Nvidia 6800 series performance?

[Xenome]

I just ordered a XFX 6800 GT, and I would like to know how these cards handle cs:s and the source engine..

So, to all people using these series; how's it going?

 

[DarkStar]

I get a pretty rock solid 90 FPS with my 6800 GT.

 

[Slainchild]

Very well, I guess.

I'm running a 9800 non-pro at 1280x1024, with all settings on high at around 40-50 fps for most maps

 

[Xenome]

Very well, I guess.

I'm running a 9800 non-pro at 1280x1024, with all settings on high at around 40-50 fps for most maps

And what does that have to do with 6800 Card series?

 

[Xenome]

I get a pretty rock solid 90 FPS with my 6800 GT.

Great. What drivers are you using? Good image quality, etc?

And the rest of your system?

 

[Slainchild]

And what does that have to do with 6800 Card series?

I was just saying that if I get that, then you should get a lot more...

 

[Milo 7]

And what does that have to do with 6800 Card series?

well if u would think a little u would see a 9800 is running the game fine and that with a 6800gt it will run even better because its 2 to 3 times better then a 9800

and darkstar what kind of pc do u have because i dont think u are getting 90fps all the time unless u are running alone on the map

 

[Xenome]

I was just saying that if I get that, then you should get a lot more...

Yes, true

But since HL2 is kinda ATI, I was just wondering ..

Right now I still have my ati 9700 pro in my pc, but I upgraded to nvidia 6800 Gt because it's faster in most games then ati, just not in HL2. ( small difference)

 

[Milo 7]

uum the 6800gt is a lot faster in hl2 then a 9700pro if u turn on AA and af and higher rez etc.

i have a x800xtpe and had a 9700pro b4 it

 

[Xenome]

Muhahha I just got the box from the mail-office.. That's one cool box

Should I do a clean format since i'm going from ati to nvidia or not?

 

[Alig]

Nah you should'nt need to format. Just make sure all ati software is removed completely before installing new card.

I went from an nvidia to my x800 and have no problems.

 

[Gary_Jones]

I didnt take the recommended settings as I got a few random "dips in performance".

But all in all it seems this game performs really well with the XFX 6800 GT.

Ive flashed mine to Ultra and have clock rates of 410/1120, for your information.

1280x960 85hz 4xAA 8xAF - quality Vsync off (turning it on seems to really hit performance on this card with source)

GOOD:
I found fps always above 80fps apart from rare occasions on aztec where it would drop to 40ish momentarily.
No slowdown for water effects/lighting etc

BAD:
Seems to be some intermittent problem with textures, that occured on source beta too, models will go white and sometimes you see a wireframe through walls - wallhack style :/
Reflections on water items sometimes dont work as expected.

Both the bad things could quite easily be driver or source related, as opposed to the card - really dont know.

 

[one day....]

I get a pretty rock solid 90 FPS with my 6800 GT.

Complete bullshit.NO WAY is it a rock solid 90,i have a 256 mb 6800 gt and i get about 80-90 when im alone on a map,but on a server with other players etc i get around 70.

 

[Mailman]

6800 GT on an Athlon 64 3500+

I get about 100 fps on the map by myself running 1280x960 with 4xAA.

 

[KBharvester]

Complete bullshit.NO WAY is it a rock solid 90,i have a 256 mb 6800 gt and i get about 80-90 when im alone on a map,but on a server with other players etc i get around 70.

hmmm...i have a 3200+ a64 and a bfg 6800 gt oc and my frames rarely drop below 80 and r usually around 150 and i have everything all the way up aa and af at 4 and 6

the stress test i get 139...

 

[Tenaku]

I have a BFG 6800 GT...

1280x1024, 4xAA, 8xAF, everything turned all the way up (including 'reflect everything' on water)
i get 87fps in the stress test, and it varies from around 160fps to 45fps or so in actual gameplay, depending on what i'm looking at. (water seems to be the biggest slow down)

 

[KBharvester]

oh and i run in 1024...my frames go over 200 alot...

