Geforce 8800gtx overclock problems with ati tool?

[Dakir]

Hello,
I was interested in some feedback on why my 8800gtx cant complete 30 seconds of ati tool at any core/shader/memory speed. An interesting fact is even on stock speed I gets spammed with tiny artifacts about 15seconds into the test. Another very interesting fact is if i run ati tool scanning for artifacts and cover up the window, like im doing right now it can go 10mins without a single artifact but when the actual window is shown it has problems.

8800gtx on a nforce 590i motherboard, all stock voltages runs about 52 degrees celcius idle, and 60 degrees celcius during gaming with fan @ 100% speed. Using rivatuner to adjust the speeds as well.


Any suggestions would help. I also cannot take the 8800gtx out and try it on another computer due to not having another computer with a pci-e motherboard.

Thanks

 

[Maziar]

Hello and welcome to the forums mate
Have u tried working with RIVATUNER ?

 

[Dakir]

Hello and welcome to the forums mate
Have u tried working with RIVATUNER ?

Yes, i use Rivatuner to adjust the speed. Only use ati tool to scan for artifacts and measure temps.

 

[V3NOM]

does it artifact during intensive games (crysis perhaps?)

 

[Dakir]

does it artifact during intensive games (crysis perhaps?)

Do you mean running ati tool while im playing crysis? Or do I notice any deformities while playing crysis?

 

[dagger]

Don't use a game to test for stability/artifacting. Use this instead.
http://www.ozone3d.net/benchmarks/fur/
Stress mode, default settings, no aa/af.

 

[Dakir]

Don't use a game to test for stability/artifacting. Use this instead.
http://www.ozone3d.net/benchmarks/fur/
Stress mode, default settings, no aa/af.

Ran that at stock speeds, no visual artifacts. 30minimum, 35average, 40maximum fps @ 1280x1024 on the stability test for 2mins. Also got my 8800gtx up to 72 degrees. I also was wondering if drivers could cause the artifacts, Im using geforce 177.92 drivers.

 

[dagger]

Ran that at stock speeds, no visual artifacts. 30minimum, 35average, 40maximum fps @ 1280x1024 on the stability test for 2mins. Also got my 8800gtx up to 72 degrees. I also was wondering if drivers could cause the artifacts, Im using geforce 177.92 drivers.

Temperatures are fine, but performance seems strange. My 8800gts get 50 min, 77max at the same settings. Maybe the new driver isn't supporting the old card properly? If I remember correctly 177.92 is a beta. I'm running on the same driver though, recently upgraded from 169 to try physx.

 

[Dakir]

Temperatures are fine, but performance seems strange. My 8800gts get 50 min, 77max at the same settings. Maybe the new driver isn't supporting the old card properly? If I remember correctly 177.92 is a beta. I'm running on the same driver though, recently upgraded from 169 to try physx.

You are correct, it is a beta driver. What version would you suggest installing since my performance is lower than average for some reason? 8800gtx should get higher scores then a 8800gts I believe...

What about forceware drivers? Currently I'm using version 120.60.

 

[dagger]

You are correct, it is a beta driver. What version would you suggest installing since my performance is lower than average for some reason? 8800gtx should get higher scores then a 8800gts I believe...

What about forceware drivers? Currently I'm using version 120.60.

Not sure. Aren't they all Forceware though? 120.60 seems really old...

 

[Dakir]

I actually tried a mixture of drivers, did some monitoring and benchmarking and I get around the same scores. All of these tests were done on the benchmark test 1280x1024 0aa/0af (1min duration) with my clocks set at: 648core, 1560shader, 1000memory on my 8800gtx.

Heres the rundown on drivers/performance:

169.21 GeForce/ 175.19 GeForce/ 177.92(beta) GeForce/
111.32 Forceware/ 111.75 Forceware/ 120.60 Forceware/
40min/ 39min/ 39min/
60max/ 59max/ 59max/
46average/ 45average/ 45average/
2750 total score/ 2681 total score/ 2684 total score/


What I cant understand is why my gtx is getting lower scores then your gts. Could it be that my nforce 590i is slowing it down? I even tried overclocking the pci-e bus from 2500stock to 2650 and I didnt see any improvement in any score. Could it be my stock voltage is set to low and my 8800gtx is getting throttled performance? I'm so lost...

 

[dagger]

I actually tried a mixture of drivers, did some monitoring and benchmarking and I get around the same scores. All of these tests were done on the benchmark test 1280x1024 0aa/0af (1min duration) with my clocks set at: 648core, 1560shader, 1000memory on my 8800gtx.

