• 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.

FarCry SM3.0 patch benchmarked

LancerJay said:
notice something when comparing the 2 pictures? look behind the phone on the desk.

showimage.html


showimage.html



optimizations ;)
Click to expand...

And the side of the gun, and the box missing on the upper right.

Unless somebody already caught those other two things.

Just started reading thread after building a puter for 2 days straight.
 
Xenozx said:
From what i understand what they did with SM3.0 ... can also be done in shader model 2.0 if they took the time to do so correct?
Click to expand...

This is true, though the situation is really the other way around. An element that is rendered with multiple SM2 instructions can be redone with a single SM3 instruction. The whole point of SM3 is that an operation which would take many clock cycles with SM2 can now be done in a single clock cycle. More operations performed per number of cycles = higher performance. This is the idea behind SIMD (Single Instruction, Multiple Data) instruction sets. Intel's MMX, SSE 1 through 3, and AMD's 3D-Now! are based on this same principle.
 
Dromedary said:
they just said that because they knew that a thread about it would get 10 pages of replies in less than 30 mins if i were guessing... yea, i wouldn't want a fanATIc with a sniper rifle pointed at my skull either
Click to expand...


considering the low framrate it shouldnt be a problem :D

BWAHAHA

Sorry couldnt resist ;)
 
I feel like every review I read about 6800s or x800 is bias in some way. Either leaning towards nvidia or ati. I want a non bias reveiw of the cards so I can get a straight answer.

Also that 6800ultra extreme? so thats just an overclocked 6800ultra? how is that fair to benchmark over a non over clocked x800XT PE? that should be overclocked too and then bench marked.
 
pistola said:
Also that 6800ultra extreme? so thats just an overclocked 6800ultra? how is that fair to benchmark over a non over clocked x800XT PE? that should be overclocked too and then bench marked.
Click to expand...
a 'x800xt-pe OC' cannot be bought tho. a 6800U OC or EE can, its just kinda hard to find or the price is :eek:
 
Has this been posted yet? http://www.firingsquad.com/hardware/far_cry_sm30/.

There is an obvious difference between the two reviews. First off you'll see the XT beating the 6800UE SM3 in most if not all high res 4xAA/8xAF benchmarks with the FS review. However, I noticed that FS uses dx9.0b instead of 9.0c like anand. I remember reading that C enables the use of SM3, while b doesn't. If thats the case what the hell is the point of doing a review of SM3.0 with far cry when SM3 isn't being used??? Am I missing something?

"So what do you need if you’re a GeForce 6800 user and you want to get SM 3.0 up and running in Far Cry? First, you’ll need the 1.2 patch. We’ve been told that Ubisoft will be releasing it on July 5th, next Monday. Next, you’ll need a copy of DX9.0c. Currently, the only way to obtain DX9.0c is to download Microsoft’s Service Pack 2 release candidate. Finally, you’ll need an NVIDIA display driver with SM 3.0 support. NVIDIA provided us with a copy of ForceWare 61.45 for our testing."

Test Setup:

"DirectX 9.0b"
 
In response to this comment:

Conversation is a great thing, however, I do believe that if you follow the trend of this thread, several fanboys have crept up in the process. Equal conversation would be a utopian thing, however and unfortunatly, it is not.
Click to expand...

I think that most guys are just ribbin' each other, not attacking the intelligence of another person. Just just good ole' fashioned giving another bud some shit. Nothing malicious nor serious about it.

Example-
I have some hard core FPS gaming lan buddies. When they saw that I had bought Rise of Nations & The Sims 3000 they gave me hell. I was lold that I was insane, taking BAD crack and turned gay. No harm done, we are doing some classic HL1 gaming this weekend.

Just a good ole boy, never meaning no harm. :D

Relax and have a little harmless fun!!!
 
agar said:
"DirectX 9.0b"
Click to expand...
it doesn't matter, you can run SM3.0 code with the dx9 original release, the only thing 9.0c adds is the ability to compile HLSL to SM 3.0 code.
 
Then does someone want to try to explain the major differences between the two?? Also, both sites use the same drivers.
This is from mikechai over at nvnews.net. He sums up the performance discrepancies between the 2 sites really nicely.

