quality...I would go with the card that actually allows you to turn off trilinear optimizations...if you like to wait long periods of time on pre-order and don't mind the somewhat questionable optimizations, the XTPE is almost unbeaten at 16x12 4xaa 8xaf, except for doom3 that is...;)

id OpenGL vs OpenGL...stop making excuses for ATi's shitty OpenGL drivers...even they admit they need to completely rewrite it...and not just for games, for thier workstation cards too...they admitted they suck at OpenGL, why can't you?

I never said they weren't. What I am saying is that the performance disparity between OpenGL on the Nvidia and ATI platforms is not as great as people make it out to be if they factor in other OpenGL titles such as IL2 Forgotten Battles and KOTOR for an example. Given the mismatch between D3D and OpenGL performance on the ATI platform, it's safe to say that yes, the performance does suck. The good news is that is at the software level and not at the hardware level and there is a good chance that ATI will finally put forth some effort in regards to their OpenGL driver implementation, since they are going to be hurt by the GeForce 6800 GT performance in Doom 3 even though that card gets absolutely slaughtered and I mean SLAUGHTERED in every other gaming benchmark out there. With that said, you go ahead and buy the fastest Doom 3 card and I will stick with the fastest gaming card. :)
 
That's about the least spectacular Inquierer article I've seen. It really doesn't say anything we didn't already know: ATI is planning to release a driver sometime in the future to make OpenGL games like D3 run better. Whoop-dee-doo.
 
Kotor runs pretty well on a 9800 but not as good as it could, and I don't think you will be worrying about framerates with either of the X800 or 6800 (any version) unless you want to run 2048x1536 or something.

So there never really has been an overriding need for exceptional OpenGL performance, up until Doom3 its been more of do you want 100 or 120fps. At most with OpenGL 1.5 you might have to turn down your AA to 2x from 4x to get acceptable performance.

Honestly other than Doom3 (which could be a very big deal) the announcement that ATi would work on its OpenGL is almost a little bit of a yawner. I'd rather see them do a specialized Doom3 driver (but that would have Quack/quake purists up in arms) as long as its faster and image quality is not compromised. Remodelling the entire OpenGL API just to run Doom3 a little better seems a little like a waste of effort and time.
 
gumplunger said:
That's about the least spectacular Inquierer article I've seen. It really doesn't say anything we didn't already know: ATI is planning to release a driver sometime in the future to make OpenGL games like D3 run better. Whoop-dee-doo.

well considering that you have a 6800GT, i can understand your excitement level for the new ATI drivers. While me on the other hand can't wait for these drivers and having been crossing my fingers ever since the D3 benchmarks were released. ;) ...right now i'm not expecting much more of a performance increase, so i won't be disappointed if they're not that great. But i still do realize i'll be able to play D3 with a good res and hopefully 4AA/16AF - that's my hopeful goal, but probably will have to settle for something like 2AA/8AF or 16AF or whatever...i don't know, i guess i'll find out :p

I really got screwed over with my card (although not really)...i got the card and a few days later D3 benchmarks released (kind of a downer, but i wasn't really surprised)...next i find out i could have gotten a x800pro VIVO and softmodded it to a x800XT :( ...the place i didn't get screwed was my wallet, i got it for free :D and it still plays all my games great at 1280x1024 6AA/16AF. so here's hoping those drivers are awesome. *goes and prays*
 
sdornan said:
How'd you get it free?

it was bought for me for doing a couple computer builds :cool: ...i got paid for my time and they threw the card in as a little extra payment
 
truth is, both the X800 and the 6800 cards are great performers with excellent IQ, and either card you buy will do great in doom3. but, the 6800 does have the edge in most opengl games, and right now doom3 is one of them. while the X800 has the edge in most directx games. however, i think we should wait until the game is out, and all the drivers are optimized for the game. then we will see the truth behind these cards. but as for now, i'm betting that the 6800 is going to stay in the lead, but the X800 may pull it up a few FPS. at least, i hope it does, i hate to see one card beat the hell out of another.
 
^eMpTy^ said:
do what? even with everything on the x800s can't beat the 6800s in OpenGL...not that I've seen anyways...in fact the vanilla 6800nu beats the XTPE in Jedi Academy...clearly ATi has OpenGL driver issues...

