Did overheating damage my GPU?

[theriju]

So yesterday I spent the day hanging around the house playing Battlefield 3 with my VisionTek Radeon 4870 (1GB) when after about 3 hours or so I started to see flickers and tearing on the screen. My gaming room at the time was around 82 degrees (hot as usual) but its really nothing new for my video card as it always operates find around this weather. I checked my Catalyst Control Center and it said card was at its usual 71 degrees.

The problem was that I realized I had been playing with the AMD overdrive utility (just testing custom clocks and such) a few days ago and I stupidly left my core clock set to 700 Mhz instead of the stock 500 and memory to 1150 Mhz instead of 900. I had pretty much forgotten to reset it. Turns out I was running the card at those clock settings with the stock cooler (I think) I have. I instantly shut off the game and reset all my clocks back to stock until my GPU cooled down.

Within about 2 hours, just to test, I started up Mass Effect 3 to see if my GPU would hold up and I had no problems for the two hours or so that I played. No tearing or flickering of any kind. It seemed as if the problem was temporary.

My question is, did I cause any permanent damage to my GPU? Although it doesn't seem like it, I keep thinking that some internal damage had occurred that drastically shortened its lifespan. I know it's about time for me to upgrade, but I was planning to give my 4870 to a friend of mine. I just don't want him to have any problems with it after I hand it over.

Thanks for taking the time to read through this. Any feedback would be appreciated.


My complete specs:

Intel Q6600 (2.4) Ghz
Ati Radeon 4870 (1GB) VisionTek
3GB Ram 667 Mhz DDR2
550W Psu

 

[coozie7]

I'm in chilly, currently windy England, you live on the Sun?
It takes a few seconds to go from a game to the Overdrive window and if the card cooler is spinning fast it can drop 8 or 10 C in that time.
But, even allowing for that, the card was still well within its safe temperature range so it is fine.
FYI all current cards and CPUs' 'throttle'-that is to say they automatically slow their clocks-if they get too hot, I think the HD4870 throttles at 105C and automatically shuts off at 115C.
If you want to monitor the temperature while playing, try GPUZ.

 

[refillable]

It may have reduced life span of the GPU, but it shouldn't make internal damage.

 

[Rockdpm]

The card should be fine. Letting a card run beyond its capabilities or past your average overclock can make something unsafe or unstable but.... The card should be fine. As long as you plan to upgrade soon. the 4870 is still a good card but its slowly becoming historic....eventually its not gonna run the latest games on medium

 

[Murissokah]

Your card is fine. The short-term consequences of overclocking are basically stability loss and a moderate rise in temperature. They both go away when you set the card back to sock frequencies.

As for long term, what makes your card die is a phenomenon called electromigration. This phenomenon consists of the elecrons pushing the ions on the IC, basically shoving them away from their original posistion and thus destroying the logical architecture.

Electromigration happens with all electronic circuits, but the effect is increased with higher currents. Higher frequencies will demand more power, and as we all know... with great power comes great current squared times impedance.

Don't freak out though, while this phenomenon is extremely important for a space mission, it is usually too slow to matter for consumer electronics. In the end, your card is very likely to outlive its usefulness.

 

[theriju]

Thanks a lot for your replies guys. Really helped a lot.

 

[theriju]

Best answer selected by theriju.