Question GPU Temperature and Voltage issue

Versuz

Reputable
Jul 6, 2017
53
3
4,565
5
This question is related to this post, but I'm posting this separately to identify the issue quickly, sorry for that.
As I mentioned in the above post lately my pc started getting BSOD (VIDEO_TDR_FAILURE ) time to time, so I installed GPU-Z and Hwinfo to monitor the system behavior. According to these tools the GPU tem reaches 88C sometimes and GPU PCIe +12V input Voltage drops to 11.2V, 11.1V and 11 sometimes . I'm not an expert and also not sure whether these software tools are reliable enough to monitor hardware power usage, but I found these values are not normal.
Is this temperature harmful to the GPU, and if these software tools not reliable is there any other way to monitor the GPU voltages, because I think I'll have to change my PSU if these values are correct.
GPUZ sensor log
HWinfo report



specs of the pc -
Gigabyte b450m ds3h
Ryzen 3600
PNY ddr4 3200mhz 8gb x 2
MSI Ventus-2X-OC RTX 3060ti
Gigabyte G750H
Antec neptune 120 argb AIO
Addlink S70 250GB
WD Blue 1tb + Seagate 1tb
MSI mag forge 100m + extra 120mm MSI argb fan
Windows 10 64bit
 
Last edited:
May 7, 2022
26
4
45
1
Aside from the common solutions everyone will probably tell like: make sure fans are spinning, set a better fan curve make sure there is no dust etc...)

Try rolling back drivers or get an older version (as old as you can without losing features or support for games you need)
from my experience some nvidia drivers easily cause higher voltages than usual hence higher temperatures.
I remember an update in 2019 or 2020 that made my 1060 run ~10C higher on average JUST after I updated drivers (I noticed because I instantly saw FPS dropping as it thermal throttled), tried asking on reddit and everyone called bs, but I still believe to this day bad drivers from Nvidia caused it.
also try out studio drivers, they are usually more stable.
 

Versuz

Reputable
Jul 6, 2017
53
3
4,565
5
Aside from the common solutions everyone will probably tell like: make sure fans are spinning, set a better fan curve make sure there is no dust etc...)

Try rolling back drivers or get an older version (as old as you can without losing features or support for games you need)
from my experience some nvidia drivers easily cause higher voltages than usual hence higher temperatures.
I remember an update in 2019 or 2020 that made my 1060 run ~10C higher on average JUST after I updated drivers (I noticed because I instantly saw FPS dropping as it thermal throttled), tried asking on reddit and everyone called bs, but I still believe to this day bad drivers from Nvidia caused it.
also try out studio drivers, they are usually more stable.
Tried both old drivers and studio drivers but still the same issue.
 
Reactions: Versuz

Versuz

Reputable
Jul 6, 2017
53
3
4,565
5
Assuming what HWiNFO is reporting is accurate, which we have to unless you're willing to test voltages yourself, the PCIe voltages are not within spec. According to https://forums.tomshardware.com/threads/pcie-cable-12v-voltage-tolerance.3225712/ (which is referencing https://xdevs.com/doc/Standards/ATX/ATX12V_Power_Supply_Design_Guide_Rev1.1.pdf) , the minimum voltage allowed out of the 12V rail, which feeds the PEG connector, is 11.4V.

My money is more on this being a power supply issue.
Initially that's what I noticed but now getting this BSOD while doing nothing, gpu temp around 50-55C and the voltage is 11.8 - 12.2V
 

ASK THE COMMUNITY