System hanging, GPU to blame?

[Molehammer]

I recently ordered almost everything for my new build from scan. It's not been great. I sent the first of my graphics cards (Nvidia GTX 275 XFX XXX edition overclock) back , and got a replacement for the faulty one.

Now the new GTX275 (same model) seems to be overheating. If I'm running something intensive (i.e. COD MW2 high specs) my system hangs and the graphics card reads around 83C just before.

I7 920 CPU, stock cooler
Gigabyte GA-EX58-UD5 Mobo
Corsair TX750 750W PSU
XFX GeForce 275 GTX XXX Core Edition GPU
6GB Corsair Dominator RAM CAS 7-7-7-24
Thermaltake Element G fan
1TB Samsung HDD


Please help, is it my motherboard killing my GPU?

Any more info needed, please say

[Phage]

What sort of case ? What air flow ?

[Molehammer]

Thermaltake Element G case, sorry, with a 230mm fan on the side, a 200mm on the front and top, and a 140mm on the back...

[rabbid]

also try cleaning drivers & installing latest

what duty cycle does your gpu fan run at load? just check its hsf is secure too.
just to clarify its ok in windows or can hang there too?
does it just hang in this game or in others too? if so all or just some?

try running benchmark/stress sw and see if you can replicate the failure whilst running gpuz or rivatuner on record. then you can inspect the logs to see what conditions were when it hung.

[Molehammer]

OK I checked all of that.

It's fine on 2D applications, but in any 3d application it seems to crash. On Gmod i ran it at full settings (1920x1080, all settings highest). It ran at around 52/53C then the screen was garbled and my display driver was reported as having crashed.

Speedfan 4.40 reports all my voltages as way off what they should be, could that be the problem?

I rang Scan technical support, too, they reccomended I set the clocks to the stock card, not the overclock to see if that helped. It didn't, so its not an unstable clock.

[Wizbit]

I have a brand new build as well only just two months old and I have the same GPU card and while I was in windows 2D mode it was fine but when I played games 3D mode after a random time the computer would BOSD and the temp max would be no more than about 75c max (which is not that high) I also noticed in GPU-Z that it would clock its self back to 400 core speed and if I looked in the windows event viewer at the same time this card underclocked & BSOD it would say the driver nvlddmkm.sys had stopped responding along with and the mini dump BSOD logs.

I tried two new nividia drivers and cleaned them out between with a drive sweeper in safe mode and I also tried under clocking the card but it still crashed then a week after this started happening the card did its last BSOD and never worked again in either my PC or my friends PC.

And now the card has been RMA back to scan who have not been that helpful (will not go into that fully at the moment) and saying it will be at least 20 days before I can have a replacement as even though they have said the card it faulty it will have to go back to XFX to be checked out .... So I guess I will not be using my new computer for ages !

I hope there is not a bad batch of these cards!!!! And I hope your card does not go like mine. I would be well miffed if my replacement card (when & if it comes) after 20 days or more did the same thing and I had to go through all the RMA hassle again!!!

[rabbid]

wizbit, thats not good.
Have you an old/ spare card at least to carry on with?

[Wizbit]

wizbit, thats not good.
Have you an old/ spare card at least to carry on with?

No spare card! I only have AGP cards. So the computer just sits there lifeless with the rest of its warrantys for all the rest of the parts drifting away. And as yet I am still waiting for an answer from XFX...... not very happy

[littlexanth]

A friend of mine says I should sell old equipment when I upgrade, but this is the exact scenario I always want to avoid. Sometimes the old spare kit comes in very handy for trouble shooting new builds.

Got any friends that can help out with a card to test with?

Littlexanth