Widescreen Gaming Forum

Just wanted to see what kind of experience everyone is having with the cat 10.2

I'm running [email protected] with 8Gb Ram
2X HIS Radeon 5850.
Windows 7 Ultimate x64

Haven't seen any real difference in crossfire eyefinity performance. I still have horrible stuttering in cross-finity (see what i did there?). Anyone have better experience?

I went back to 10.1 coz of crossfire problems in some games, performance in Vantage was about the same though.

good in the games i am currently playing (dirt2, BFBC2). Crysis not happy (stutter) and also, tried to force DX9 in BFBC2, massive flicker (worked in 9.12hf, but no CF).

also, side note, my mouse is working in CS:S and Crysis under Steam (previously, no mouse pointer unless windowed)

ran overdrive auto-tune, came back with 895/1300, however, massive flicker. back to stock clocks which is fine with me really.

Just wanted to see what kind of experience everyone is having with the cat 10.2

I'm running [email protected] with 8Gb Ram
2X HIS Radeon 5850.
Windows 7 Ultimate x64

Haven't seen any real difference in crossfire eyefinity performance. I still have horrible stuttering in cross-finity (see what i did there?). Anyone have better experience?

I have the exact same issue with 2x5870s... With a single 5870 everything run great but with cross-finity huge stuttering issue despite the frame boost! Vsync also doesn't help either.

The proper solution for me is to sell the 5870s and the adapter and get 2xfermi cards that will definitely work with 3 displays.

my performance in source games is really strange. Its not microstuttering, but its stuttering really bad. For example, in Team Fortress2 I am running at 5760x1200 and I can tell the performance is really good, but its like I'm dropping about 5 frames of animation every 1/2 second. I can run with 8x AA on a single 5850 so crossfire should be no problem, but its bizarre. I can adjust the AA levels all I want and it makes no difference. A similar problem happens in GRID, if I try to run any higher than 2xAA then the game grinds to a halt. Not sure if the driver is bringing the second 5850 out of reduced power mode. Any thoughts?

I'm a fan, they fixed my two biggest gripes:

1> Main monitor won't properly go into sleep mode (does now)
2> Eyfinity profiles won't load correctly (do now!)

Hooray. I have extensive benchmark results I could post too if people want; I've got a benchmark script that I put driver releases through their paces through, it's just that there is alot of data... Overall this driver was a touch slower than 10.1, but not enough to be noticeable.Note that I'm not running crossfire.

I found sort of a fix. Like I mentioned earlier, I primarily play Source and Codemasters games. I found that if I disable AA in TF2, and enable it in the CCC my performance in Cross-Finity (EyeFire?) is greatly improved. I still get a little bit of stuttering, but this is worlds apart from where the stuttering was greatly affecting playability. Also this seems to be most important

ENABLE VSYNC IN CCC. This has been the biggest performance gain for me. Again I'm using two 5850 crossfired and eyefinitied so maybe this just works for me. Please don't be afraid to try though.

Right now I've got TF2 running 5760x1200 with 16x AA and 16x AF enabled through CCC - in game settings have AA disabled and Trilinear filtering.

GRID is running 5760x1200 with 4x AA and I'm shooting to get 8X AA enabled. Strangely it seems that you can't interrupt the opening animated sequence otherwise the loading screen grinds to slide show speeds.

I found sort of a fix. Like I mentioned earlier, I primarily play Source and Codemasters games. I found that if I disable AA in TF2, and enable it in the CCC my performance in Cross-Finity (EyeFire?) is greatly improved. I still get a little bit of stuttering, but this is worlds apart from where the stuttering was greatly affecting playability. Also this seems to be most important

ENABLE VSYNC IN CCC. This has been the biggest performance gain for me. Again I'm using two 5850 crossfired and eyefinitied so maybe this just works for me. Please don't be afraid to try though.

Right now I've got TF2 running 5760x1200 with 16x AA and 16x AF enabled through CCC - in game settings have AA disabled and Trilinear filtering.

GRID is running 5760x1200 with 4x AA and I'm shooting to get 8X AA enabled. Strangely it seems that you can't interrupt the opening animated sequence otherwise the loading screen grinds to slide show speeds.

I enabled AA (x4) in CCC and VSYNC also. My FPS do not appear to be much different, but, things do seem to be much smoother. Going to play more.

Thanks for the suggestions

These drivers work okay with a single 5970 for eyefinity but add a 5870 to the mix for some tri-fi and eyefinity doesn't work as the fps TANK beyond belief and it becomes nothing more than a slide show.

Tri-fi 5970 + 5870 works fine with a single monitor but is utterly pathetic when eyefinity is enabled. :roll: :cry:

No new resolutions. Going from 10.1 to 10.2 gave me 4-12fps loss in Crysis and GTA IV in Eyefinity. That alone, is enough to go back to 10.1.

