Is 80-95 degrees normal for a HD5870 PCS+?

[Blacktric]

The new Powercolor HD5870 PCS+ I've bought gets as hot as 80-95 degrees according to GPU-Z (to be exact: GPU #1: 80-83, GPU #2: 90-95 and GPU #3: 80-85). Is this normal? And is it healthy to leave the card for 5-6 hours under %100 load with those temps? I'm using a Coolermaster HAF-932 case with 3 25cm fans and a 92mm fan. According to HW monitor, the system temp is around 45 degrees in idle and ambient temperature is around 30-33.

 

[(sic)Klown12]

A little high for that cooler, but the chip can handle it. The stock cooler for 6970s would allow the chip to hit 85-87c(GPU #1 sensor, the one that AMD uses for CCC) and they can run like that without issue.

If you want to bring temps down, use Afterburner to create a custom fan curve.

 

[slayernine]

That seems a bit high for a 5870. I have an XFX 5870 and it is currently idling 50 degrees. After doing a bit of light gaming it jumps up to just over 60 degrees.

With a stock cooler my temps were about 10 degrees higher so 60-75 roughly.

 

[YoungGun21]

Turn up the fan to about 70% and watch your temps drop 20C

 

[OS]

my 6950s run that hot at default stock also

it's kind of BS but i assume since the fan control has temp feedback its probably ok

 

[mdlam]

those temps are pretty normal for the crappy fan they have on those cards. Get like a 30 dollar cheapie cooler on it and you'll drop temps down to about 70 under the worst load and have it run silent. I did a setsugun 2 on a 6950 ref 2gb unlocked, overclocked it to 1k, and it rarely passed 80, and even when it did, was pretty much silent....to me as long as it worked without temperature-induced artifacts and was silent, I was happy.

 

[dma0991]

80C is pretty standard for a GPU but take precautions for anything above 90C. Although I've seen GPUs running >100C before, I don't think it would be good for the card in the long run.

If you're running a triple Crossfire and without a single slot spacing you might be choking your cards for cold air. The small gap between each cards might be a potential problem.

 

[3DVagabond]

The new Powercolor HD5870 PCS+ I've bought gets as hot as 80-95 degrees according to GPU-Z (to be exact: GPU #1: 80-83, GPU #2: 90-95 and GPU #3: 80-85). Is this normal? And is it healthy to leave the card for 5-6 hours under %100 load with those temps? I'm using a Coolermaster HAF-932 case with 3 25cm fans and a 92mm fan. According to HW monitor, the system temp is around 45 degrees in idle and ambient temperature is around 30-33.

I'm not 100% certain, but I remember reading that GPU #1 is the GPU temp, GPU #2 is the VRM, and GPU #3 is the VRAM. If someone else knows better pipe up on whether this is correct or not?

 

[GaiaHunter]

http://www.techpowerup.com/reviews/Powercolor/HD_5870_PCS_Plus/32.html

 

[BFG10K]

those temps are pretty normal for the crappy fan they have on those cards.

Uh, the cooler is better than the reference version: http://www.hexus.net/content/item.php?item=22037&page=5

Those temps are far too high, unless the OP is running Furmark burn mode or something.

 

[(sic)Klown12]

I'm not 100% certain, but I remember reading that GPU #1 is the GPU temp, GPU #2 is the VRM, and GPU #3 is the VRAM. If someone else knows better pipe up on whether this is correct or not?

That's what everyone thought, but W1zzard, the creator of GPU-Z, has admitted that he doesn't know exactly where each sensor is in the core. On my 6970, when you hold the mouse over one of the graphs it pops up a message stating it.

 

[tweakboy]

That is not normal, are these load temps you posted ? or idle ?

The card will function and live even @ 100c for some, but of course we dont want it there it will blowup soon. YOu need to get those temps to 60's and 70's raise fan to 100 percent and see what happens. gl