Training 1600x1200, 4xAA, 4xAA+8xAF

FS 6800U, 77.5, 52.7, 42.2
FS XT PE, 76.1, 55.5, 50.1
AT 6800U, 73.6, -NA-, 56.0
AT XT PE, 73.7, -NA-, 48.9

Research 1600x1200, 4xAA, 4xAA+8xAF
FS 6800U, 77.6, 57.3, 44.8
FS XT PE, 72.1, 55.0, 51.3
AT 6800U, 77.6, -NA-, 59.3
AT XT PE, 69.6, -NA-, 49.3

Regulator 1600x1200, 4xAA, 4xAA+8xAF
FS 6800U, 73.6, 47.3, 38.5
FS XT PE, 70.9, 44.3, 40.4
AT 6800U, 73.1, -NA-, 54.6
AT XT PE, 65.0, -NA-, 38.9

Volcano 1600x1200, 4xAA, 4xAA+8xAF
FS 6800U, 76.6, 53.6, 45.2
FS XT PE, 70.7, 49.6, 47.8
AT 6800U, 76.9, -NA-, 61.8
AT XT PE, 67.1, -NA-, 45.4


Edit: I think this may be the diff:

Anand ues "Windows XP Professional SP2 RC2 with DX9.0c and the Summer 2004 DirectX SDK Update" while FS uses "Windows XP Professional SP1+ DirectX 9.0b".
 
agar said:
Has this been posted yet? http://www.firingsquad.com/hardware/far_cry_sm30/.

There is an obvious difference between the two reviews. First off you'll see the XT beating the 6800UE SM3 in most if not all high res 4xAA/8xAF benchmarks with the FS review. However, I noticed that FS uses dx9.0b instead of 9.0c like anand. I remember reading that C enables the use of SM3, while b doesn't. If thats the case what the hell is the point of doing a review of SM3.0 with far cry when SM3 isn't being used??? Am I missing something?

"So what do you need if you’re a GeForce 6800 user and you want to get SM 3.0 up and running in Far Cry? First, you’ll need the 1.2 patch. We’ve been told that Ubisoft will be releasing it on July 5th, next Monday. Next, you’ll need a copy of DX9.0c. Currently, the only way to obtain DX9.0c is to download Microsoft’s Service Pack 2 release candidate. Finally, you’ll need an NVIDIA display driver with SM 3.0 support. NVIDIA provided us with a copy of ForceWare 61.45 for our testing."

Test Setup:

"DirectX 9.0b"
Click to expand...


I just got done looking at that review.

First of course there is the DirectX 9.0b in the test setup. Now that could of been just a typo but then when you go and look at the benches it just doesn't seem very accurate compared to the other reviews. The 6800's with SM 3.0 implemented are scoring lower then the regular 6800's with SM 2.0.

It leaves me to believe they fudged somewhere.

The benches at Tech Report give a little more accurate representation i think.

http://techreport.com/etc/2004q3/farcry/index.x?pg=3

See look here:

http://www.firingsquad.com/hardware/far_cry_sm30/page11.asp

1600x1200 w/ 4x AA + 8x AF [Research Demo] the 6800u with SM 3.0 scored 44.8 fps.

Now look here:

http://techreport.com/etc/2004q3/farcry/index.x?pg=3

1600x1200 w/ 4x AA + 8x AF [Research Demo] the 6800u with SM 3.0 scored 50.71 fps.

Version 1.2 of Far Cry will apparently come with four built-in demos for benchmarking. Those demos take place on the four levels mentioned in the NVIDIA presentation. Rather than use those pre-recorded demos, however, we elected to record five of our own—one on each of the four levels NVIDIA mentioned, and one on the "Control" level. The demos are downloadable via a link below.
Click to expand...

Tech Report set up their own demo's instead of using the four demo's that are now built into Far Cry.

We set the game's image quality options as high as possible, including "Very High" for every option except water quality, which we set to "Ultra High." We tested without antialiasing or anisotropic filtering, and then we tested with 4X AA and 8X anisotropic filtering.
Click to expand...

Some screenies:

http://techreport.com/etc/2004q3/farcry/screen1.jpg

http://techreport.com/etc/2004q3/farcry/screen2.jpg

The new Far Cry patch does indeed seem to be a decent showcase for Shader Model 3.0's potential. The GeForce 6800 cards gain up to about 10% in average frame rates when using the SM3.0 code path. No, the differences aren't going to convert the masses into NVIDIA fanboys overnight, but they do show NVIDIA wasn't kidding about Shader Model 3.0 offering some advantages.

