• We’re currently investigating an issue related to the forum theme and styling that is impacting page layout and visual formatting. The problem has been identified, and we are actively working on a resolution. There is no impact to user data or functionality, this is strictly a front-end display issue. We’ll post an update once the fix has been deployed. Thanks for your patience while we get this sorted.

What problem to use for stress testing an overclocked Video card?

aatf510

Golden Member
ATItool artifact scan doens't seems to be sufficient, and 3Dmark06 loop seems to crashes even at stock speed?
What else is there to use if I want to make sure a overclocked Video card is stable?
 
3DMark06 is actually an excellent stabilty test, IMO. Games that push the card a lot are also good tests, Oblivion and FEAR come to mind.
 
Any sufficiently demanding 3D game will do. Especially Oblivion - if it doesnt crash in 20 minutes of Oblivion, chances are it will not crash after hours of other games.
 
Originally posted by: munky
Any sufficiently demanding 3D game will do. Especially Oblivion - if it doesnt crash in 20 minutes of Oblivion, chances are it will not crash after hours of other games.


QFT. But you can always use rthdribl...
 
If 3dmark 06 keeps crashing it is unstable. Mine wouldn't run 06, until I got better cooling, IMO.
 
Originally posted by: d3lt4
If 3dmark 06 keeps crashing it is unstable. Mine wouldn't run 06, until I got better cooling, IMO.

But my 3dmark06 always crashes after the same test even at stock speed, and my friend has the exact same problem crashing at the exact same point. Therefore, I think it's a bug rather than my card being not stable.
 
I've run 3dmark overnight and not crashed. If it's crashing for you, you either have a bad card or you have insufficient cooling in your computer case. The two test I currently use are RTHDRIBL and 3DMark.
 
RTHDRIBL

And the forest benchmark in 3dmark05. The one with the fairy flying around, I find it really easy to spot artifacting in that one. Also, Doom 3 and Quake 4 are pretty sensitive.
 
Originally posted by: akugami
I've run 3dmark overnight and not crashed. If it's crashing for you, you either have a bad card or you have insufficient cooling in your computer case. The two test I currently use are RTHDRIBL and 3DMark.

Right. OP, just because your GPU is at stock speeds doesn't meen it cannot be bad/unstable, especially if you have a factory OC'd GPU. When nVidia designs a GPU they design it to operate at rated clockspeed in something probably like 99.99% of cores built. They know that at some point it's better to introduce some bad parts than to advertise a slower part. Then when eVGA or BFG decided to market their card factory overclocked they decide they will accept say a 99% reliability rate. There is no guarantee any user won't be in one of the small percentages with a bad card. If you cannot get your card to run stable, and have no OC issues, good drivers, and your case has OK cooling then you may have a bad card. And don't let anyone tell you to get better case cooling if you don't want to. These cards are not designed to be run in a abnormally hot case, but the card as sold should run fine under normal conditions. RMA if you cannot get it to run 3d mark at stock speeds, that's what RMA programs are there for.
 
I forgot to mention that bad drivers can also cause instability. Make sure you have the latest BIOS, motherboard drivers as well as video card drivers. Make sure you do proper removal of the older video card drivers. Both ATI and nVidia suck and leave renmants of older drivers behind that can cause instability. Download Driver Cleaner Pro (free software) and it will help you remove traces of old drivers. Generally what people recommend, and what I also recommend, is to uninstall the older drivers, reboot into safemode, run Driver Cleaner Pro and have it clean out your old video card drivers. Reboot and then install the new drivers.
 
I am making the assumption because both my friend's watercooled XFX 7900GT and my stock cooling eVGA 7900GT crashes at the exact same point (right after HDR2 is finished) if we choose to loop 3Dmark06 HDR1 and HDR2, and it only crashes at the same spot in serveral trails and it only crashes after running for over an hour. However, if we let our cards run overnight with the 3dmark06 Demo, then our machines are fine.
 
Well I don't want to tell you for sure there is anything wrong, but 3Dmark is really popular, and wouldn't any problem that caused it to crash be widely known?
 
could be a bad download of 3Dmark. I've had it crashing on one system too, ran everything else fine. but as many has said use RTHDRIBL, it was the one that I got the highest temps in my setup.
 
Back
Top