• Some users have recently had their accounts hijacked. It seems that the now defunct EVGA forums might have compromised your password there and seems many are using the same PW here. We would suggest you UPDATE YOUR PASSWORD and TURN ON 2FA for your account here to further secure it. None of the compromised accounts had 2FA turned on.
    Once you have enabled 2FA, your account will be updated soon to show a badge, letting other members know that you use 2FA to protect your account. This should be beneficial for everyone that uses FSFT.

Question about my GTX 690 Temps

N

northfaceseen

Limp Gawd
Joined
Dec 28, 2011
Messages
129
Hey,

I got a EVGA GTX 690 last week and I have noticed the one of my gpu cores GPU 1 runs about 8-9C higher than GPU 2. I have checked the paste on both cores and applied Phobia Grease and I also checked to see if my case flow was obstructing my gpu fan flow, but my cores are still very far apart on temps. From what I have tested if I block airflow from flowing out of the front of the card my GPU 2 temps rise, so I am thinking GPU 1 is the core closest to the I/O panel. My question is, is it normal to have such a wide degree difference between cores if the heatsinks are making proper contact and the gpu fan flow is unobstructed, and would I have a just case in RMA'ing this card?

Thanks for your guys input
 
Current Temps my card hits while running Heaven 3.0 at 2560x1600 extreme tessellation are GPU 1 82-83C and GPU 2 72-73C with fan running around 60-70% in a 70F room.
 
I certainly dont think its a reason to RMA. Those temps are not unreasonabley high for a dual GPU doing heavy benching. I would be surprised if both GPUs run the same temp, and that seems like more of a difference than I would expect- but its doesn't seem bad. CPU core temps are often that different, and even the heaven bench may use one GPU more than the other. If you had a 20 degree difference I would say thats a reason for concern but I have a 10c degree most of the time in my CPU benchmarks, not sure why it would be that different for GPUs.
 
I'm not sure what the safe temp is for the 690 for core temps but I would get that info off the nvidia web site or call customer service. Any time I see temps in the 80's it worries me but my gtx 480 ran in the 90's and it was ok because nvidia said it could run at 95C without causing failure.
 
Is there a GPU load difference between the two? I've found on mine that GPU2 is often 10-30% below GPU2 load depending on what other things I have going on on side monitors or in the background. This is mostly with V-sync on and neither GPU fully loaded.
 
Are you using more than one monitors? I have two monitors hooked up and my GPU 1 is roughly 3c higher than my GPU2 on idle. If you are then its probaly just kicking in a different performance state/load than gpu2.
 
My 690 GPU's are almost always at different temperatures. My GPU A is around 5 deg C hotter than GPU B most of the time. I'm sure there are many factors. In SLI systems, doesn't one GPU always have a little higher workload than the other GPU?
 
the gpu 1 is always the main gpu and have more load over the other. I guess in other words its just gpu1 running along everything and the 2nd one actually kicks in when its needed of course its pretty much active all the time but still.
 
Thanks for you inputs. I do agree that one GPU will get more of the workload. I guess my concern is that my $1,050 video card is running abnormally high on temps with the fan being close to the max. For example today I let Heaven run for 1hr and my temps with all factory clocks and power were hitting 84C on GPU 1 with my GPU fan running 75-80%. From what I have seen on reviews alot of people are rarely hitting above 80C on stock settings with the fans rarely going above 60%. I am concerned the card I got barely squeaked by nvidia's tests.
 
I was playing some BF3 today and noticed that my GPU temps are actually pretty close to 8-9 deg C difference like the op was saying. I forgot to take note of the fan speed but I think it usually is around 60% and my GPU temps are usually high 60's to low 80's. All stock frequencies/voltages.

If the heatsink is making good contact with the GPU's then the only thing you can do (with stock cooler) is to move some more air through your case. Maybe see what your temps are like with your case open.
 
Are you running multiple monitors with that 690? If so that is known to cause a 10 degree C upward shift in some nvidia cards. For instance my GTX 480 would jump 10 degrees or so when I simply plugged in my second monitor to it.

I imagine this would account for the problem.
 
That difference in temps doesn't seem too much out of place. In my quad sli setup after 15-20 minutes of heaven at max settings, GPU 1 is usually at 80, GPU 2 at 74, GPU 3 at 70 and GPU 4 at 68. So there IS a 6-7 degree variance in cores of the same card.
 
You're not using the GPU shroud on the HAF-X are you? If so, don't. The 690 exhausts hot air out of both sides of the card. If you use the GPU shroud with a fan facing the card, it will make 1 core hotter than the other.
 
northfaceseen said:
Hey,

I got a EVGA GTX 690 last week and I have noticed the one of my gpu cores GPU 1 runs about 8-9C higher than GPU 2. I have checked the paste on both cores and applied Phobia Grease and I also checked to see if my case flow was obstructing my gpu fan flow, but my cores are still very far apart on temps. From what I have tested if I block airflow from flowing out of the front of the card my GPU 2 temps rise, so I am thinking GPU 1 is the core closest to the I/O panel. My question is, is it normal to have such a wide degree difference between cores if the heatsinks are making proper contact and the gpu fan flow is unobstructed, and would I have a just case in RMA'ing this card?

Thanks for your guys input
Click to expand...

I've noticed temp differences in my 690 as well. I mounted a Koolance waterblock (a process I was hesitant about given the cost of the card) and I've been thinking that perhaps my application of pk-1 thermal compound may have been lacking but I really don't want to drain my loop and dismantle the thing because it's such a hassle. Considering that the temp difference is only 2-5c I don't think I'll bother. When gaming my temps are about 42c on one GPU and 47 on the other. The difference is less at lower temps so perhaps your higher difference with higher temps isn't abnormal.
 
I went ahead and RMA'd my card and just got my replacement. My new card is way better with its temps. For example my old card would hit 84C on GPU 1 in Heaven and my new card both of the GPU's are with 1C of each other while maxing at 74C. My fan speed is also running about 10-15% lower than my first card with auto profile. I feel much better with this new card.
 
Temp difference of upto that level are normal across GPUs. Just chill and relax. And after heavy benching 82-83 is not bad at 70% fan speeds.
 
Update: My card no longer shows temperate difference greater than 1 or 2c and since "curing" the thermal compound I put on my block temps have dropped quite a bit.

I'm glad to hear you got a good card ;-)
 
You must log in or register to reply here.
Back
Top