This patch also seems to have cleaned up some problems with the GeForce 6800 code path in the game. Even with Shader Model 2.0, GeForce 6800 performance is way up. Image quality also seems to be quite a bit better than it was on the GeForce 6800 with Far Cry version 1.1. My eye is accustomed to seeing the game run on an Radeon 9800, and the GeForce 6800 cards looked just fine to me with the new patch. I played through a few levels of the game, including some with lots of pixel shader-laden effects on the walls and floors, and I didn't notice any corner cutting or blocky shading or texturing. Of course, some of these changes may be attributable to newer NVIDIA drivers, but the effect is the same.
Click to expand...
 
burningrave101 said:
See look here:

http://www.firingsquad.com/hardware/far_cry_sm30/page11.asp

1600x1200 w/ 4x AA + 8x AF the 6800u with SM 3.0 scored 44.8 fps.

Now look here:

http://techreport.com/etc/2004q3/farcry/index.x?pg=3

1600x1200 w/ 4x AA + 8x AF the 6800u with SM 3.0 scored 50.71 fps.
Click to expand...

In response to that, you can definitely see the difference in FS vs. Techreport, however they are no where near what anand had.

67.4 6800UE
59.3 6800U.

This doesn't seem right, and now that techreport doesn't show to much difference form FS, I am becomming a little skeptical at Anand's results.

http://www.anandtech.com/video/showdoc.html?i=2102&p=7
 
the Tech Reports benches don't show a blow out like Anand's.....The PE still wins some and is very close in others....

By the time I decide what to buy the next gen cards will be out!
 
agar said:
In response to that, you can definitely see the difference in FS vs. Techreport, however they are no where near what anand had.

67.4 6800UE
59.3 6800U.

This doesn't seem right, and now that techreport doesn't show to much difference form FS, I am becomming a little skeptical at Anand's results.

http://www.anandtech.com/video/showdoc.html?i=2102&p=7
Click to expand...

Remember, this is an NVIDIA provided demo and highlights the performance benefits of the new rendering path. At the same time, if we had benchmarked this level, we would have done the same thing by following the path on which the game takes the player. This kind of benefit does exist in the game, but so do times when (as we have shown) no real improvement can be seen.
Click to expand...

http://www.anandtech.com/video/showdoc.html?i=2102&p=7

Bottom of the page.

Version 1.2 of Far Cry will apparently come with four built-in demos for benchmarking. Those demos take place on the four levels mentioned in the NVIDIA presentation. Rather than use those pre-recorded demos, however, we elected to record five of our own—one on each of the four levels NVIDIA mentioned, and one on the "Control" level. The demos are downloadable via a link below.
Click to expand...

http://techreport.com/etc/2004q3/farcry/index.x?pg=2

Tech Report test setup.

Benchmarks

Far Cry 1.2 (custom and standard demos)
Click to expand...

FiringSquad test setup.

http://www.firingsquad.com/hardware/far_cry_sm30/page2.asp

Their all using different demo's, their all getting different results. Reason being, depending on where you record the demo in the game, it will influence how the cards perform. nVidia probably knows EXACTLY where the 6800u will perform the best with SM 3.0 and thats the route they chose to record their demo.

The X800XT PE and 6800u nither one are going to perform the same anywhere in the game. It will be back and forth. Thats why i'm severely skeptical of the HardOCP review because it just leads me to believe they happened to choose an area where the X800XT PE is able to perform the best.
 
burningrave101 said:
Thats why i'm severely skeptical of the HardOCP review because it just leads me to believe they happened to chose an area where the X800XT PE is able to perform the best.
Click to expand...

I agree with your analysis, however those are some pretty loaded words you are saying there.
 
agar said:
I agree with your analysis, however those are some pretty loaded words you are saying there.
Click to expand...

I'm not saying they did it on purpose, i just dont think it was accurate using FRAPS and i think the X800XT PE might of had an advantage somehow. Because the X800XT being able to run 1600x1200 and outperform the 6800u at 1280x1024 is just not sensible.

Thats just my opinion though. Your welcome to your own.

And if the Anandtech review isn't enough to chew on, then check this one out lol.

With all the controversy of graphics testing in the past six months, it would be insane for me to present these numbers to you without telling you exactly how they were tested. All three demos that you will see used in FarCry (Fort, Dam and Regulator) were recorded using the in-game "demorecord" function in the developer mode of the game. During the demo record, and while in god mode, the enemies of FarCry do not interact with the player, so I chose levels in which there were a lot of characters and in-game character interaction if possible. The recorded demos are therefore actual gameplay, just in a very repeatable form. The demos were played back and benchmarked using the latest version of FRAPs to get the framerates and framegraphs.
Click to expand...

http://www.pcper.com/article.php?aid=55&type=expert

FarCry - Regulator Demo

This particular level was recommended by NVIDIA to be used to demonstrate the new SM3.0 patch in FarCry due to its use of grasses and trees that take advantage of the geometry instancing ability of DirectX 9.0c. Also, the intro to the level uses a heavily lighted room that allows the SM3.0 to do multiple light renderings in a single pass instead of requiring one for each light.
Click to expand...

