[SOLVED] Ryzen 2700x Infrequent overheating

Aug 13, 2020
3
0
10
Computer Type: Desktop Custom build

GPU: RTX 2070 Super stock

CPU: AMD Ryzen 7 2700X Processor with Wraith Prism LED Cooler

Motherboard: Crosshair hero 7 wifi

RAM: Corsair Vengeance LPX 16GB (2x8GB) DDR4 DRAM 3000MHz

PSU: 750w

Case: Thermaltake Armor+MX ATX Mid Tower. Fans: Noctua 120mm 1700rpm (rear exhaust), PSU fan (rear exhaust), Noctua 200mm 1700rpm (side intake, blows on top of CPU), 2 smaller intake fans on front of case.

Operating System & Version: Win10 1909

GPU Drivers: Nvidia 451.67

Chipset Drivers: 5.12.0.38 (I think? Not 100% sure how to verify this)

Background Applications: Steam, origin, discord, geforce experience, HWinfo64

Description of Original Problem: Random/infrequent CPU overheating (~90c)

Troubleshooting: Adjusted CPU voltage offset in bios -0.1v. Set more aggressive CPU fan curve. Cleaned case of dust (I do this about once a month anyways).



A bit more background. Idle temps ~35c. This has happened twice. First time was about a month ago when I was playing No Man's Sky. I had been playing about 45 mins when I afk'd for a few mins. When I returned, it was about 90c and throttling. During that play session, temps ranged from about 40-60c, with average temp of 43c. I shut down, booted back up and didn't see the issue again for about a month. During that month I continued to play plenty of games (Control, Apex Legends, VR games, No Mans Sky etc etc) without any overheating.

Today was the second occurrence. Playing No Mans Sky for about 30 mins and had alt-tabbed, went afk and when I returned, it was about 85c and throttling. Shut down right away and here I am now. It has been about 2 hours since reboot, and I have sat with both Apex and/or No Mans Sky running without overheating. Max temp is 60.5c, average temp is 43.7c.

Any help on troubleshooting this would be greatly appreciated!
 
Solution
Computer Type: Desktop Custom build

GPU: RTX 2070 Super stock

CPU: AMD Ryzen 7 2700X Processor with Wraith Prism LED Cooler

Motherboard: Crosshair hero 7 wifi

RAM: Corsair Vengeance LPX 16GB (2x8GB) DDR4 DRAM 3000MHz

PSU: 750w

Case: Thermaltake Armor+MX ATX Mid Tower. Fans: Noctua 120mm 1700rpm (rear exhaust), PSU fan (rear exhaust), Noctua 200mm 1700rpm (side intake, blows on top of CPU), 2 smaller intake fans on front of case.

Operating System & Version: Win10 1909

GPU Drivers: Nvidia 451.67

Chipset Drivers: 5.12.0.38 (I think? Not 100% sure how to verify this)

Background Applications: Steam, origin, discord, geforce experience, HWinfo64

Description of Original Problem: Random/infrequent CPU overheating (~90c)...
Aug 13, 2020
3
0
10
Another option is a higher end heatsink. Such as a Scythe Mugen 5 Rev B or Thermalright Macho Direct. They offer exceptional cooling performance for around $50.

The AMD Wraith are nice factory heatsinks but don't compete with heavier duty solutions.

That's an option. I'm hesitant to blame the heatsink setup though simply because it will go a month without any overheating while being used under heavy load. I'm not an expert, but to me that feels bios related, like PBO has a bug that is overvolting bigtime or something like that.
 
Aug 13, 2020
3
0
10
So I've been trying to figure out how to do exactly that. I have yet to be successful in even disabling PBO, which is annoying lol. Also, my voltage offsets don't seem to be taking effect. For example, I have a -0.5v offset currently and I'm still seeing 1.5v when PBO kicks in. Do I have the wrong idea of what offsetting is?

Also, what are 1 2 4 &c tests?
 
Computer Type: Desktop Custom build

GPU: RTX 2070 Super stock

CPU: AMD Ryzen 7 2700X Processor with Wraith Prism LED Cooler

Motherboard: Crosshair hero 7 wifi

RAM: Corsair Vengeance LPX 16GB (2x8GB) DDR4 DRAM 3000MHz

PSU: 750w

Case: Thermaltake Armor+MX ATX Mid Tower. Fans: Noctua 120mm 1700rpm (rear exhaust), PSU fan (rear exhaust), Noctua 200mm 1700rpm (side intake, blows on top of CPU), 2 smaller intake fans on front of case.

Operating System & Version: Win10 1909

GPU Drivers: Nvidia 451.67

Chipset Drivers: 5.12.0.38 (I think? Not 100% sure how to verify this)

Background Applications: Steam, origin, discord, geforce experience, HWinfo64

Description of Original Problem: Random/infrequent CPU overheating (~90c)

Troubleshooting: Adjusted CPU voltage offset in bios -0.1v. Set more aggressive CPU fan curve. Cleaned case of dust (I do this about once a month anyways).



A bit more background. Idle temps ~35c. This has happened twice. First time was about a month ago when I was playing No Man's Sky. I had been playing about 45 mins when I afk'd for a few mins. When I returned, it was about 90c and throttling. During that play session, temps ranged from about 40-60c, with average temp of 43c. I shut down, booted back up and didn't see the issue again for about a month. During that month I continued to play plenty of games (Control, Apex Legends, VR games, No Mans Sky etc etc) without any overheating.

Today was the second occurrence. Playing No Mans Sky for about 30 mins and had alt-tabbed, went afk and when I returned, it was about 85c and throttling. Shut down right away and here I am now. It has been about 2 hours since reboot, and I have sat with both Apex and/or No Mans Sky running without overheating. Max temp is 60.5c, average temp is 43.7c.

Any help on troubleshooting this would be greatly appreciated!


This has nothing to do with your PC or hardware.

The issue is that the game is still running its 3D game engine even in the lobby...and it's running the system maxed out never giving the GPU or CPU a rest like they get in game as things vary when you're moving around. You can verify this in Task Manager by looking at the performance tab and you'll see the cpu and gpu still hammering away even when you've tabbed out.

I've had the same issue in other recent games like Ghost Recon...some recent games like COD have started to address the issue by allowing you to set a FPS cap for the lobby which I set to 30fps. Problem solved.

The simple solution, if the game wont allow you to set an fps cap for the lobby, is if you need to go afk for more than a couple minutes just exit the game.
 
Solution
So I've been trying to figure out how to do exactly that. I have yet to be successful in even disabling PBO, which is annoying lol. Also, my voltage offsets don't seem to be taking effect. For example, I have a -0.5v offset currently and I'm still seeing 1.5v when PBO kicks in. Do I have the wrong idea of what offsetting is?

Also, what are 1 2 4 &c tests?


PBO is going to push the voltage up regardless of the offset...if you want the offset to work to reduce temps you'll have to figure out how to disable PBO in the bios. IMO instead of using voltage offsets to control temp (and risk destabilizing your system) it's better to just improve your CPU cooling and let the chip do it's thing...otherwise you should've just bought a 2600.