Heres the rundown on drivers/performance:

169.21 GeForce/ 175.19 GeForce/ 177.92(beta) GeForce/
111.32 Forceware/ 111.75 Forceware/ 120.60 Forceware/
40min/ 39min/ 39min/
60max/ 59max/ 59max/
46average/ 45average/ 45average/
2750 total score/ 2681 total score/ 2684 total score/


What I cant understand is why my gtx is getting lower scores then your gts. Could it be that my nforce 590i is slowing it down? I even tried overclocking the pci-e bus from 2500stock to 2650 and I didnt see any improvement in any score. Could it be my stock voltage is set to low and my 8800gtx is getting throttled performance? I'm so lost...

As long as the motherboard use pciex16, it shouldn't bottleneck. Do not overclock pcie bus, lock it at 100mhz. Otherwise it may cause problems. And don't change pcie voltage, it's not healthy for the card. You can't change the gpu voltage without volt modding anyway.

It might just be Furmark being skewed. It runs on opengl instead of directx, and is meant for stressing, not really benchmarking. Try 3dmark instead.
http://downloads.guru3d.com/3DMark06-v110-Basic-Edition_d1297.html

There are also game benchmark frontend tools you can use, like Crysis and UT3.
http://downloads.guru3d.com/Crysis-Benchmark-Tool-1.05-Final-download-1791.html
http://downloads.guru3d.com/Unreal-Tournament-3-Benchmark-Tool-download-1816.html

The point is, try to confirm that there is actually a problem before doing things that may possibly mess up a perfectly good system when there's nothing wrong in the first place.

Besides, if it's only slightly below, it really doesn't matter. It's not like you'll notice a difference anyway. Also, my8800gts is overclocked. You're better off comparing to someone with a stock 8800gtx like yours.

My 3dmark06 results, for reference:

 

[Dakir]

This is what my results came out to, does everything look ok to you?

 

[dagger]

This is what my results came out to, does everything look ok to you?

http://i112.photobucket.com/albums/n200/Dakir/Scores.jpg

Everything on gpuz looks perfectly normal, yet the 3dmark scores are more like a stock 8800gt. Your 8800gtx is overclocked nearly 100mhz too. It's strange how that card is oced past 8800 Ultra level, and still get that kind of performance.

Maybe 3dmark06 is skewed? Have you tried Crysis benchmark?

 

[Dakir]

Everything on gpuz looks perfectly normal, yet the 3dmark scores are more like a stock 8800gt. Your 8800gtx is overclocked nearly 100mhz too. It's strange how that card is oced past 8800 Ultra level, and still get that kind of performance.

Maybe 3dmark06 is skewed? Have you tried Crysis benchmark?

No idea if it is, i'll try crysis benchmark later. Its strange I pay for an 8800gtx and get performance of a card 1/4 its price...

 

[Dakir]

Downloaded and ran that crysis benchmark and heres what I got, this is also with my card overclocked at 660 core, 1560 shader and 1045 memory.

9/8/2008 12:56:22 AM - XP
Beginning Run #1 on Map-island, Demo-benchmark_gpu
DX9 1280x1024, AA=No AA, Vsync=Enabled, 32 bit test, FullScreen
Demo Loops=3, Time Of Day= 9
Global Game Quality: High
==============================================================
TimeDemo Play Started , (Total Frames: 2000, Recorded Time: 111.86s)
!TimeDemo Run 0 Finished.
Play Time: 51.67s, Average FPS: 38.71
Min FPS: 17.55 at frame 143, Max FPS: 53.69 at frame 970
Average Tri/Sec: 38127312, Tri/Frame: 984939
Recorded/Played Tris ratio: 0.93
!TimeDemo Run 1 Finished.
Play Time: 43.81s, Average FPS: 45.66
Min FPS: 17.55 at frame 143, Max FPS: 57.15 at frame 997
Average Tri/Sec: 45461740, Tri/Frame: 995746
Recorded/Played Tris ratio: 0.92
!TimeDemo Run 2 Finished.
Play Time: 43.60s, Average FPS: 45.87
Min FPS: 17.55 at frame 143, Max FPS: 57.15 at frame 997
Average Tri/Sec: 45707852, Tri/Frame: 996501
Recorded/Played Tris ratio: 0.92
TimeDemo Play Ended, (3 Runs Performed)
==============================================================

Completed All Tests

<><><><><><><><><><><><><>>--SUMMARY--<<><><><><><><><><><><><><>

9/8/2008 12:56:22 AM - XP

Run #1- DX9 1280x1024 AA=No AA, 32 bit test, Quality: High ~~ Overall Average FPS: 45.765

 

[dagger]

That's actually pretty good. 45 fps is perfectly playable. The only problem is min fps at 17. It'll be annoying to feel that lag from time to time. It's normal for first run to be slower, due to content loading.