Far Cry 1.2 - Regulator - 1600x1200 - 4x AA / 8x AF

6800u (SM 3.0) average fps = 60.8
X800XT average fps = 40.5

FarCry - Regulator Demo w/ Flashlight

We decided to add the use of the Flashlight in the Regulator demo as this is a dark level and you may actually use it during the game here. Also, the light adds an additional pass for the graphics to render and thus may show us more of the abilities of SM3.0.
Click to expand...

Far Cry 1.2 - Regulator w/ Flashlight - 1600x1200 - 4x AA / 8x AF

6800u (SM 3.0) average fps = 46.3
X800XT average fps = 35.2

Far Cry 1.2 - Regulator Demo @ Tech Report - 1600x1200 w/ 4x AA / 8x AF

http://techreport.com/etc/2004q3/farcry/index.x?pg=3

6800u (SM 3.0) average fps = 37.75 fps
X800XT PE average fps = 38.33 fps

It makes you wonder which card is actually performing the best when each review is showing different results lol.
 
i was waiting for something like this to push me either x800xt or 6800u but it hasn't changed me at all, reviews are a little to contrasting for me to make a real decision. we need just 1 no nonsense benchhmark using a series of maps not just one recommended by nvidia or ati or whatever
 
tornadotsunamilife said:
i was waiting for something like this to push me either x800xt or 6800u but it hasn't changed me at all, reviews are a little to contrasting for me to make a real decision. we need just 1 no nonsense benchhmark using a series of maps not just one recommended by nvidia or ati or whatever
Click to expand...

Well the reason why nVidia recommended those maps is because thats where SM 3.0 is showing up the most right now. It wouldn't do much good to run around and benchmark in a map that uses very little SM 3.0 support lol.
 
tornadotsunamilife said:
i was waiting for something like this to push me either x800xt or 6800u but it hasn't changed me at all, reviews are a little to contrasting for me to make a real decision. we need just 1 no nonsense benchhmark using a series of maps not just one recommended by nvidia or ati or whatever
Click to expand...
That is the very fundamental base from which encourage "cheating", you should get used to using your own brain power, culling the data out of cooked-reviews and idiotic-reviews.
 
they need to benchmark the whole game from start to finish to have good results that really tells me how the cards behave on avarage in the games from now on ;)
 
burningrave101 said:
Well the reason why nVidia recommended those maps is because thats where SM 3.0 is showing up the most right now. It wouldn't do much good to run around and benchmark in a map that uses very little SM 3.0 support lol.
Click to expand...

What is the use of benching heavy SM3 maps only when there are plenty of others that aren't SM3 heavy? Those other maps will not have the benefit of SM3 and will not run as well. Trying to draw a conclusion about the performance of the game from only SM3 heavy maps is completely ignorant. It severely skews the results. The only conclusion you could draw from it, is that SM3, in some cases, gives a performance boost.

At this time, there is only one game that should soon have limited SM3 support. Before someone tries to flame me over that statement, remember, that to get full use out of something like SM3 it needs widespread use throughout a game. To properly do that, you will need the game programmed specifically for it. Otherwise you just get patches of it here and there and thus, only get the performance increase in those certain areas.

Also, before someone tries to yell fanboy, look at the vid card I'm running. If you can't tell, I'm not too worried about hardcore gaming. Hell, the last game I bought and played was RTCW. I'm also not going to buy either one of these cards unless I win the lottery. I just don't have the use for something more than what I currently have.

Hopefully, I have put some of this in perspective at least a little for some of the people here.
 
The X800XT PE and 6800u are pretty much on par performance wise with each other. Ither card will give you more performance then your CPU can handle unless you run high resolutions w/ AA + AF.

You can't go wrong with ither one of them.

The 6800u has a little better features and is stronger in OpenGL. Driver support is also better for the 6800's, especially if your into Linux.

What is evident from all these new reviews is that the 6800GT is faster then the X800Pro. I've got several new reviews here now, along with a couple i dont think any of you have seen yet and i'm going to try and host some images tonight so i can post the benches in the thread.

I've got to go to work now so the rest of you play nicely while i'm gone :p.
 
I am in no way a fan boy of either really, i just find it extremely odd and weird that in 1 patch the ultras go from HH x800 farcry 1 month or so ago to what it was in anandtechs review..

