F5F Stay Refreshed Hardware Desktop GPU temperatures rise sharply and computers stop working during game launches?

GPU temperatures rise sharply and computers stop working during game launches?

GPU temperatures rise sharply and computers stop working during game launches?

Pages (2): 1 2 Next
S
Shad0wHydra13
Senior Member
716
02-27-2022, 12:03 AM
#1
Hello.
I’m experiencing issues with my PC when launching games. The GPU fans immediately hit full speed and temperatures rise to about 80°C, with hotspots near 100°C. After that, the screen goes black and I need to restart the system. Everything functions normally during idle. My assumption is that the GPU thermal management is faulty, possibly due to a damaged heatsink.
I’ve tried:
- BIOS update
- Full driver reinstall using DDU
- Using older driver versions
Specific details:
CPU: Ryzen 5 5600
GPU: Nvidia RTX 3060Ti
Motherboard: B550M Steel Legend
RAM: 16GB DDR4
Power supply: Coolermaster 700W Gold
S
Shad0wHydra13
02-27-2022, 12:03 AM #1

Hello.
I’m experiencing issues with my PC when launching games. The GPU fans immediately hit full speed and temperatures rise to about 80°C, with hotspots near 100°C. After that, the screen goes black and I need to restart the system. Everything functions normally during idle. My assumption is that the GPU thermal management is faulty, possibly due to a damaged heatsink.
I’ve tried:
- BIOS update
- Full driver reinstall using DDU
- Using older driver versions
Specific details:
CPU: Ryzen 5 5600
GPU: Nvidia RTX 3060Ti
Motherboard: B550M Steel Legend
RAM: 16GB DDR4
Power supply: Coolermaster 700W Gold

K
Komodo88
Senior Member
749
02-27-2022, 12:09 AM
#2
Check Realiability History/Monitor and Event Viewer for any error messages, alerts, or informational logs around the time of BSODs.
= = = =
Which games? Source?
Restart drivers directly without using third-party tools or installers.
Download them manually from the manufacturer's site, reinstall, and adjust settings. Ensure the source is genuinely from the manufacturer's domain.
Just because a product name, model, or brand appears in the link or path isn't proof the site is the official one
= = =
Power down the system, unplug it, and open the case.
Remove dust and debris carefully.
Inspect all connectors, cards, RAM, jumpers, and case connections to confirm they are securely installed.
Use a bright flashlight to look for any signs of damage.
K
Komodo88
02-27-2022, 12:09 AM #2

Check Realiability History/Monitor and Event Viewer for any error messages, alerts, or informational logs around the time of BSODs.
= = = =
Which games? Source?
Restart drivers directly without using third-party tools or installers.
Download them manually from the manufacturer's site, reinstall, and adjust settings. Ensure the source is genuinely from the manufacturer's domain.
Just because a product name, model, or brand appears in the link or path isn't proof the site is the official one
= = =
Power down the system, unplug it, and open the case.
Remove dust and debris carefully.
Inspect all connectors, cards, RAM, jumpers, and case connections to confirm they are securely installed.
Use a bright flashlight to look for any signs of damage.

D
Demorocks8
Member
116
03-03-2022, 06:35 AM
#3
Here are a few examples of games: GTA V, Ready Or Not, BF6 beta.
Performed a clean install using Nvidia, not DDU, and cleaned the PC thoroughly.
Didn't see any GPU damage; radiators appeared normal with no discoloration. A few screws were slightly loose, so I tightened them.
The only error codes I encountered during the black screen were "Windows didn't shut down correctly," which was due to manual off-switching.
PSU cables seemed okay; I swapped the 8pin and 6pin connections on the GPU.
Games are running now, but temperatures remain high at 70-80°C with fans working hard.
D
Demorocks8
03-03-2022, 06:35 AM #3

Here are a few examples of games: GTA V, Ready Or Not, BF6 beta.
Performed a clean install using Nvidia, not DDU, and cleaned the PC thoroughly.
Didn't see any GPU damage; radiators appeared normal with no discoloration. A few screws were slightly loose, so I tightened them.
The only error codes I encountered during the black screen were "Windows didn't shut down correctly," which was due to manual off-switching.
PSU cables seemed okay; I swapped the 8pin and 6pin connections on the GPU.
Games are running now, but temperatures remain high at 70-80°C with fans working hard.