 

[A|2SENIC]

I think that they are very solid cards, get one especially if your going to be playing games besides half-life 2.

 

[Alientank]

I have a BFG 6800 GT...

1280x1024, 4xAA, 8xAF, everything turned all the way up (including 'reflect everything' on water)
i get 87fps in the stress test, and it varies from around 160fps to 45fps or so in actual gameplay, depending on what i'm looking at. (water seems to be the biggest slow down)

Same here. 89 fps in the stress test. 1280x1024 4x aa 8x af, everything as high as it can go including reflections set to reflect everything. Game has no problems, it's very fast. System in sig.

 

[Uhu]

Didnt you guys realize that AA doesnt work with the nvidia cards?

 

[KBharvester]

oh yeah? i was wondering why i was still seeing jaggies..

 

[Uhu]

I have a 6800 GT and was trying the hell out to get FSAA to work. But I googled a little and found some statements that Source engine itself isn capable of "rendering" FSAA on DX9 Nvidia Cards. But it it is said to be possible with a patch. A driver or HL patch? dunno

The framerate drops with activated AA though. The algorithms seem to be run but it wont display.

 

[ShinRa]

Complete bullshit.NO WAY is it a rock solid 90,i have a 256 mb 6800 gt and i get about 80-90 when im alone on a map,but on a server with other players etc i get around 70.

then the rest of your comp must suck, seeing as i havent fallen below 100 fps in cs:s yet with everything maxed out (cept resolution, i prefer 10x7 to 16x12)

 

[Alientank]

I have a 6800 GT and was trying the hell out to get FSAA to work. But I googled a little and found some statements that Source engine itself isn capable of "rendering" FSAA on DX9 Nvidia Cards. But it it is said to be possible with a patch. A driver or HL patch? dunno

The framerate drops with activated AA though. The algorithms seem to be run but it wont display.

Really? It looked pretty good then without AA on! Cause I had it set to 4x aa. I'll turn it off though if it doesn't work.

 

[Uhu]

you can check it it doesnt work. I like it more without AA. The whole image is sharper and above 1024 you dont need it imo =)

 

[Mailman]

That's odd. I'm quite sure that FSAA is running on my 6800 GT. No jaggies here.

 

[Uhu]

That's odd. I'm quite sure that FSAA is running on my 6800 GT. No jaggies here.

Open cs_italy and look at houses and edges. You will see it is not working

 

[Mailman]

When I looked at the edges last night I saw that they were antialiased. I'm at work right now, but I will post some screenshots when I get home.

 

[Milo 7]

then the rest of your comp must suck, seeing as i havent fallen below 100 fps in cs:s yet with everything maxed out (cept resolution, i prefer 10x7 to 16x12)

can i see some screenshots of u with heavy action like lots of shooting etc? with fps counter on

 

[ShinRa]

when i get home. im at school right now.

 

[Uhu]

When I looked at the edges last night I saw that they were antialiased. I'm at work right now, but I will post some screenshots when I get home.

yes would be nice

thats a screenshot from me. Try to get to the same position on CS_ITALY

Heres my specs:
Geforce 6800 GT, 256 MB, A64 3200, 1 GB RAM
1280x1024 everything on highest . Doesnt matter what settings I choose. Edges always look like this. Even with forced AA in the forceware drivers

http://www.megagames.com/news/html/pc/half-life2withoutaa.shtml

 

[fishy2000]

Complete bullshit.NO WAY is it a rock solid 90,i have a 256 mb 6800 gt and i get about 80-90 when im alone on a map,but on a server with other players etc i get around 70.

man i dont know what you all are smoking but i can show ss of me with my 100-140 fps with everyone esle in the server...

my fps never goes down past 85...

nvidia 6800 ultra 256mb 1 gig of reg. ecc and amd 64-fx 51

 

[Milo 7]

man i dont know what you all are smoking but i can show ss of me with my 100-140 fps with everyone esle in the server...

my fps never goes down past 85...

nvidia 6800 ultra 256mb 1 gig of reg. ecc and amd 64-fx 51

well SHOW us the PICS then!!!!! NOW!!!!!!!!!!! :imu:

 

[Xenome]

Ok my card is in now, and guys remember one thing:

USE THE 66.71 DRIVERS!!!

