Can't boot after stress test - GPU fried?

paganompu

Distinguished
Jan 28, 2012
9
0
18,510
Hi all,

My PC shut down after running Furmark once the GPU hit 72 deg. And it won't boot: fans start spinning for a split second and stop - unless I pull the graphics card or disconnect its first power adapter. Basically the same scenario as posted here

Build is:
i5 2500k no OC
MSI 460 GTX oc'd to 900 I think with upped voltage
8GB corsair ram
Corsair mx650 psu

I was fairly surprised at all this, because I used to use the same OC profile for the card a lot. Though that was a year or two ago, I've lately played pretty much only indie and low gfx games. I have had the card for about three and a half years now, could the age have made it more susceptible to damage from overlooking?

So do you guys figure it's fried? There was no burning smell. I can live with the loss of the card, I just hope the motherboard and psu are all good...

 
Your PC shutdown after hitting 72 degrees? In Celsius I'm guessing?

Graphics card should not have fried due to that temperature but it could be because of the overclock itself.

The Power Supply should be good enough to supply the graphics card.

Anyway, the only true way to find out is actually getting another graphics card if you can borrow or find one and test it with your computer. Make sure the graphics card has a PCIe external power port so you can connect that to the power supply. From there, you can see if the card works or not. If you do not get that working, then it may be the power supply.

Over time, power supplies degrade. Even CPUs and GPUs degrade over time, and they degrade faster when they are overclocked meaning that it may be possible to push them to a high limit for a overclock but now that max overclocking limit for that component has decreased a bit.
 
I didn't have a second graphics card to borrow, so I went ahead and tried it on my girlfriend's computer (A6-3500, Corsair CX430 PSU). Behavior was exactly the same as described in my first post. That left me pretty confident that the graphics card was at fault. I went ahead and ordered a 2GB XFX Radeon R9 270 as a replacement.

Here's where things get a little hairy.

I plugged in the XFX card and powered on the system. No signal, no POST beep - but the fans on both the card and the CPU spin as normal, unlike with my old 460. Now I've figured that my system is well and truly screwed. However the exact same thing happened on the other PC as well. I did take note that that system only had a 430W PSU instead of the minimum recommended 450W, but surely it would at least go through POST?

The way I figured, the XFX must be a lemon because it surely would have worked on the second PC even if mine was broken. I RMA'd it and just received the reply today: "Displays fine, currently running on Heaven Benchmark @ 1080p. Heaven ran fine for 20 mins, tested all outputs on the card and they all display."

I genuinely don't know what to think or suspect now. Anyone have any ideas?

---------------

For reference, this is the detailed specs of both computers:

My PC:
Intel i5-2500K no OC
MSI P67A-GD53
1GB MSI GTX 460 Cyclone
650W Corsair Enthusiast 650TXV2
8GB (2x4GB) Corsair Vengeance DDR3
120GB Corsair Force Series 3
1TB Western Digital WD10EALX

2nd test PC:
AMD A6 3500
MSI A75MA-P35
8GB (2x4GB) Corsair DDR3
Kingston SSDNow V300 120GB
1TB Seagate ST1000DM003 Barracuda

---------------

Some other points I should probably mention:
- the 2nd PC was set to PEG in BIOS (important since it uses APU for day-to-day video output)
- I tried both PCI-E slots when testing both cards
- I tried all outputs on each card (except DP since I don't have the cable for that)