Still, all the benchmarks showed is basically your 8800gtx that's been oced way past 8800ultra is not performing as it should. It doesn't pinpoint whatever problem that may be there. Sorry if this turns out to be just a waste of time.

My results, same settings, except in 64bit instead of 32bit, as I use Vista x64.

9/8/2008 6:45:01 PM - Vista 64
Beginning Run #1 on Map-island, Demo-benchmark_gpu
DX9 1280x1024, AA=No AA, Vsync=Enabled, 64 bit test, FullScreen
Demo Loops=3, Time Of Day= 9
Global Game Quality: High
==============================================================
TimeDemo Play Started , (Total Frames: 2000, Recorded Time: 111.86s)
!TimeDemo Run 0 Finished.
Play Time: 42.51s, Average FPS: 47.05
Min FPS: 36.66 at frame 174, Max FPS: 57.86 at frame 875
Average Tri/Sec: 46305128, Tri/Frame: 984225
Recorded/Played Tris ratio: 0.93
!TimeDemo Run 1 Finished.
Play Time: 39.35s, Average FPS: 50.82
Min FPS: 36.66 at frame 174, Max FPS: 58.54 at frame 1021
Average Tri/Sec: 50592176, Tri/Frame: 995518
Recorded/Played Tris ratio: 0.92
!TimeDemo Run 2 Finished.
Play Time: 39.29s, Average FPS: 50.90
Min FPS: 36.66 at frame 174, Max FPS: 58.90 at frame 729
Average Tri/Sec: 50712164, Tri/Frame: 996227
Recorded/Played Tris ratio: 0.92
TimeDemo Play Ended, (3 Runs Performed)
==============================================================

Completed All Tests

<><><><><><><><><><><><><>>--SUMMARY--<<><><><><><><><><><><><><>

9/8/2008 6:45:01 PM - Vista 64

Run #1- DX9 1280x1024 AA=No AA, 64 bit test, Quality: High ~~ Overall Average FPS: 50.86

 

[Dakir]

Could it be that since its not plugged into a pci-e 2.0 socket its not running full speed? Other then that... I really have no idea why its running sluggish compared to other 8800gtx's...

 

[dagger]

Could it be that since its not plugged into a pci-e 2.0 socket its not running full speed? Other then that... I really have no idea why its running sluggish compared to other 8800gtx's...

Can't be pcie2.0. 8800gtx doesn't even support pcie2.0. So if you plug it into a new motherboard with pcie2.0 slot, it'll just run at the lower bus speed anyway.

As for other 8800gtxs, I have 8800gts, and no one has yet came with 8800gtx for comparison. Maybe it just seems poor but is really normal? I don't think min fps dropping to 17 fps can be normal though. It's just far too big a difference.

 

[Dakir]

Yeah, it runs crysis at that resolution fine, but there are random spikes like that every once in a wile. Maybe I just have a bummed 8800gtx.

 

[redgar]

I think theres something wrong with benchmarking the GTX pretty much across the board. I have had problems with benchmarks in every program I have run on my BFG 8800gtx OC. I get the artifact bug at stock speed, and lower scores on Furmark and 3dMark than people running the cheaper 8800GT.

I am starting to believe there is some sort of issue with benching these things without the results becoming skewed. I have not come across a game yet, including Crysis, that I couldn't run flawlessly on max settings without lag or hitching on my 8800gtx. Yet somehow, it always looks like a pos on benchmarks.

Wish I knew what the deal was... and wish I could find another artifact test like ATI tool that worked. I can't go more than 10 seconds without supposed artifact errors with no overclock.

 

[assasin32]

Go into settings, under artifact scanning and use the 'old method' which says its older and more capatible method. Thats what I used when OCing my 320mb 8800gts, seems to have worked fine for me. It took me forever to find out they had that.

 

[smasher]

My Settings Are The Same 648/1550/1000 My 06 Score
http://i94.photobucket.com/albums/l102/smasher_05/PC%20Photos/06Mark13378.jpg on X6800 Extreme

 

[redgar]

Ya, the biggest difference in our scores is the processor. My friend with the 8400 wolfdale @ 3.6 scored about the same as you also with the 8800gtx running 630/950.

I score a 12,600 with my e6600 @3.2 and my 8800gtx @ 630/1474/975. I find I can't run those video settings reliably in that I get a freeze up from time to time.

I do 615/1439/950 for games atm.

 

[Dakir]

With my q6600 @ 2.6ghz and 8800gtx @ 660 core, 1560 shader and 1045 memory and using the new 180 nvidia beta drivers I only get 12900 tops. Not even comparing cpu scores just graphic card scores and mine is the lowest by a good deal and also overclocked the most. Also using windows media center as my OS 32bit.