I've never had any flicker/line issues, so I cun't speak to that.

Monitor goes to sleep correctly now. Hard to believe that that would have made it through QA on 10.1 . I've only played GTA IV and have not noticed any changes good or bad.

I'm going to post this because after looking through I'm a little concerned about the performance loss on some older games. This is the biggest % drop I've seen from a driver release in awhile, but interestingly enough alot of the newer games performance is relatively untouched. I've used this script for 10 years now adding and removing new games as "necessary". If it's too wordy pls let me know and I'll edit it down.

There are actually multiple runs for each game, but I took those numbers out because it's too much info for a post. I'm also taking out the system information relayed in each log file to conserve space, see my sig for system specs. I have results from 4x AA/16x Aniso, 4x/0x, 0x/16x, 0x/0x all at 1600x1200. (4:3 because it's easier to script for older games, but the pixel count is still in a good midrange)

First 10.1 @ 0x/0x:

Benchmark Log File created on 1/27/2010 at 3:26:16 PM
Tests performed at 1600x1200x32, 0x FSAA & 0x Anistropic (Quality setting)
********************************************************************************
Results for Quake3:
AVERAGE FPS: 870.4
***************************************
Results for UT2004FlyBy:
AVERAGE FPS: 616.8
***************************************
Results for UT2004Botmatch:
AVERAGE FPS: 120.8
***************************************
Results for Comanche4:
AVERAGE FPS: 160.8
***************************************
Results for HalfLife2:
AVERAGE FPS: 551.3
***************************************
Results for SeriousSam2:
AVERAGE FPS: 341.1
***************************************
Results for Doom3:
AVERAGE FPS: 323
***************************************
Results for HalfLife2_Lost_Coast:
AVERAGE FPS: 317.6
***************************************
Results for Supreme_Commander:
AVERAGE FPS: 86.05
***************************************
Results for Company_of_Heroes_DX9:
AVERAGE FPS: 370.25
***************************************
Results for Company_of_Heroes_DX10:
AVERAGE FPS: 172.8
***************************************
Results for Crysis_Island_DX9_HIGH:
AVERAGE FPS: 79.84
***************************************
Results for Crysis_Island_DX10_VERY_HIGH:
AVERAGE FPS: 44.32
***************************************
Results for Dirt2_DX9_High:
AVERAGE FPS: 123.07
***************************************
Results for Dirt2_DX11_Ultra:
AVERAGE FPS: 71.61
***************************************
Benchmarks finished at 4:20:33 PM

******************************************************************************
Now 10.2

Benchmark Log File created on 2/17/2010 at 3:54:42 PM
Tests performed at 1600x1200x32, 0x FSAA & 0x Anistropic (Quality setting)
********************************************************************************
Results for Quake3:
AVERAGE FPS: 836.15
***************************************
Results for UT2004FlyBy:
AVERAGE FPS: 589.3
***************************************
Results for UT2004Botmatch:
AVERAGE FPS: 121.3
***************************************
Results for Comanche4:
AVERAGE FPS: 158.7
***************************************
Results for HalfLife2:
AVERAGE FPS: 550.4
***************************************
Results for SeriousSam2:
AVERAGE FPS: 340.6
***************************************
Results for Doom3:
AVERAGE FPS: 305.6
***************************************
Results for HalfLife2_Lost_Coast:
AVERAGE FPS: 318.5
***************************************
Results for Supreme_Commander:
AVERAGE FPS: 86.19
***************************************
Results for Company_of_Heroes_DX9:
AVERAGE FPS: 368.8
***************************************
Results for Company_of_Heroes_DX10:
AVERAGE FPS: 175
***************************************
Results for Crysis_Island_DX9_HIGH:
AVERAGE FPS: 79.75
***************************************
Results for Crysis_Island_DX10_VERY_HIGH:
AVERAGE FPS: 44.81
***************************************
Results for Dirt2_DX9_High:
AVERAGE FPS: 121.86
***************************************
Results for Dirt2_DX11_Ultra:
AVERAGE FPS: 72.49
***************************************
Benchmarks finished at 4:49:51 PM

