295 gtx artifacts

[richard_ab]

Hi so basically a month or so ago i bought a new 295 gtx to replace my old 7900gs and im just not happy with it. I get artifacting in nearly all my games these are Company of heroes, Red alert 3, Need for speed shift, World of warcraft, crysis and in benchmarking programs like 3D mark 06 and 3d Mark vantage, sometimes the artifacts persist even after closing the game and appears on windows as loads of dots. Some times it blue screens for me or crashes. I also get errors reporting the nvidia display driver stopped repsonding .... after crashes.

This is my set up

Motherboard: GA-ex58-DS4 (gigabyte)
CPU: Intel i7 920
RAM corsair DDR3 6 gb
PSU: 780w Hiper
Graphics card = BFG NVIDIA GeForce GTX 295 1792MB PCIe 2.0 (B)
OS: Windows 7 64bit RC
Driver 191.07 (this is whql certified and supports windows 7 which im running)

Now i know its not a temperature problem as my gpu doesn't go over 70 (and gives me problems at even 55) unless im running furmark stability test.

When i run the following programs i don't have artifacts, ATI tool test, occt GPU test (i do get errors listed when i run the GPU memtest dunno if its related). Furmark gpu stability test and for some reason i dunno why batman arkham asylum runs well apart from the fact that it freezes every 2 minutes which i must press ctrl+alt+delete to get out of and get back in the game (oddly enough same thing for mirrors edge except mirrors edge has artifacting on the skin). Now i took it back to the store and they said the problem is SLI and that its 'too new for the current games' which sounds like a load of bull sh** to me (sli could be a problem but not that the VGA is too new lol after all reviews were done on CoH and i haven't seen them mention errors..). I have also ran my vga on my dads pc who has a very similar build except for his motherboard is an ASUS ramage extreme 2 and he gets artifacts as well.

3D mark programs when they crash for me give me an errors saying something like direct 3d error device removed i dunno if that is related ...

Last thing, the artifacts do improve when i disabled multi gpu which yes it points to SLI but SLI is a few years old now and games like NFS shift or crysis or the 3d mark programs wouldn't give me problems as they are SLI compatible right?

I think i have given enough details hehe (ye i know i over did it lol) i just want to fix this because as u all know this card ain't cheap and i feel like i wasted my money....


Thanks

 

[richard_ab]

BTW i have also formatted my pc and tried different drivers but nothing worked, i used driver sweeper to remove the drivers

 

[drums101]

Try using driver sweeper while in safe mode....If that dont help I would RMA the card to get a replacement. If the temps are what u say they are then there is nothing physicall wrong with the cooling the new gpus can handle up to 100c before problems start to occur.

 

[tsd16]

Hi so basically a month or so ago i bought a new 295 gtx to replace my old 7900gs and im just not happy with it. I get artifacting in nearly all my games these are Company of heroes, Red alert 3, Need for speed shift, World of warcraft, crysis and in benchmarking programs like 3D mark 06 and 3d Mark vantage, sometimes the artifacts persist even after closing the game and appears on windows as loads of dots. Some times it blue screens for me or crashes. I also get errors reporting the nvidia display driver stopped repsonding .... after crashes.

This is my set up

Motherboard: GA-ex58-DS4 (gigabyte)
CPU: Intel i7 920
RAM corsair DDR3 6 gb
PSU: 780w Hiper
Graphics card = BFG NVIDIA GeForce GTX 295 1792MB PCIe 2.0 (B)
OS: Windows 7 64bit RC
Driver 191.07 (this is whql certified and supports windows 7 which im running)

Now i know its not a temperature problem as my gpu doesn't go over 70 (and gives me problems at even 55) unless im running furmark stability test.

When i run the following programs i don't have artifacts, ATI tool test, occt GPU test (i do get errors listed when i run the GPU memtest dunno if its related). Furmark gpu stability test and for some reason i dunno why batman arkham asylum runs well apart from the fact that it freezes every 2 minutes which i must press ctrl+alt+delete to get out of and get back in the game (oddly enough same thing for mirrors edge except mirrors edge has artifacting on the skin). Now i took it back to the store and they said the problem is SLI and that its 'too new for the current games' which sounds like a load of bull sh** to me (sli could be a problem but not that the VGA is too new lol after all reviews were done on CoH and i haven't seen them mention errors..). I have also ran my vga on my dads pc who has a very similar build except for his motherboard is an ASUS ramage extreme 2 and he gets artifacts as well.

3D mark programs when they crash for me give me an errors saying something like direct 3d error device removed i dunno if that is related ...

