Overheating and Extremely Loud GTX 980Ti

VideoklipBG

Commendable
Jan 10, 2017
52
0
1,560
Hello.
I think I've mentioned something about my GPU once in old thread but there was no discussion about and also it doesn't matter.

So in short - My Gigabyte GeForce GTX 980Ti G1 Gaming Windforce 6GB 384bit was bought around year and a half ago or maybe year and a couple months. The Setup overall was built around 08-09 2016.

I've had many problems with almost everything (Incompatibility at some point, problems, defective parts, etc.) and within a couple weeks I had to return and replace a Motherboard, PSU, CPU Cooler and 2x GPUs... Don't ask what and why, it's just my "luck"... 😀

The problems were from defective parts to coil whine from the PSU and the GPU(s), and many other...


So now I'm with the 980Ti Windforce Edittion by Gigabyte (Which is known to have some coil whine but is not significant for me) And many reviews, people, opinions say that it's very quiet, cool and it's performing pretty well in almost everything.

Well... It's exactly the opposite... I don't know if some part of the issue(s) is with the cooler design, the chip itself or something is defective.

When everything is idle (5-10% RAM usage, 0-1% max CPU usage, 0% GPU usage) the card is around 35 to 50°C. The ambient temperature of the room is around 21-22°C (Autumn - Winter).
In conditions which nothing is overclocked, added voltage, power limit increased, everything is stock.

When playing couple of very popular not-heavy games (Like CS:GO, Minecraft, Euro Truck Simulator 2, War Thunder, Paladins, etc.) in Low to Medium settings (No Anti-Aliasing, no effects, much foliage, GPU intensive things) and the GPU usage is approximately 20% to 30% on average - The card is staying minimum at 60°C (Never falls under).

If for example: I'm not streaming, recording, doing such heavy tasks and things, just playing casually with again - low to medium settings (Which results in average 25% - 30% to 40% peak GPU usage and 20% - 30% CPU usage) everything becomes worse and worse with the time running these games and doing the things.

60-70°C minimum / average, 75-80+ after 5 or 10 minutes... In the best cases - 87°C with up to 50% usage. I've reached 98°C before with some benchmarks and programs, but this is not just insane... I don't actually know...

I've tried running some benchmarks, even today for example Unigine Heaven, Unigine Valley, Fire Strike and similar... The fans are louder than hair dryer or vacuum cleaner. Almost exactly like turbine / jet engines.

The case airflow is not that bad, most of the cables are sorted at the back of the case even if there's no much space. The SSD and the HDDs are sitting with 1-2 fingers space between each other. The front fan is very quiet, but it's working perfectly fine, the airflow can be felt at the bottom of the GPU (under the fans).
The exhaust fan is the CPU Cooler (Be Quiet! Dark Rock 3) which is blowing out the air from iside ot outside very well.

Side panel open, table / desk fan blowing to the motherboard - Nothing helps even with a couple degrees.
Using noise cancelling headphones doesn't help at all, because the fans are making little bit of "whining" like noise, plus the air flow noise between the copper pipes.
(40-50% Fan speed or 1800-1900RPM is the maximum that is not annoying and terrible even if can be heard a little bit outside of the room).

I've tried RMA before for overheating issues (Couple months after I built the setup) and "no one cannot do anything"... Because the GPU is working "fine" without fault or "big problems"...
So in practice I can play only very light games on the lowest settings possible and with locked framerate at 60 or 120 (Because I have 60Hz dual-monitor setup, but I'm gaming only on the main monitor. The secondary is helping me a lot when multi-tasking.)

I'm pretty sure that if I ruin the warranty (which have around half a year left) and apply custom thermal compound like Arctic MX-4 or Thermal Grizzly - That will not help or maybe with a half degree celsius down...

I'm sorry for this insanely long post and for the mistakes, because my English is not that good.
 
The 980ti's are not as cool running as the 10 series, but surely you should not be complaining about noise. Especially not on that card. I honestly thing the reason is most likely thermal paste. If the card is running and you don't notice anything on the screen then its definitely running just fine. However, with the fans spinning that chip shouldn't be that hard to control thermally. If you are sure that your ambient is 20-22C than you should not have a problem.

up to you if you want to try and get a replacement or replace the paste yourself. I would call them and just tell them it gets loud and then crashes. IF they receive it they will likely change the paste and send it back.

As for me, i would do it myself. I do a lot of this stuff and I buy some of the better pastes in large quantities for repairing and modding not only mine, but other's PC's as well.
 
Full pc specs please, case model also.
Post screenshot from MSI Afterburner hardware monitoring window during game (when high temperatures are reached). Graps in screenshot: cpu usage/temp, gpu usage/temp, fps.
Hardware monitoring window looks similar to this:
Monitoring.png
 
Excuse me for the delay, I just now captured screenshots without any certain tests or playing around to get the absolute maximum temps under certain load, but these are like "starting to the high temps" so if I start playing some matches in a game or something the things will be worse.
Running CS:GO and after this ETS2 on low to just about medium settings, framerates locked to 200, 120 and 60 with VSync at the end on the second "test". After shutting down the PC and waiting for everything to cool down completely I just started some games again and after running for a couple of minutes I captured the monitoring graph.

