Test GPU overclock stability?

Hey guys Dell tech is coming in the next two hours to replace my faulty GPU and once I find that the new one is stable at factory settings I intend on overclocking it to get the most out of Homefront, COD, BC2, etc. Ive overclocked successfully on my old R1 but I need a good program to test OC stability. Dell used Video Card Stability Test for this but then again its Dell and they still recommend me use 258.xx drivers. Any ideas for a GPU stability test?

-DeeVu

 

You could use OCCT with it's built in artifact scanner but I find most if not all stability tests less stressful then on R1 so run it longer to compensate- OCCT used to load my R1 to 98%, now it's ~50% with my R2.

You could also use 3DMark, loop the Crysis benchmark tool and Far Cry 2 benchmark tools.

 

Cheers. I will do that! Just waiting on my new MoBo. Unfortunately in order for my current GPU i have to underclock it to keep my R2 from crashing. So I will definitely look into these suggestions!

 

Just looked at OCCT and the words 'It will make your CPU hotter than never' and 'that will make your GPU really hot' do tend to scare one a bit.

Just had my Desktop PC open, stripped down (CPU cooling hood removed) and running while I decided how to 'engineer' a STAT3(6GB) 1TB HDD into a circa 2004 PC.

Then wham!!!! Dead silence and nothing. Like really nothing!!! Scary nothing.

Powered the PC back on and the boot loader told me I had a temperature 'issue'. If I attempted a boot the power tripped off again.

Hours later (with cooling hood on) the PC(now cooler) is fine and boots and runs with no problem.

The M11x C2D CPU has a max. temp of 105C and the Nvidia documents the same generic temperature for the GT 335M with temperature throttling at lower levels. (hopefully).

So running applications like Furmark and OCCT we're attempting to 'bake' the chips and with added load/heat attempt to generate memory errors or throttling. That is if the chip is less than perfect and/or no errors if the chip is good.

If the CPU/GPU is less than perfect to start with, hammering them in this way is not going to help them. CPU/GPU in the realm of normal or better may not be advisedly impacted as it's able to handle the extra load/heat.

@DavyGT, you mentioned that initially you could load the M11xR1 up to 98%, but now(undefined time and activity) you're only able to load the system up to 50%.

Does that imply that the system is now not operating at the same level it was before when you were able to get 98% load?

Or are we saying the OCCT application has now 'de-stressed' it's application to limit potential damage?

What do you think prompted almost 50% drop in load?

Or am I just not reading what you've written incorrectly.

Just thought I'd start the ball rolling with a good olde noob question.

 

There is MSI Kombuster as well. Kombuster doesn't catch some of my unstable overclocks though, even at full settings running for a couple of hours, while Oblivion will crash within 5 minutes.

 

That's the purpose of the program. Still stays within safe limits anyway.

Why did you leave your PC running while you were upgrading?

Oops, I meant to add, with my R2.

It's just OCCT, Furmark and ATiTool loading lower then usual. Games still run fine.

I don't know why my R2 is having lower load. Maybe it's because the PCIE bus is getting saturated with Optimus traffic or nVidia drivers are throttling the programs. They're the same versions as on my R1.

Kombustor was hopeless when I used it. Ran it overnight and passed but my R1 crashed in 10 minutes of Crysis.