********************************************************************************
10.1 @ 4x/16x:
Tests performed at 1600x1200x32, 4x FSAA & 16x Anistropic (Quality setting)
********************************************************************************
Results for Quake3:
AVERAGE FPS: 842.5
***************************************
Results for UT2004FlyBy:
AVERAGE FPS: 569.2
***************************************
Results for UT2004Botmatch:
AVERAGE FPS: 119.9
***************************************
Results for Comanche4:
AVERAGE FPS: 158
***************************************
Results for HalfLife2:
AVERAGE FPS: 434.7
***************************************
Results for SeriousSam2:
AVERAGE FPS: 250.1
***************************************
Results for Doom3:
AVERAGE FPS: 306.6
***************************************
Results for HalfLife2_Lost_Coast:
AVERAGE FPS: 247.9
***************************************
Results for Supreme_Commander:
AVERAGE FPS: 82.22
***************************************
Results for Company_of_Heroes_DX9:
AVERAGE FPS: 314.5
***************************************
Results for Company_of_Heroes_DX10:
AVERAGE FPS: 148.55
***************************************
Results for Crysis_Island_DX9_HIGH:
AVERAGE FPS: 54.11
***************************************
Results for Crysis_Island_DX10_VERY_HIGH:
AVERAGE FPS: 34.16
***************************************
Results for Dirt2_DX9_High:
AVERAGE FPS: 98.36
***************************************
Results for Dirt2_DX11_Ultra:
AVERAGE FPS: 60.17
***************************************
Benchmarks finished at 3:26:07 PM

********************************************************************************
10.2 @ 4x/16x
Tests performed at 1600x1200x32, 4x FSAA & 16x Anistropic (Quality setting)
********************************************************************************
Results for Quake3:
AVERAGE FPS: 799.1
***************************************
Results for UT2004FlyBy:
AVERAGE FPS: 549.15
***************************************
Results for UT2004Botmatch:
AVERAGE FPS: 118.7
***************************************
Results for Comanche4:
AVERAGE FPS: 156.9
***************************************
Results for HalfLife2:
AVERAGE FPS: 432.1
***************************************
Results for SeriousSam2:
AVERAGE FPS: 247.2
***************************************
Results for Doom3:
AVERAGE FPS: 294.3
***************************************
Results for HalfLife2_Lost_Coast:
AVERAGE FPS: 247.5
***************************************
Results for Supreme_Commander:
AVERAGE FPS: 82.38
***************************************
Results for Company_of_Heroes_DX9:
AVERAGE FPS: 308.45
***************************************
Results for Company_of_Heroes_DX10:
AVERAGE FPS: 149.5
***************************************
Results for Crysis_Island_DX9_HIGH:
AVERAGE FPS: 53.98
***************************************
Results for Crysis_Island_DX10_VERY_HIGH:
AVERAGE FPS: 34.76
***************************************
Results for Dirt2_DX9_High:
AVERAGE FPS: 100.46
***************************************
Results for Dirt2_DX11_Ultra:
AVERAGE FPS: 60.93
***************************************
Benchmarks finished at 3:54:33 PM

***************************************

Again, sorry for the long post. Hopefully this info will be useful to someone. (Dave? ;-D )
Edit: Compressed post more.

Oh no! Quake 3 now runs at 800 fps instead of 842 fps! :lol:

Oh no! Quake 3 now runs at 800 fps instead of 842 fps! :lol:

It's fun to compare with hardware from 10 years ago. I have Quake 2 in there too if you want that. :wink:

Edit: and actually the reason I look at that stuff is because (and I'm no driver programmer) I suspect some of those things may speak to lower-level driver code. I'd hypothesize that if the cards aren't tied up doing much of anything but setup and texturing, results like those speak to the efficiency of the base elements of what needs to be done for every frame. AAAANNNNDDD it's fun to compare with stuff like this:

Benchmark Log File opened on 1/26/01 at 3:48:31 PM
System Specifications:
CPU Speed:Pentium 3 1031(128x8) Motherboard:Abit BE6-II
Memory: 384MB @ 2-2-2 Video: Voodoo 5 @ 185Mhz 1.04.01
Audio: SB Live! Value w/ Liveware 3
All tests performed at 800x600x16 w/ 2x FSAA unless otherwise noted

Results for Quake 3:
AVERAGE FPS: 58.17

on 1 hand I think I finally fixed my monitor flickering on and off with the active adapter.. on the other hand when crossfire is enabled in eyefinity mode I am getting screen tearing that drives me nuts.. thats all I noticed so far. Eyefinity=win but crossfire eyefinity=fail... very frustrating and so far up to cat 10.2 its been a waste going the crossfire route.

I've been playing around all day and these new driver seem to have something in them that is giving more support for reading monitor info thru an apple DP adapter.

Well, after reading this looks like I'll stay with the 9.12 hotfix once again. Experienced frame drop in eyefinity crossfire with 10.1 and looks like it is the same with everyone here with 10.2 . Oh well . One of these days improvement is bound to happen.

Is anyone here running crossfire and eyefinity with no stuttering or screen tearing with any of the catalyst drivers?

Looks like 10.3 could potentially bring very large performance improvements to either DX11 or DiRT2 in DX11 specifically...

http://www.widescreengamingforum.com/forum/viewtopic.php?f=64&t=18459&p=184726#p184726