cancel
Showing results for 
Search instead for 
Did you mean: 

Gtx 980 vrm

wheneva14
Level 7
Hello,

I recently returned my MSI 970 and bought a Asus 980, i was planning to overclock the card but after benchmarking with out of box settings, the VRM temperature caught my attention. It is @ 100C when running Heaven Bench,

Is this normal? Because the past few cards ive owned never had these high VRM temperatures.

What is the safe temperature zone for the VRM?

Thank you
11,123 Views
18 REPLIES 18

Chino
Level 15
Which GTX 980 model did you buy? Also what program are you using to read the VRM temperature? If possible, post a screenshot.

wheneva14
Level 7
Hello,

I bought the Asus STRIX-GTX980-DC2OC-4GD5 from newegg
I used HWINFO64 for the VRM temperature.
Here is a little screen shot
46562

Chino
Level 15

1stcowgirl
Level 7
wheneva14 wrote:


Is this normal? Because the past few cards ive owned never had these high VRM temperatures.


ill try to keep it short,.

THiS iS ASuS!!!! 100C for VRM is OK for these cards

if you want to make sure, READ THIS: (read what MarshallR@ASUS said on the 3rd page)
https://rog.asus.com/forum/showthread.php?46389-ASUS-GTX780Ti-DC2OC-_-VRM-are-Very-Very-HoT!
ºi7-4790k, ºZ97-A, º4x4GB Ripjaws X 1600mhz, ºASUS GTX 780Ti OC [DC2OC] ºCorsair GS700, ºCM 690 IIIº LG E2241T ºWin 7 Ulti' SP1 x64

BRSxIgnition
Level 9
So 105C is the end of the safe zone for 970 and 980 cards?

My Strix 970 hits 98C under regular gaming use and 100C after running the Valley Benchmark for 45 minutes.

All the other temperatures are as follows:

  • CPU: 18° MIN, 57° MAX, 36.7° AVG
  • Motherboard: 23° MIN, 39° MAX, 33.3° AVG
  • GPU (Core): 20° MIN, 71° MAX, 52.6° AVG
  • GPU (VRM): 23° MIN, 100° MAX, 72.5° AVG

BRSxIgnition wrote:
So 105C is the end of the safe zone for 970 and 980 cards?

My Strix 970 hits 98C under regular gaming use and 100C after running the Valley Benchmark for 45 minutes.

All the other temperatures are as follows:

  • CPU: 18° MIN, 57° MAX, 36.7° AVG
  • Motherboard: 23° MIN, 39° MAX, 33.3° AVG
  • GPU (Core): 20° MIN, 71° MAX, 52.6° AVG
  • GPU (VRM): 23° MIN, 100° MAX, 72.5° AVG


To all friends interested:

i find that the best test is using a game like BF4 or WatchDogs.
***NOTE***
the VRM temps may have 10c change between season of Summer vs Winter.
my 780ti vrm can get to 109c on a hot summer day.
and now is winter so its staying around the 95c-100c.
oh yea, another test can be the use of DSR in nVidia control panel

If you are worried then the only solution is using GPU-Tweak to make the card fan faster (it will keep your VRM at 90c MAX)
this is tested all over!
in the summer the par between the gpu & vrm is 20c-25c (meaning: gpu=80c -> vrm=100c+) _ raising the fan speed will solve the ratio, gpu will be reduced to 70c and vrm will reduced to 90c)

p.s.
Dont worry, ASUS Build Can take this kind of heat! _ (tested!)
ºi7-4790k, ºZ97-A, º4x4GB Ripjaws X 1600mhz, ºASUS GTX 780Ti OC [DC2OC] ºCorsair GS700, ºCM 690 IIIº LG E2241T ºWin 7 Ulti' SP1 x64

Chino
Level 15
72.5C average is well within the safe zone.

Just to put your mind at ease:

MarshallR@ASUS wrote:
Compared to most components heat won't affect the chokes so much in the long run as they are environmentally sealed, non-moving parts. Compared to capacitors that are 'wet' inside, which is why we use the high quality 105C, 10s of thousands of hour life-span Japan made variety. It's more important here than your motherboard. What really kills stuff is humidity+heat. But, again, being environmentally sealed they won't suffer as bad as other parts can do. The old-old-old chokes that were exposed to air would oxidise and/or vibrate, but that's no longer the case. I don't want to quote our marketing because it seems to take away from a technical discussion, but the "super alloy" part is genuinely an upgraded alloy from the low-profile or low-cost chokes on reference cards, which is part why DCII is a bit more expensive than reference.

As for the PCB itself getting hot - it's also meant to as it's unavoidable as the solder joint is a better conductor than any thermal paste. It's mitigated as part of the PCB design: the tracing and layers. You can remove/replace/add thermal components that will lower the temperatures certainly, which is why after market cooling has always existed. The thermal pad might not be the 'best' solution but it's the most long-lasting. You can replace it with thermal paste and maybe every 6 months have to replace it. As a company we can't ask our users to do that. You could run it under water with everything at 30C and it would last ... 20 years! Will you still be using it in 20 years though?

There's always a cost/benefit on our side to doing lots of things: continually adding more thermal mass; the quality of components; expected lifespan of the card and making sure it sees out warranty; and being price competitive all weigh in. No company will(or, should) cost save to the point of increasing your RMA numbers above the inevitable few %age points, because it's grossly uneconomical and you don't rebuy next time.

BRSxIgnition
Level 9
Thing is, that average is taking into account the idle temperatures from a good hour or so prior to the benchmarks. (Around 30 - 40c)

When actually playing the game/benchmarking, the VRM is constantly at that temperature. (98c+)

For clarity sake, I will run the tests directly after start up to avoid the low idle temperatures and record those.

Chino
Level 15
Haven't used HWINFO64 in a while. But I think you can configure the OSD to show what you want to see, correct? If that's so, just configure it to show the VRM temperature and go game/benchmark while monitoring in real time. Can't get any more accurate than that. 😉