Cooling On 6970

Associate
Joined
10 May 2011
Posts
24
Hi, I have a Sapphire 6970 which has always troubled me with temps. It doesn't get overly hot but I think it's hotter than normal. At the moment I have it on 50% fan speed idle between 38c and 45c. It's caused my PC to reboot a few times during BF3 gameplay and I noticed that it has melted the plastic around the heatsink. I currently have it in a HAF X with all fans on intake but H70 fans as exhaust. I also have a cheap Cyclone fan in my bottom drive bays blowing cold air onto the card but it seems to have made no difference at all. I know it's not right for the GPU to get hot enough to cause a reboot but are those idle temps about right? Thanks in advance.
 
If it's a leaf blower style cooler best to leave it on auto or better yet set up a gradual fan profile probably as they need to spin higher than 50% to cool well most of the time.It's expensive only 4 I would consider are the

arctic cooling accelero twin turbo II
Arctic cooling accelero extreme plus II/III

Zalman VF3000A < you will need to buy VRM heatsinks as I don't think the default ones fit
Zalman sell a 6970 vrm heatsink but good luck finding it for sale

Thermalright shaman < again may have to source additional vrm heatsinks not sure what actually comes with it now

Barring that I would maybe take the stock cooler off carefully and reapply some fresh thermal paste see if that helps.

I am waiting on a 6970 to arrive in the post and had planned to put an accelero twin turbo II on it myself as I think especially cable tying your own 120mm fans would do wonders for it, everyone will recommend the extreme plus II/III but they are like £50 that is silly money just for a heatsink.

Oh there is a Gelid cooler that will fit too but I think again the vrm heatsinks lacking.
Buying a great cooler for a card used to be far cheaper all this £50 stuff is crazy.

So I take it you have tried using the gpu duct on the haf x then ?
 
Last edited:
If it's a leaf blower style cooler best to leave it on auto or better yet set up a gradual fan profile probably as they need to spin higher than 50% to cool well most of the time.It's expensive only 4 I would consider are the

arctic cooling accelero twin turbo II
Arctic cooling accelero extreme plus II/III

Zalman VF3000A < you will need to buy VRM heatsinks as I don't think the default ones fit
Zalman sell a 6970 vrm heatsink but good luck finding it for sale

Thermalright shaman < again may have to source additional vrm heatsinks not sure what actually comes with it now

Barring that I would maybe take the stock cooler off carefully and reapply some fresh thermal paste see if that helps.

I am waiting on a 6970 to arrive in the post and had planned to put an accelero twin turbo II on it myself as I think especially cable tying your own 120mm fans would do wonders for it, everyone will recommend the extreme plus II/III but they are like £50 that is silly money just for a heatsink.

Oh there is a Gelid cooler that will fit too but I think again the vrm heatsinks lacking.
Buying a great cooler for a card used to be far cheaper all this £50 stuff is crazy.

So I take it you have tried using the gpu duct on the haf x then ?

Yeah I'm using the duct atm with the crappy little fan at the end of it. I have looked into aftermarket coolers but like you say £50+ is crazy considering the amount I paid for the card in the first place. As for the paste option, I have already tried replacing that with arctic paste but I had the same temps. I've just been on BF3 for 45mins and checking MSI afterburner overlay I was between 98-99% usage and at 70-73% which isn't as bad as usual.
 
Maybe having top fans as exhaust will help case temps , i ran my 6950 passively @ stock using Arctic cooling accelero extreme plus II in my Haf X.
 
Maybe having top fans as exhaust will help case temps , i ran my 6950 passively @ stock using Arctic cooling accelero extreme plus II in my Haf X.

The only reason I have top fans as intake is because if they are exhaust they fight against the push/pull on my H70
 
..:: Update ::..

I've removed the crappy little fan and screwed a 120mm to the end of the air duct as an intake onto the GPU but that seems to make no difference....
Any suggestions?
 
you have to run the gpu cooler at higher than 50% those leaf blower coolers are noisy there's really nothing you can do except put up with the noise if you want lower temps

use hw monitor to see what temperature the card is hitting during gaming I would say if it's 80c or below then the crash / reboot is maybe being caused by something else
 
you have to run the gpu cooler at higher than 50% those leaf blower coolers are noisy there's really nothing you can do except put up with the noise if you want lower temps

use hw monitor to see what temperature the card is hitting during gaming I would say if it's 80c or below then the crash / reboot is maybe being caused by something else

I used MSI AB to monitor temps last night during BF3 gameplay after I fit a 120mm fan at the end of my air duct and it got to about 71 degrees. The only reason I'm assuming it's the GPU causing restart is becuase the GPU plastic housing has melted on the inside. Also I can no longer game with more than 1 display. Obviously if I could turn up the fan I would but I think if it gets any louder my mrs will throw my PC out the window lol.
 
71 c would not cause a system reboot that isn't close to hot enough could try removing the hdd bays or as many as you can so the front fan can push air through unobstructed but the first things I am thinking of if I am crashing is unstable overclock or failing power supply.

Maybe the vrm chips on the card are not being cooled properly did you check those temps ?
 
71 c would not cause a system reboot that isn't close to hot enough could try removing the hdd bays or as many as you can so the front fan can push air through unobstructed but the first things I am thinking of if I am crashing is unstable overclock or failing power supply.

Maybe the vrm chips on the card are not being cooled properly did you check those temps ?

At the moment my CPU is oc'd on ASUS step-up to 3.70 on i7 950 GPU isn't oc'd at all. I'll remove the hdd bays not being used (don't know why I didn't think of that) and I don't know how the check the vrm chips... Any software that can tell me that or would I have to grab a temp monitor from work?
 
Thanks Dude.

Idle I have:
GPU Temp #1: 36c
GPU Temp #2: 35.5c
GPU Temp #3: 37c

From what I can see online the GPU Temp #3 is VRM. I'll be jumping on some BF3 soon so after an hour or so I'll check temps again
 
If it's rebooting, it's not because it's overheating at them temps. lol

Likely rebooting due to the game itself crashing out or an driver error.

My 6950 (uses the same cooling design as yours) runs at over 61'c idle and 70/80'c load when gaming in BFBC2 or BF3 for hours at running three monitors at 5760x1080/1200
 
If it's rebooting, it's not because it's overheating at them temps. lol

Likely rebooting due to the game itself crashing out or an driver error.

My 6950 (uses the same cooling design as yours) runs at over 61'c idle and 70/80'c load when gaming in BFBC2 or BF3 for hours at running three monitors at 5760x1080/1200

It hasn't rebooted in a while and I didn't think that it would @ 70 odd degrees but majority of the restarts occurred whilst playing mainly BF3 on 2 monitors and considering the plastic around my GPU has melted I think it's safe to say the GPU was definitely overheating
 
Back
Top Bottom