"not that I've seen anyways" = I have only played iD software titles

It's well documented that ATI's OpenGL performance is not up to snuff in relation to their physical hardware. However, the negative level of performances in general is grossly overstated and the negative level of performances are mostly tied to OpenGL titles developed by id Software, not by other developers. That doesn't mean that id Software is doing something wrong obviously, they are just utilizing different rendering techniques that just so happen to allow them to put out the most impressive game engines on the market. Doom 3 is also getting a boost from other feature sets of the 6800x platform, so in a straight OpenGL to OpenGL comparison, the difference is even less between the two platforms. For me, there are just more games that fall on the borderline of not being acceptable at 1600x1200x4xxx8/16af on the Nvidia platform as opposed to the ATI platform, which doesn't have that problem and I can also live with Doom 3 just running at a meager 50 fps at 1600x1200x0aax16af. :)
 
TheCrimsonCrusader said:
That's not true at all. The X800XT PE and 6800u are NOT tied for the performance crown in D3D. The settings you choose to use and the part of the level you test determines which is faster. Translation: If you want the faster product and don't care about having the best image quality possible, which means giving up high levels of AA and AF, go with Nvidia's high-end product. If you want the faster product and do care about having the best image quality possible, which means NOT giving up high levels of AA and AF, go with ATI's high-end product.

I really wish review sites would quit utilizing underpowered CPUs' and focus on Intel's Pentium Extreme series and AMD's Athlon 3800+ or FX-53(939&940) series, which are as good as it gets now for the respective platforms and gives the true capabilities of the respective video cards and yes, where ATI has an advantage due to its level of CPU scaling.

ATI is getting most of their performance from a more refined architecture with a higher core and memory speeds. Also, ATI's "bri-linear" filtering technique is still on par with Nvidia's "trilinear" filtering, which means Nvidia still has some work to do. As far as OpenGL performance between the two platforms, perhaps we should call it ID'S OPENGL performance to be more specific. Check out the results with IL2 Forgotten Battles running OpenGL and that isn't the only example. As far as all of the other OpenGL games out there such as Call of Duty, and the Jedi series, yes, the Nvidia platform blows ATI away on those, but once again, we are talking about ID'S OPENGL, which also have a number of other factors involved that favor the Nvidia platform besides OpenGL. Another thing is that while Quake Arena, Call of Duty, the Jedi series, etc. may smoke the ATI platform, does it really matter when that doesn't have an impact on the future and ATI already runs those games at 100+ frames per second anyways? I think not.

Quit being a lemming of review sites and YOU need to get out and perform real world tests as opposed to reading more reviews if you think the 6800u beats the X800XT in the majority of D3D games, much less tie them. When benchmarking different storage devices, review sites don't even try multiple controllers, which can make a HUGE difference in benchmark results. That's just another example of short-sighted reviewing techniques.

Doom 3 can beat ATI by 12fps with everything on, because I will take these instead along with still getting over what will be an estimated 45 fps or more at 1600x1200x0aax16af on the platform mentioned below

Here some SCOREBOARD for you:


FX-53 (socket 940)
Asus SK8V motherboard
Corsair XMS3200PTL (1 gig)
ATI X800 XT (4.8 beta Catalyst)
Nvidia 6800 Ultra (61.77)
No overclocking for the system or video card is used


C&C Generals: 1600x1200x4aax16af:

ATI: 60
Nvidia: 36



Colin McRae Rally 4 1600x1200x4xaa:

ATI: 94
Nvidia: 80


Ground Control 2: 1600x1200x4xaax8af

ATI: 73
Nvidia: 76


F1 Challenge: 1600x1200x4aax16af:

ATI: 47
Nvidia: 50


Far Cry (1.2 patch) 1600x1200x4aax16af

ATI: 49
Nvidia: 38


Thief 3: 1600x1200x4aax8xaf

ATI: 59
Nvidia: 33


Unreal Tournament 2004: 1600x1200x4aax16af
ATI: 78
Nvidia: 55


Tron 2.0: 1600x1200x4aax16af

ATI: 88
Nvidia: 54


Stalker: 1600x1200x4aax16af:

ATI: 127
Nvidia: 108


Painkiller: 1600x1200x4aax16aa:

ATI: 117
Nvidia: 85


Splinter Cell: 1600x1200x4aax16aa:

ATI 48
Nvidia: 44


Tomb Raider: Angel of Darkness 1600x1200x4aax16aa

ATI: 71
Nvidia: 55


Max Payne 2: 1600x1200x4aax16aa:

ATI: 145
Nvidia: 122


Kotor: 1600x1200x4aax16af:

ATI: 38
Nvidia: 33


IL2 Forgotten Battles(OpenGL): 1600x1200x4aax16af

ATI: 41
Nvidia: 27


Half-Life 2: 1600x1200 (no aa or af)

ATI: 107
Nvidia: 96


Lock-On: 1600x1200x4aax16af:

ATI: 19
Nvidia: 16


X2: 1600x1200x4aax16af:


ATI: 54
Nvidia: 52


Hmm these numbers look very close to driverheaven, looks like your af opt for nv are off
 
The X800XT PE and 6800u are NOT tied for the performance crown in D3D.

Is anyone else tired of hearing about winning a "crown?" It's been repeated over and over until it's a very old cliche. :(
I think it's time to come up with a better term for the performance leader. :)
 
Back
Top