Last thing, the artifacts do improve when i disabled multi gpu which yes it points to SLI but SLI is a few years old now and games like NFS shift or crysis or the 3d mark programs wouldn't give me problems as they are SLI compatible right?

I think i have given enough details hehe (ye i know i over did it lol) i just want to fix this because as u all know this card ain't cheap and i feel like i wasted my money....


Thanks

Use a temp monitor and ATI tool. Your card in particular may have trouble handling that temp. I know it sounds crazy because that temp is generally fine in terms of video cards, but in my recent experience testing my BFG 285GTX OC, its gotta be the case.

For example, i was getting artifacting, and driver errors after less than 5mins of play on any game(less demanding games may take about 10mins). I went through the driver/bios/chipset/power merry go round, and dismissed heat as an issue because it wasnt getting very hot.

But I decided to do some testing anyhow. Pulled up a couple temp monitors ATITool and 3dmark.

on automatic fan settings 3dmark was failing due to a driver error before it could finish(my temp wasnt breaking 70 mind you). Cranked my fan speed up to max and it made it through 3dmark just fine.

On to ati tool on automatic fan settings, I ran the artifact scanner. within about 1 min (when the temp broke 60) it started to artifact badly. Turned it off, cranked the fan speed up, let it cool down. Started Artifact scanner up again, it ran fine for about 15 minutes, until the temp started breaking 60 then began artifacting. I repeated the process with both 3dmark and ati tool, to ensure that it was linked to the temp and not a freak occurance. Just as previously tested, once the temp breaks 60 all hell breaks loose. The final test was in games, as predicted it would quickly show issues and fail on automatic fan speed, and last significantly longer when the fan speed was cranked up to max.

Obviously it is not a simple software/driver issue, it has got to be linked to heat despite the fact the heat threshold its failing at sure as hell isnt hot in terms of video cards. I think i proved the correlation to be pretty clear, that it failed sooner at a lower fan speed when it hit 60c, and took much longer to fail at max speed, in fact it wouldnt fail until it hit 60c.

Try doing what I did and see if you get similar results. If it is a driver issue, the temperature as long as its within reason, should be irrelevant to when the card fails. i.e. if its a driver issue, I doubt you would only start seeing problems at a certain temp whereas its completely fine if it is below that temp.
I will actually be returning my card and getting a new one, the folks I spoke to at bfg were not exactly friendly. Furthermore, I found out they factory overclocked this card BEYOND what is specified on the box, which could be causing further instability.

 

[nforce4max]

Check the volts on the 12v rails to see if they are dipping while in use and if they are above 11.5v or greater RMA it (the card).

 

[richard_ab]

Thanks for the help both of you, will test it soon. Regarding the temperature i get your point but im not sure how much the problem is temperature as there is no specific threshold at which artifcating starts to occur but i will try what u said. Also i will check with the voltage its just im not sure exactly how, i have CPUIDHWmonitor installed just not sure which voltage is the one i need to look out for.

Thanks

 

[richard_ab]

Ok so i ran ati tool with speedfan to monitor my temp, I got yellow artifacts instant (at a temp below 50 around 48for gpu2 and 44 for gpu1) they were idling at 42. Odd thing is ati tool artifact scanner doesn't detect them. i closed the scanner and started again straight away and straight away i get the error : Display driver NVIDIA Windows Kernel Mode Driver, version 191.07 stopped responding and has succesfully recovered. So i guess its software related. By the way this error came when the temperature hit 58 degrees in gpu2 and gpu 1 was 48. Could the difference in temperatures be related maybe one gpu is damaged which is why i get better results in Multigpu disabled? regarding the voltages i still haven't figured out how to check if someone could explain please

Thanks.

 

[nforce4max]

Well it's time to make use of that warranty and RMA. It shouldn't artifact at such low temps.

 

[richard_ab]

Anyone else has any ideas which might help?

 

[richard_ab]

Could the problem be PhysX?

 

[ridic23]

Nope.....Physx probs let games crash not artifact......seems to be a faulty GPU....ur better off getting a replacement....since the kernel mode error pops even at low temps then the possible suspect is the GPU....got mine replace for the exact reason....If u happen to have a second system u could try the card in that just to be sure....

 

[richard_ab]

ye i do have a second system with very similar specs 6 gb ddr3 ram asus ramage extreme 2 motherboard win 7 64 bit i7 cpu 2.66ghz and same 780w PSU.

I ordered cod 6 which is the main game im going to be playing from now on, if i have any problems im taking it back to the shop.

Thanks