Have I bricked my card?

Today i swapped the stock thermal paste to IC diamond and the results were very good on the cpu and second gpu, however the primary gpu showed no improvement. So i decided to reapply it again with a thinner layer because i assumed i over applied on that component and i booted again and ran furmark to test temps. The temps on the gpu rose to 143C!!!
So i shutdown straight away and reapplied the paste again with a thicker layer and now my machine will not POST.
I've tried reseating the gpu and a power drain to no avail.
Ideas?

 

I don't think GPU's can get to 143C. I think it would be smoking at 120C.

 

There was no burning smell or anything and cosmetically it is fine... dunno :S

 

How did you get your temperature reading? 143C is unlikely.

 

with gpuz and hw32...
Still wont post. It just comes up with 2 blinking lights (numlock and capslock w/e, the 2 on the left) which when i looked it up means replace the gpu

 

I'd say if 143*C is accurate you probably fried it.

 

The paste was put on by a technician that came to replace my screen and i said while he was at it could he swap the paste.
So after it stopped working i rang dell and told them that he swapped the paste (which he did originally) but that it was with the dell TIM (knowing full well they would wash their hands of it otherwise).
Now they say they need to corroborate the story with the engineer who will no doubt tell them it was ic diamond.

I shouldn't have told them about the paste at all. I bet they wont swap it now. Would this be covered under accidental damage?

 

as long as they don't know you did it you'r self it's fine

 

Thing is i dont want to get the tech guy into trouble lol cos he was one of the few who knew what he was doing.

 

those who don't know don,t get in trouble for much worst so if thatt one get in trouble for knowing ...... you did him a favor in hinting him that he should find a new job

EDIT on the other hand you are the first one that i heard to suceede in killing an ati in a alienware by overheating they usualy shut them selfs down befor they reaches danger levels (mines shut down at 107 while the rest of the system keep runing so i can click listen to music even myskype keep runing but the graphic goes bye bye) so for all we know the cards might have been defectuous from the begining

 

Well that's something i suppose lol. I've set an unfortunate record it seems.

I just hope to god this works out well and i get a replacement card without paying for it

 

Just as a note, IC Diamond is not a paste that should be spread, really. They recommend on their own site that you put a blob in the middle of the component, and let the heatsink pressure spread it.

I still can't get my head around the fact that you managed to push it to 143C. That's absolutely insane! It should have shut off as soon as it was reaching above 110C!

Unless you did a really poor paste job, I doubt it could have risen that high that fast. How quickly did it shoot up, literally within seconds? It might have been a defective part to start with.

I wonder...technically, you could see if you can get it replaced under accidental damage. I'm not sure if it'd cover that though.

 

I don't think you should have mentioned the paste and gotten the tech involved in all this. You could have just said the GPU died and that would have been the end of it. Anyway, if they're going to corroborate with the tech then you'll still likely get a replacement though who knows how they'll deal with the tech.

Besides, temperature gains from applying TIM on the GPU is negligible at best, I wouldn't recommend it. The 143C does seem unlikely and the only way it could possibly have happened is if you somehow forgot to screw down the heatsink and the GPU had a broken sensor.

 

Well, I have a different point of view. In fact, I have 2 master HD 5870 which one I bought it from Dell just in case the first one died. So, what happened was, I tried overclocking both cards to see its maximum stable clock with one is around 850/1000 and another close to 900/1100 but the one with the 900 clock has a broken heatsink with the memory heatsink slightly off the copper pipe ( what I suspected ) as furmark can quickly ( within 30 seconds ) increase the temperature of the memory to 118 without shutting down or throttling. Upon swapping the heatsink, the problem solved, memory temperature maxed out at 92 during furmark test with ambient of 30'c.

 

could ati have removed that protection from hier cards or mabe dell's 5870 bios did it?

i've never seen an ati get that hot withou shutting it self down befor these 5870 or defective ones

 

Well I can confirm that the heatsink was ON and that also that it only took about 5-10 seconds to reach that temperature. I agree it was like there was no heatinsk on temp wise - maybe i spread the paste too much away from the middle of the chip?? In any case the cards should have shut off straight away to protect themselves surely?
I know from experience that most other cards will have a shutoff limit so is it likely that my mistake with the paste uncovered a defective card anyway?

It's ironic that i wanted to keep my primary gpu temps from reaching the 90's and it ended up totally fried.

 

No, it wont. Accidental Damage/Complete Care is to cover spills and falls/drops - that is it. Contact support for clarification on terms or check the support site and view the warranty terms link.

Any further discussion on falsifying info needs to stop. We have a strict policy here on NBR against this type of discussion - see forum rules relating to fraud/illegal activity.

Closing thread.