T
TornadoBasak
Member
72
03-06-2022, 03:16 PM
#4
Execute "dism" and "sfc /scannow" using the command prompt.
Goal is to locate and resolve any faulty or corrupt files. It seems manual shutdowns may have caused file damage and configuration issues.
Reference links:
https://www.windowscentral.com/how-use-d...s-10-image
https://www.lifewire.com/how-to-use-sfc-...es-2626161
You can search online for additional details about "dism" and "sfc /scannow".
T
TornadoBasak
03-06-2022, 03:16 PM #4

Execute "dism" and "sfc /scannow" using the command prompt.
Goal is to locate and resolve any faulty or corrupt files. It seems manual shutdowns may have caused file damage and configuration issues.
Reference links:
https://www.windowscentral.com/how-use-d...s-10-image
https://www.lifewire.com/how-to-use-sfc-...es-2626161
You can search online for additional details about "dism" and "sfc /scannow".

Y
y0rdin
Junior Member
35
03-08-2022, 03:14 PM
#5
i found some damaged files and corrected them with the guide. sfc/scannow no longer detects any corrupt files, so that issue is resolved, but the fans are still behaving erratically and the temperature stays between 70-80°. Possibly the heatsink is the cause.
Y
y0rdin
03-08-2022, 03:14 PM #5

i found some damaged files and corrected them with the guide. sfc/scannow no longer detects any corrupt files, so that issue is resolved, but the fans are still behaving erratically and the temperature stays between 70-80°. Possibly the heatsink is the cause.

L
LEJCS
Junior Member
22
03-10-2022, 07:19 AM
#6
Is this a new issue that has arisen and the card was previously cooling fine beforehand? Or is this a second hand card that has had the problem from the beginning?
L
LEJCS
03-10-2022, 07:19 AM #6

Is this a new issue that has arisen and the card was previously cooling fine beforehand? Or is this a second hand card that has had the problem from the beginning?

J
jjmonkey13
Member
236
03-22-2022, 01:17 AM
#7
If the hot spot reaches at least 100°C (the die usually requires a refurbishment), all fan curves will be disregarded until one of these conditions is met:
-hot spot drops below 100°C.
-hardware stops working because it couldn't meet the first condition.
J
jjmonkey13
03-22-2022, 01:17 AM #7

If the hot spot reaches at least 100°C (the die usually requires a refurbishment), all fan curves will be disregarded until one of these conditions is met:
-hot spot drops below 100°C.
-hardware stops working because it couldn't meet the first condition.

N
Nadiiine
Junior Member
1
03-25-2022, 04:32 AM
#8
It's a used card. A fan problem occurred recently, though the black screens are likely from about a week ago.
N
Nadiiine
03-25-2022, 04:32 AM #8

It's a used card. A fan problem occurred recently, though the black screens are likely from about a week ago.

A
ACarKilledMe
Junior Member
34
03-31-2022, 09:45 PM
#9
Repasting for a GPU should be handled by an expert.
It's more complex than working with the CPU.
A
ACarKilledMe
03-31-2022, 09:45 PM #9

Repasting for a GPU should be handled by an expert.
It's more complex than working with the CPU.

M
mindcraft54
Member
66
04-01-2022, 08:44 AM
#10
It sounds like it needs a repaste although it is worth checking gpuz and confirming the vbios to make sure it has not been messed around.
Depending on your technical level and the type of cooler used by the manufacturer, it is not that difficult. The biggest thing I find is the thermal pads and any damage to them that can be caused. You mentioned a few screws were loose which to me implies that someone may have already had a go at taking this apart. Possibly unsuccessfully.
On the rear of the card there is normally a sticker on one of the screws related to warranty. Was this there?
M
mindcraft54
04-01-2022, 08:44 AM #10

It sounds like it needs a repaste although it is worth checking gpuz and confirming the vbios to make sure it has not been messed around.
Depending on your technical level and the type of cooler used by the manufacturer, it is not that difficult. The biggest thing I find is the thermal pads and any damage to them that can be caused. You mentioned a few screws were loose which to me implies that someone may have already had a go at taking this apart. Possibly unsuccessfully.
On the rear of the card there is normally a sticker on one of the screws related to warranty. Was this there?

Pages (2): 1 2 Next