I got 3620 points or something in 3dmark05 with the standard ones from nvidia site, and now I got 4619

Unfortunatly I can't test CS:S yet, because I don't have a credit card.

But what a card this is lol, it even looks beautifull! (card itself)
Gonna play some doom 3 now that came with it.
( even tough I already completed it with my old card)

When HL2 is out I will be playing that of course, don't worry

/me is happy

 

[DarkStar]

and guys remember one thing:

USE THE 66.71 DRIVERS!!!

I got 3620 points or something in 3dmark05 with the standard ones from nvidia site, and now I got 4619

Unfortunatly I can't test CS:S yet, because I don't have a credit card.

I wouldn't be so hasty on using those 66.71 drivers. Nvidia has been optimizing the 66.XX series of drivers specifically to give large boosts in 3Dmark05. Which the new drivers accomplish quite handily. However, for me at least the 66.XX series is terrible when it comes to actually running games. My CS:Source FPS dropped about 10 when I was using those drivers.

3Dmark is not the end all, be all yardstick for performance. It's generally better to do your own benchmarking with games that you actually PLAY.

 

[Xenome]

I wouldn't be so hasty on using those 66.71 drivers. Nvidia has been optimizing the 66.XX series of drivers specifically to give large boosts in 3Dmark05. Which the new drivers accomplish quite handily. However, for me at least the 66.XX series is terrible when it comes to actually running games. My CS:Source FPS dropped about 10 when I was using those drivers.

3Dmark is not the end all, be all yardstick for performance. It's generally better to do your own benchmarking with games that you actually PLAY.

Well with the standard drivers I got black triangles or lines in 3dmark05, now with 66.71 they are gone.
Somebody else on futuremark.com had the same problem.

The standard drivers just suck with 3dmark05.

Further more, I can't comment on fps in games, because I haven't tested, but doom III is running NICE..

 

[nyorker]

my rig

amd 64 3200, 1 gig ram, gainward gef 6800 128 meg.. getting 35-100 fps on all maps. i usually run around 70 fps,,, on full maps with action i can go as low as 40fps maybe even 35.. but its playable. im at 1024 with 6x and 16x , everything on hi.. is this good??? i thought i would get no lower than 70 but since its not a gt or ultra card i wont worry about it.. my cpu should be fast enough.. sometimes the game does lag, but i think its network issues...

 

[Milo 7]

so does anybody have their screenshots yet with 100+fps during heavy action?

 

[Alientank]

What settings should we be using for drivers for a 6800 GT? I have the 66.70's, and AA and AF are set to application controlled, Image settings are set on "quality", vsync is off, force mipmaps is off, texture clamp is on, extension limit is off, trilinear optimization is on, anisotrophic mip filter optimization is off, and anisotrophic sample optimization is off. Is this how I should have it all set? Basically looking for the best IQ but running as fast as it can, and I know high quality mode is no better looking then quality is but is a huge performance gap. Does this all look right or should I change it? Oh I'm running the game max details 4x aa 8x af if that's needed to be known.

 

[Uhu]

What settings should we be using for drivers for a 6800 GT? I have the 66.70's, and AA and AF are set to application controlled, Image settings are set on "quality", vsync is off, force mipmaps is off, texture clamp is on, extension limit is off, trilinear optimization is on, anisotrophic mip filter optimization is off, and anisotrophic sample optimization is off. Is this how I should have it all set? Basically looking for the best IQ but running as fast as it can, and I know high quality mode is no better looking then quality is but is a huge performance gap. Does this all look right or should I change it? Oh I'm running the game max details 4x aa 8x af if that's needed to be known.

Make a screenshot at 6x AA then turn it off and make a SS at 0x AA.
Compare them:
come to the solution that there is no difference with nvidia cards