[SOLVED] MSI RTX 3080 GAMING Z TRIO LHR UNDERVOLT HELP!

Selynelar

Reputable
Feb 19, 2019
186
9
4,615
Hi all,

My concern is that I have sold an MSI RTX 2080 TI GAMING X TRIO and bought an RTX 3080 GAMING Z TRIO LHR card. The previous card had around 80-82 degrees when I was playing for example BF 2042. In Battlefield V it was around 70-75 degrees. I tought it maybe deserved a repaste. But what I realized that new card(3080 GAMING Z TRIO) is having similar temperatures. One of my friend suggested to undervolt the card to reduce the temperature.

First question:
I have done the undervolting to 825mv and 1875mhz,the temp dropped to 74 degree stable in BF 2042. In idle sometimes I see around 50 degrees. I realized that the card boost is only 1830mhz officially. Is it possible that I have damaged the card? Sorry for this silly question.
Second question:
Is these temperatures are normal and in long therm can damage the card?
Third question:
Could someone help me to undervolt the card? Or any suggestion to reduce the temperatures?

My specs:

Cpu/Cooler:I7 10700K/Noctua NH-D15 ChromaxBlack
Motherboard:MS Z490 UNIFY
Case: BE QUIET PURE BASE 500DX
Gpu: MSI RTX 3080 GAMING Z TRIO - all 3 pcie connections connected seperately/ newest driver installed
SSD: Samsung 970 EVo Plus 1TB M.2 Pcie Nvme
RAM: HyperFury 2x16GB 3200mhz
Psz: Seasonic Focus Plus 850W Platinum
Case fans: 3xBe Quiet Silent Wings 3 (3pin) 14cm Fan
Display: ACER XB273UGS

Thanks for your replies.
 
Solution
What's the problem? Nvidia cards will generally run 70-82°C when pushed under punishing workloads, it's part of GpuBoost3, it'll boost until voltages and temps will no longer support such activity. Even medium loads can see those temps as the card will boost as high as it can, to give you best performance.

It's entirely normal behavior under a load.

The idle temps are normal, only in the sense that idle temps respond to airflow in the case. If airflow is lousy, idle temps will be higher. Airflow will be a byproduct of case design, fan layout, fan curves and obstructions.
50c In idle is the max safe Value. But to be on the safe side, i think the temps are above normal
No such thing. Idle temps are case by case depending...

Lutfij

Titan
Moderator
1| I don't think you've damaged anything on the card. Though, did you re-paste the card with a good quality thermal paste? On that note, did you tamper with the warranty void sticker on the back of the card?

2| Prior to removing the RTX2080Ti from your system, did you run DDU to remove all GPU drivers from your platform? When the RTX3080 was dropped in, did you manually install the latest drivers from Nvidia in an elevated command, i.e, Right click installer>Run as Administrator?

3| If you haven't, read through these guides;
View: https://www.reddit.com/r/nvidia/comments/koub76/comment/ghw265v/?utm_source=reddit&utm_medium=web2x&context=3
 

Karadjgne

Titan
Ambassador
What's the problem? Nvidia cards will generally run 70-82°C when pushed under punishing workloads, it's part of GpuBoost3, it'll boost until voltages and temps will no longer support such activity. Even medium loads can see those temps as the card will boost as high as it can, to give you best performance.

It's entirely normal behavior under a load.

The idle temps are normal, only in the sense that idle temps respond to airflow in the case. If airflow is lousy, idle temps will be higher. Airflow will be a byproduct of case design, fan layout, fan curves and obstructions.
50c In idle is the max safe Value. But to be on the safe side, i think the temps are above normal
No such thing. Idle temps are case by case depending. Load temps define safe limits. A gpu could have a load Delta of 10°C and still be perfectly fine, as long as that Delta wasn't with a max of 90°C. Most gpus will idle slightly above cpu idle temps, just due to airflow patterns, so even with stock Ryzens running 40-60°C idle temps, 50ish°C for a gpu isn't all that abnormal.
 
Last edited:
Solution

Selynelar

Reputable
Feb 19, 2019
186
9
4,615
Thanks for everyone's advise and help.

1| I don't think you've damaged anything on the card. Though, did you re-paste the card with a good quality thermal paste? On that note, did you tamper with the warranty void sticker on the back of the card?
No. I did not touch this card. The repaste was about the 2080 Ti and I did not touch that one as well. Because still it has warranty on it. And I did not want to ruin that one.

2| Prior to removing the RTX2080Ti from your system, did you run DDU to remove all GPU drivers from your platform? When the RTX3080 was dropped in, did you manually install the latest drivers from Nvidia in an elevated command, i.e, Right click installer>Run as Administrator?

I always install like that. Safe mode -->DDU--> Remove GPU driver and Turn off computer to install the new GPU --> In options: Prevent windows to download drivers
Install the new card with run as admin.

The case fans at the front, how many do you have and is any part of the lowest fan below the gpu?
I have two 14cm intake dc fans in front and one 14cm dc fan exhaust at the back. On the top nothing. The case should have good airflow. That is why I have chosen that. Cpu has awesome temps btw. (knock-knock-knock :) )
Would the bottom fan be blowing any air below the gpu?
It is supposed to blow straight onto the Gpu.
Any fans set to exhaust hot air out of the case?
One 14cm dc fan exhaust at the back


So overall as I understand it is nothing wrong these temperatures?
But still do you think should be in long term any problem if the cards runs usually on these temps during load?
Do you suggest to do undervolt or not neccessary?
And at the end so do you think that playing with undervolt where I have undervolted the card and raised the clock speed did I do any harm to the card or I dont need to worry about that?

Thanks for your help. I really appreciate that.
 

Karadjgne

Titan
Ambassador
Undervolting is the new Black, so to speak. It used to be OC was the way to go to get more performance, but Intel, nvidia, amd are pushing their components to 90-95% of possible performance anyway, so that last 5% means seriously higher power consumption for little to no gain.

Now with things like GpuBoost3, the card is pushing itself as high as the silicon can take and remain stable, limited by efficiency between voltages and temps. Ryzen goes a step further, maximum performance is gotten at the 60°C mark, the processor 'thinks' at maximim efficiency, regardless of clock speeds.

So tinkering with voltages and lowering temps can easily see higher, longer sustained boosts, which equates to higher fps, both highs and lows. Upto a point. That point will change depending on the card, the silicon, airflow, components etc. It's a fine line to balance voltage limits, loss, usage vrs temps.

Temps aren't an issue as such. The card will lower its boost, and voltages, automatically when it reaches 83°C. You see a sustained temp, but if you had a stable fps output to start with you'd see a small decline in fps output as the boost lowers. Since games are never stable, what you get is you just don't get the exact fps possible, which you don't see anyway.
 

Selynelar

Reputable
Feb 19, 2019
186
9
4,615
Thanks a lot to all of you. Unfortunetaly I cannot mark as best answer to all of you. But all the answers have been usefull. At the end I did not do any undervolt so I ended up playing with the case setup in the bios which was quite helpful. Now temps are low even on cpu as well. I just set the temperatures on case fan and voltages when to raise rpm and now seems like it helped a lot.

Thanks again to All!!!