It's looking pretty normal at the beginning, where the GPU and the Games are warming and starting up. There are many Alt+Tab-bing to look at the temps, usage and the other stuff.

This is just starting and the lowest - mid temperatures when limmiting by graphics, resolution and framerate in games.
The fan speed is really high, over 50% minimum / average which makes super loud noise, not only from airflow. Over 1800-1900RPM it's pretty much terrible... (Maybe on this card, I don't know if Gigabyte Windforce edittions are not supposed to be in a person's room over 40-50% fans speed... 😀)
That doesn't made so much sense, but anyway.

First "test run" for less than couple of minutes:

https://i.imgur.com/3rFEQWC.jpg

After restarting the PC for 10-15 minutes, waiting to cool down and with side panel open:

https://i.imgur.com/x74orZw.jpg

I wasn't able to capture screenshot exactly where the game(s) is running and the temps are jumping, because even if using windowed / windowed fullscreen mode when clicking away for example on MSI Afterburner the game is dropping the framerate and "idling" like Alt+Tab-bing...
When I'm able to record more accurate results from medium - high load and the temps I will try to share them. Or try somehow to record a video, I don't know.

I usually use custom fan curve with constant 20% fans on idle and optimization not to sound like jet engine in mid to more - high load. Also slightly better temps but with 0.5% to 1% difference...

All specifications are in my profile's information tab, but here are they:

Motherboard: Gigabyte GA Z170 HD3
PSU: Seasonic X-Series, 1050W, 80 Plus Gold
CPU: Intel Core i7 6700K 4.00GHz 4-Cores, 8-Threads
GPU: Gigabyte GeForce GTX 980 Ti G1 Gaming Windforce 6GB GDDR5 384bit
RAM: 16GB (2x8GB) Adata XPG 2666 MHz
CPU Cooler: BeQuiet! Dark Rock 3
SSD: Adata SX930 XPG 120 GB
HDD 1: Seagate SV35 2 TB 64MB Cache
HDD 2: Toshiba P300 1 TB 64MB Cache
Case: CoolerMaster K350

*Quick Edit: If it's not clear, because I didn't explain properly... Like always... These are the best - low temps when starting up things. Low to medium load on the GPU for 50-60 seconds to 2-3 minutes overall;*
 
Didn't see cpu usage/temps in those screenshots.
Anyway - GPU temperatures don't seem troubling (63-74 C).

Heat buildup means that airflow inside the case is not sufficient. You may want to dial up rpms of frontal intake fan (or change it to some other higher performance model). Also rear exhaust fan may be necessary.
 
The average CPU usage in most games is around 20-30%, if a game is running with more frames and the GPU is used a bit more (40-50%) then the CPU also has around 5-10% more load.
The idle CPU temperatures are 24-30C, under 100% load with benchmark, stress test or some software it varies around 60C to almost 70. Never exceeds that, it's very quiet and stable.
Of course with stock voltage and clock speed. Only X.M.P is enabled, because I have 2666MHz 2x RAM Modules and this utilizes the full speed and performance.

https://i.imgur.com/Ku0ZCC4.png

I've tried full speed on the front fan (Which is a little bit loud) and the CPU fan with more "aggressive settings". This doesn't help much, because like I said before I think - The side panel has many ventilation holes and fan mounting option. When removed the temps are about the same. Tried also desk fan directly facing the motherboard (with the ambient air temperature) and this also doesn't help much... Only in the CPU cooling a bit, because the big heatsink is staying in front.

And also - these GPU temps are of the lowest they can be on starting application or load. If it stays at 70, even at 70 and something it will be more than paradise.. 😀

So maybe I will try again to RMA the card, but there is one, like problem... I'm from Bulgaria, not very big European country. Here we have very unreliable companies and there are many problems / issues especially with hardware returning and replacement. Very few more reliable companies overall. And like I said in the very first post - I've replaced a lot of things, and also 980 Xtreme Edition, I think.. With 980Ti because of crashing, artifacts and other issues. And now again with this 980Ti... It wasn't very easy and I was almost screwed for a couple reasons...

As a last as resort I think I can try new thermal paste, but it's very unlikely to give a different result.. One part of the issue, I think is the size of the fans, because they are like 80mm or something, I don't know exactly on this model.. Another part may be bad chip / broken / now working properly heatsink and cooling overall, but I don't know why and how if that's the case.

*Edit: The front fan model is CoolerMaster Silencio FP 120 PWM*
 
Another option is to reduce GPU overclock.

Your card is factory overclocked to a significant degree.
Nvidia reference clocks for gtx980ti are 1075/1000 (boost/base).
Your card is overclocked to 1241/1152 (gaming mode) and 1279/1190 (OC mode).

Looks like your card cooling solution can't handle the overclock.
 
Hmm, I don't think underclocking will help much. That's only thing that I've not tried for now.. But adding +30 or +50 MHz on the core / boost clocks and +80MHz on the memory clock the temps are going like 1-2 degrees up..

So basically I can over or underclock a little bit and nothing will happen that much. Which is very weird and strange, but.. That is the case.

Yes, this model board is factory overclocked to some point and it has two options - Gaming and OC Mode. And in the most cases I don't find much difference, maybe on some benchmarks, but nothing more than that.

Kind of very strange and weird "issues" or something, but yeah..
 

TRENDING THREADS