GTX 580 DirectCu II high load temps

C

Camberwell

Gawd
Joined
Jan 20, 2008
Messages
947
I got my new card a couple of days ago, and I notice that when I first set the fan to 100%, and then stress test with OCCT, temps get up to 96/97C. Isn't this pretty high for this cooler with the fan on full?

The airflow in my case is pretty good. I have 2 x 120mm Scythe's at the front, and another at the rear. I have removed the metal fan grills and the dust filters, and when I game/bench I keep the front door of the case open, so I don't think this is the problem.

Do I need to RMA?
 
sounds like you have good airflow in your case, yah that seems way too high for that big ass non reference cooler! how are your gaming temps? how are the temps with the cards fans set on auto? what are you using to monitor temps? are you sure that your fans speeds are increasing when you set it to 100%? i'm sure you'd hear them spin up. i'd call asus and set up an rma, that card is a ticking time bomb of failure imo.
 
HotShotMedic said:
sounds like you have good airflow in your case, yah that seems way too high for that big ass non reference cooler! how are your gaming temps? how are the temps with the cards fans set on auto? what are you using to monitor temps? are you sure that your fans speeds are increasing when you set it to 100%? i'm sure you'd hear them spin up. i'd call asus and set up an rma, that card is a ticking time bomb of failure imo.
Click to expand...

Funny thing is, gaming temps are 70C with fan at 30% (my own fan profile), so pretty good considering the oc @ 900MHz. I use Afterburner to monitor, and when the fans are at 100%, trust me you hear it!
 
strange man, 70C at 30% sounds great. not sure why its getting so hot w/ fans at 100%, even w/ stress testing.
 
Did you happen to upgrade your driver to 270.16 (I believe that is the latest version). When I upgraded my Galaxy 580 to this driver it immediately started over heating. I ended up reverting the driver to the old one and the problem went away. I have opened up an Nvidia Forum post on the subject.

When i say overheat I am saying I believe my cards health was in jeapardy. I think if I had left that driver in a few minutes more the card would have just flat out burned up.
 
So what happens when you game normally and put the fans at 100%? Either the fans are defective, and not working properly at high speeds, or OCCT is designed to set Nvidia cards on fire.
 
Maybe your heatsink isn't making good contact as that temp is very high for the card.

My GTX 580 DirectCU II never gets above 58-63c (idle 34-37c) when running stock clocks / fan profile. If I go 900 on the core and set the fan to 40% it might go as high as 67-69c but has never broken 70c. This has been the case since early Feb -- and is still true today using the latest 270.61 WHQL drivers.
 
I am using 270.61. This only happens when I pre-set the fans to 100% and then stress test my oc of 900MHz @ 1.15 V.

During gaming I also don't get above 69/70C with fans at 40% with this oc, so it seems to only be the stress test which is the issue. If I dial down the voltage from 1.15 V to 1.10V, temps get much better during stress testing, although I have to drop the core to 875MHz for stability. If I game normally and set the fans at 100% then temps drop significantly. I didn't think OCCT was as bad as say Furmark?
 
If the temps only get past 70c when using stress testing software than I see no reason for issue. So long as it performs in games and you don't witness instability or heat related throttling -- the card is more than likely working as intended.

I've never used furmark or OCCT for more than a minute or two...what is the point? Stress test by running games or maybe a round of 3dmark/heaven/etc.
 
WorldExclusive said:
Temps are high, either remount the heatsink or RMA. My 570 DCUII doesn't past 60C @ 45%.
Click to expand...

As I mentioned, during normal use with fans at 40% it does not go over 70C with an oc of 900MHz @ 1.15V, it is only when I stress test with this oc that the temps get that high....
 
way too hot for a card with an aftermarket cooler like that, I'd start monitoring your fan % in afterburner and see if it's working properly. I know you said the fan gets loud at 100% but the bios on the card kicks that in no matter what. It would suck if you fan is forced too low by accident in afterburner and it's simply not apparent now. I think the heatsink may not be making proper contact on the card as posted earlier if everything checks out.

In that case call the manufacturer and ask what you can do, if they say it's okay to check and re apply tim then I'd go for it.
 
Lord_Exodia said:
way too hot for a card with an aftermarket cooler like that, I'd start monitoring your fan % in afterburner and see if it's working properly. I know you said the fan gets loud at 100% but the bios on the card kicks that in no matter what. It would suck if you fan is forced too low by accident in afterburner and it's simply not apparent now. I think the heatsink may not be making proper contact on the card as posted earlier if everything checks out.

In that case call the manufacturer and ask what you can do, if they say it's okay to check and re apply tim then I'd go for it.
Click to expand...

According to every piece of monitoring software that I can find, the fans are running at 100% (3450rpm) when I set them manually (and also automatically in the bios if left to itself). Thing is, I'm pretty sure my RMA will be denied if I try to explain that I get these high temps when stress testing an oc at max voltage, because for the rest the temps are fine.....
 
Camberwell said:
According to every piece of monitoring software that I can find, the fans are running at 100% (3450rpm) when I set them manually (and also automatically in the bios if left to itself). Thing is, I'm pretty sure my RMA will be denied if I try to explain that I get these high temps when stress testing an oc at max voltage, because for the rest the temps are fine.....
Click to expand...

Well I dunno but in my opinion being that you bought a card with a robust cooler because you've read reviews and seen advertised on the box that it should run xyz cooler than the reference design, that since it's not efficiently cooling the card up to standards then it is a issue with the card not performing up to spec. More than likely a bad install of the cooler. Either they should give you the thumbs up to fix it yourself (if your willing) or they should rma the card.

You could easily make up a good argument here. I don't think that card should go over 79-85 max in any stress program and I'm being generous. It probably should be about 5-10 degrees lower than that.
 
Lord_Exodia said:
Well I dunno but in my opinion being that you bought a card with a robust cooler because you've read reviews and seen advertised on the box that it should run xyz cooler than the reference design, that since it's not efficiently cooling the card up to standards then it is a issue with the card not performing up to spec. More than likely a bad install of the cooler. Either they should give you the thumbs up to fix it yourself (if your willing) or they should rma the card.

You could easily make up a good argument here. I don't think that card should go over 79-85 max in any stress program and I'm being generous. It probably should be about 5-10 degrees lower than that.
Click to expand...

I think you're right, I'll apply for an RMA on Mon and see what happens. As you say, I bought the card for the cooler, and I've never had a card get that hot! I'll let you know how I get on, thanks for all the contributions guys!
 
Damn that's weird. My 570 directCU is oc'd to 930 core and overvolted and I still don't pass 70c with 50% fan speed @ 99% gpu load.
 
You must log in or register to reply here.
Back
Top