GPU Shutting Down Under Load

[Ythia]

Hey Everyone,

This is the first time I run into this sort of Issue so I wanted to know if anyone was familiar or had experienced this sort of problem and what solution they had found for it.

First things first, here is the full config:

• CPU : AMD Ryzen 7 7800X3D
• GPU : Asus ProArt OC GeForce RTX 4070 Ti SUPER
• MB : Asus TUF GAMING B650-PLUS
• RAM : Corsair Vengeance 64 GB (2 x 32 GB) DDR5-6000
• PSU : Corsair RM850x 850 W 80+
• Cooler : Deepcool Assassin 4S
• Storage :
  •  Samsung 970 Evo Plus 1 TB
  •  Samsung 970 Evo Plus 2 TB
  •  Seagate ST2000LX001 2TB (Previously used as an external drive on my old laptop)
• Case : Fractal Design North
• Fans : Lian Li SL-INF
  • 1 x 120mm (Rear Exhaust)
  • 3 x 120mm (Front Intake)
  • 2 x 140m (Top Exhaust)
  • All connected to the same L Connect controller
• OS : Windows 11 Home
• Bios Version : American Megatrends Inc. 3024 8/2/2024

 

Now for the situation:

I've had a few time while my computer was under load, mostly playing games, my screens would turn black and when I looked at my computer case the GPU was visibly turned off. Fans not spinning and red light indicating it's powered but turned off. Everything else was working fine, case fan spinnings game audio still working, discord still working etc... After a few moments the GPU would sometimes turn back on... But I'd still need to hard reboot the entire PC to get things working back again...

 

At first I checked the temps but they seemed fine... CPU max temp was about 70°C and GPU max temp was about 60°C . So I immediately thought it might be a PSU issue... Downloaded OCCT and started a Power Stress Test which went on for 4 minutes before having the issue happen and the GPU turn off in the exact same manner... However... I'm used to PSU failures causing a complete system shutdown and not just the GPU turning off so before I go off and buy an RM1000x I just wanted to check with you guys if there wasn't any other issue that could be happening.

 

Note I have the following BIOS Settings:

- EXPO 1 turned on for RAM

[Marsrover91]

So I think you might be right on the PSU issue, had almost this exact thing happen. I would buy one with the intention of returning it and use it to test if it works great if not at least you know that it's not PSU. The only other thing could be the card itself but I haven't seen one just shut down during load.

[Blue4130]

49 minutes ago, Ythia said:

Hey Everyone,

This is the first time I run into this sort of Issue so I wanted to know if anyone was familiar or had experienced this sort of problem and what solution they had found for it.

First things first, here is the full config:

• CPU : AMD Ryzen 7 7800X3D
• GPU : Asus ProArt OC GeForce RTX 4070 Ti SUPER
• MB : Asus TUF GAMING B650-PLUS
• RAM : Corsair Vengeance 64 GB (2 x 32 GB) DDR5-6000
• PSU : Corsair RM850x 850 W 80+
• Cooler : Deepcool Assassin 4S
• Storage :
  •  Samsung 970 Evo Plus 1 TB
  •  Samsung 970 Evo Plus 2 TB
  •  Seagate ST2000LX001 2TB (Previously used as an external drive on my old laptop)
• Case : Fractal Design North
• Fans : Lian Li SL-INF
  • 1 x 120mm (Rear Exhaust)
  • 3 x 120mm (Front Intake)
  • 2 x 140m (Top Exhaust)
  • All connected to the same L Connect controller
• OS : Windows 11 Home
• Bios Version : American Megatrends Inc. 3024 8/2/2024

 

Now for the situation:

I've had a few time while my computer was under load, mostly playing games, my screens would turn black and when I looked at my computer case the GPU was visibly turned off. Fans not spinning and red light indicating it's powered but turned off. Everything else was working fine, case fan spinnings game audio still working, discord still working etc... After a few moments the GPU would sometimes turn back on... But I'd still need to hard reboot the entire PC to get things working back again...

 

At first I checked the temps but they seemed fine... CPU max temp was about 70°C and GPU max temp was about 60°C . So I immediately thought it might be a PSU issue... Downloaded OCCT and started a Power Stress Test which went on for 4 minutes before having the issue happen and the GPU turn off in the exact same manner... However... I'm used to PSU failures causing a complete system shutdown and not just the GPU turning off so before I go off and buy an RM1000x I just wanted to check with you guys if there wasn't any other issue that could be happening.

 

Note I have the following BIOS Settings:

- EXPO 1 turned on for RAM

What do the logs show? Open even viewer and see what is causing the gpu to crap out.

[johnno124]

Is the PCi-E/12vHWP cable fully seated? The 2 previous ASUS cards i have had, it should be a white light for power, red indicated that the PCIe cable was not connected/seated properly

 

[Ythia]

10 hours ago, Blue4130 said:

What do the logs show? Open even viewer and see what is causing the gpu to crap out.

I looked through the Event Viewer and there was nothing that seemed related to GPU. Also when reproducing the crash with OCCT it didn't pick up on any errors either

[Ythia]

8 hours ago, johnno124 said:

Is the PCi-E/12vHWP cable fully seated? The 2 previous ASUS cards i have had, it should be a white light for power, red indicated that the PCIe cable was not connected/seated properly

 

I've tried re slotting the 12vHWP cable as well as reducing the bend the cable had. I also had some new driver updates that I installed, Currently running some stress tests... If it crashes again I'll swap out the PSU Cables and use a different socket on the PSU just in case.

[Ythia]

Alright,

Now that I'm confident on the stability and fix it seems like the issue was the 12vHWP cable that was too bent on the GPU side... I unplugged it and replugged it making sure that the cable had as much length as it needed to have a soft bend in it and the GPU has been stable ever since.

Thanks everyone for the help!