Is my GPU dying?

[SMOKEU]

My shader unlock 6950 was fine mining at 930MHz @ 1.200V all day long. I had it at that speed for a couple of weeks with no issues, but it randomly started crashing again and again. It won't even run stable at 900MHz @ 1.200V without giving a BSOD after 5-10 minutes of mining. I haven't changed any settings which would make it do this. It's still fine for gaming at 920MHz @ 1.200V. Temperatures are always below 80°C. I just get the feeling that the GPU is on its way out and I can't understand why it was perfectly stable but now it's not.

If I flick the BIOS switch back to the original BIOS it's even more unstable. It has to be over volted on the stock BIOS just to mine at stock clocks without crashing.

[goxed]

My shader unlock 6950 was fine mining at 930MHz @ 1.200V all day long. I had it at that speed for a couple of weeks with no issues, but it randomly started crashing again and again. It won't even run stable at 900MHz @ 1.200V without giving a BSOD after 5-10 minutes of mining. I haven't changed any settings which would make it do this. It's still fine for gaming at 920MHz @ 1.200V. Temperatures are always below 80°C. I just get the feeling that the GPU is on its way out and I can't understand why it was perfectly stable but now it's not.

If I flick the BIOS switch back to the original BIOS it's even more unstable. It has to be over volted on the stock BIOS just to mine at stock clocks without crashing.

I would first try a different computer, and then check if the fan is spinning good and heatsink fins are not gunked out.

[SMOKEU]

Now it's crashing while gaming and it was fine at the same clocks a few days ago. I removed the heatsink and cleaned everything out a few weeks ago and replaced the thermal paste. I'll see if I can find another computer to put the card into to test it.

[crazyates]

When you repasted it, did you make sure the VRM pads were in good condition and make good contact with the cooler? What are the VRM temps? (you can use GPU-Z to find this).

[SMOKEU]

When you repasted it, did you make sure the VRM pads were in good condition and make good contact with the cooler? What are the VRM temps? (you can use GPU-Z to find this).

The thermal pads seemed to be in a good condition. GPU-Z says the VReg temperature is around 44°C after it's been mining for a while at stock clocks and volts, and the temperatures don't go over that. It seems a bit low as I'd expect it to be a lot warmer than that.

[crazyates]

When you repasted it, did you make sure the VRM pads were in good condition and make good contact with the cooler? What are the VRM temps? (you can use GPU-Z to find this).

The thermal pads seemed to be in a good condition. GPU-Z says the VReg temperature is around 44°C after it's been mining for a while at stock clocks and volts, and the temperatures don't go over that. It seems a bit low as I'd expect it to be a lot warmer than that.

Ya that is WAAY too low to be right, unless you're watercooled. Normal should be 80-100C.

[SMOKEU]

GPU-Z has always said the Vreg temperature is something ridiculously low. Even when over volted to 1.200V with a 930MHz overclock while mining it's never gone above 60°C even after a few hours non stop mining. This has also been the case with previous Windows 7 installs with the same card.