All in all what i want to know is how this will effect half life 2 doom 3 coming up, i have already ordered my x800xt but if this is the kind of performance boost i can expect in upcoming games then i will cancel my order today.
 
FlyinBrian said:
I am in no way a fan boy of either really, i just find it extremely odd and weird that in 1 patch the ultras go from HH x800 farcry 1 month or so ago to what it was in anandtechs review..

All in all what i want to know is how this will effect half life 2 doom 3 coming up, i have already ordered my x800xt but if this is the kind of performance boost i can expect in upcoming games then i will cancel my order today.
Click to expand...

It actually wasn't one patch. The 6800s have been significantly increasing in speed with each driver release. The latest one with the patch put them over the top, thats all.

Also remember back then may sites benched Nvidia's full trilinear versus ATI's brilinear because they thought ATI was producing full trilinear when they weren't and ATI requested them to benchmark in that manner; that further skewed benchmarks.
 
..it would seem that both cards have there plus's and minus's ...I have an XT and an Ultra OC on pre-order ...which ever comes first ..thats what I'll go with ..if anybody knows where to actually buy or pre-order the 6800 Ultra EE ..can yuh post a link please and thank you? :)
 
burningrave101 said:
The 6800u has a little better features and is stronger in OpenGL. Driver support is also better for the 6800's, especially if your into Linux.
Click to expand...

Ok....burningrave101....stop parroting. :rolleyes: I have seen you repeat this time and time again and have ignored it but really....you don't even use linux so how on earth would you know? You don't. For the record....IMO NEITHER works well in Linux. While nVidia has put a little more effort into it then ATI....neither is a simple install trust me. Even current implementation on nvidia linux drivers is jacked up. 4 stack anyone? The current version linux driver for nvidia doesnt work well with Fedora core 2 or red hat enterprise...you have to completely reinstall the 2.6 kernel from source....for those that dont follow what that means it would be like ripping out the guts of windows XP and recreating fgrom scratch.....sound like an advantage?

I didn't think so.....
 
FS's and Techreports numbers completly contradict Anands numbers.

From the Volcano level 1600x1200 4xAA/8xAF

AT:
6800U: 61.8
XT/PE: 45.4

FS:
6800U: 45.2
XT/PE: 47.8

Thats a pretty large difference in scores. FS uses a 3800+, while Anand uses a 3400+, so you would expect frames to be a little higher, like FS shows for ATi. How did they get such a difference in scores.. ? Anands review doesnt show the 6800's taking a bigger hit with AA/AF like we've come accustomed to. Id like to see [H] and B3D to a review. As it is now, Im not sure which way to believe.

I dont believe Anand has 4xAA on.
 
but don't forget the 1.2 patch will only support sm3.0, which is more or less just a performance boost we will probably see the eye-candy features in 1.3
 
the guy that did their review has posted over at beyond 3d stating that the dx9b inthe setup wasa copy paste error. they are using sp2 with dx9c. hes also pretty sure that anand did not havs 4xaa on for his tests, something about having to set the drivers to app specific and forcing it ingame, instead of fromt eh drivers.
 
ah ok.

should be interesting on upcoming reviews on SM3.0 titled games. or even that 1.3 patch.
 
whoa thats aweseome, in that review flyinbrian linked us to and OCed x800pro BEAT the x800XT

they dont even need a 2nd molex to do that ;)
 
retardedchicken said:
whoa thats aweseome, in that review flyinbrian linked us to and OCed x800pro BEAT the x800XT

they dont even need a 2nd molex to do that ;)
Click to expand...
fcoc.png

heh i dont see a oced x800pro beating a x800xt. close, but not beating.

maybe it was too low of a res tho.
 
Bad_Boy said:
fcoc.png

heh i dont see a oced x800pro beating a x800xt. close, but not beating.

maybe it was too low of a res tho.
Click to expand...

It will be interesting to see how many reviews that use "custom" benchmark demo's instead of the Nvidia cherry picked flavor get the same results.
 
MrHappyGoLucky said:
It will be interesting to see how many reviews that use "custom" benchmark demo's instead of the Nvidia cherry picked flavor get the same results.
Click to expand...

maybe you should read the review lol. its from May 04, 2004. nothing to do with SM3.0.
:p
 
Every and each reviewers now have the nice option of cherry picking their favorite maps for their favorite graphic hardware sponsors in the best lights.
 
What's up with the large performance drops for the NV3x series. X-bit labs has performance dropping upwards to 26% with the new patch. Hope NV3x still has some driver juice in it.
 
You must log in or register to reply here.
Back
Top