Question PC shuts down occasionally under load (I already changed the PSU) ?

[elux1337]

Hey everyone, I’m having an issue with my PC where it shuts down completly. It's like a short circuit.

It‘s occasionally when under load when playing Flight Simulator, when tabbing out or when loading into a game, but not consistently. Temperatures are normal.

First it was occasionally, then it got consistent when stress testing. But after updating my BIOS, and it was more stable but now its still crashing sometimes. I swapped the PSU already.

After it crashed i have to turn it on again at the on switch at my case. Wake-On-Lan doesn't work then.

System specs:

* CPU: Ryzen 7800X3D
* GPU: Gigabyte RTX 4080
* PSU: BeQuiet 850W (recently replaced with a new one, issue persists)
* Motherboard: MSI B650 Tomahawk WiFi
* RAM: G.Skill DDR5 6000MT/s

What I’ve tried so far:

1. Replaced the PSU (problem persists).
2. Updated the BIOS (slightly improvement).
3. stress tests work perfectly fine. (OCCT, Prime95, FurMark)
4. Memtests
5. Event Viewer (is empty)
6. Rebuild PC
7. Fresh installed Windows
8. Updated drivers

What’s the most likely hardware fault to cause this issue?

 

[Lutfij]

BeQuiet 850W (recently replaced with a new one, issue persists)
Be Quiet! is the brand of the unit while 840W is the advertised wattage of the unit. What is the model of the unit?

Motherboard: MSI B650 Tomahawk WiFi
+
Updated the BIOS (slightly improvement).
For the sake of relevance, what BIOS version are you on for your motherboard?

RAM: G.Skill DDR5 6000MT/s
Got a link to this ram kit?

CPU: Ryzen 7800X3D
How are you cooling the processor?

You forgot to mention the make and model of your case, the number of fans and their orientation in said case.

Fresh installed Windows
Windows 10 or 11? Did you recreate the bootable USB installer to rule out a corrupt installer? Did you install the OS in offline mode, to later install all drivers in an elevated command?

 

[elux1337]

BeQuiet 850W (recently replaced with a new one, issue persists)
Be Quiet! is the brand of the unit while 840W is the advertised wattage of the unit. What is the model of the unit?

Motherboard: MSI B650 Tomahawk WiFi
+
Updated the BIOS (slightly improvement).
For the sake of relevance, what BIOS version are you on for your motherboard?

RAM: G.Skill DDR5 6000MT/s
Got a link to this ram kit?

CPU: Ryzen 7800X3D
How are you cooling the processor?

You forgot to mention the make and model of your case, the number of fans and their orientation in said case.

Fresh installed Windows
Windows 10 or 11? Did you recreate the bootable USB installer to rule out a corrupt installer? Did you install the OS in offline mode, to later install all drivers in an elevated command?

1. PSU: 850 Watt be quiet! Straight Power 11 Modular 80+ Gold

2. BIOS Version: 7D75v1L (Agesa Combo 1.2.0.2b)

3. RAM Kit: 32GB (2x 16GB) G.Skill Trident Z5 NEO EXPO black DDR5-6000 DIMM CL30-38-38-96 Dual Kit

4. Cooled by a BeQuiet Pure Rock 2

5. Case is a NZXT H510 with two NZXT 120mm Fans in and the same 120mm Fans out (one at the back and one on top)

6. Windows 11 installed via Boot Stick

 

[manytools]

Have you tried uninstalling the gpu driver and try to play without installing manually?

 

[elux1337]

Have you tried uninstalling the gpu driver and try to play without installing manually?

Yes. I uninstalled it with DDU.

 

[manytools]

Have you used insulation gaskets when building the pc?

 

[elux1337]

Have you used insulation gaskets when building the pc?

If you mean the rings at the screws, no. But there's insulation at the Mainboard itself. But I don't think that's causing the problem, since it's only crashing under load. My guess is either faulty GPU, Mainboard, CPU or RAM but i'm not sure which of them is the most likely to cause these random shutdowns under load.

 

[manytools]

My thought was may the heat causing mobo or other component to thermal expansion. But if thats not the case, crash or mouse stutters that caused by alt tabbing while under heavy load is mainly because caching or fast recall from ram problem. However system failure is a complete different thing here. So again if any contact causing short circuit is not the case, it may even be VRM's that are overheating or faulty at the first place. Is any overclock have done other than Expo/Xmp ?

 

[SkyNetRising]

my PC where it shuts down completly. It's like a short circuit.
What’s the most likely hardware fault to cause this issue?

This typically happens with cpu overheating or PSU protections kicking in.
You said temperatures are normal. What exactly temperatures are you experiencing on cpu and gpu?
Most likely PSU can't handle transient power spikes from GPU.

Adjust power limit for your graphics card with MSI Afterburner.
Lower power limit by 10%.

 

[elux1337]

My thought was may the heat causing mobo or other component to thermal expansion. But if thats not the case, crash or mouse stutters that caused by alt tabbing while under heavy load is mainly because caching or fast recall from ram problem. However system failure is a complete different thing here. So again if any contact causing short circuit is not the case, it may even be VRM's that are overheating or faulty at the first place. Is any overclock have done other than Expo/Xmp ?

Thanks for the reply. I have nothing overclocked.

 

[elux1337]

This typically happens with cpu overheating or PSU protections kicking in.
You said temperatures are normal. What exactly temperatures are you experiencing on cpu and gpu?
Most likely PSU can't handle transient power spikes from GPU.

Adjust power limit for your graphics card with MSI Afterburner.
Lower power limit by 10%.

Thanks for the suggestions.
Since i changed the PSU and it doesn‘t shutdown when stress testing i don‘t think its a overheating issue or caused by the PSU. I used a different model of the PSU.

Temps under stress is CPU 89, GPU at 70-80 (hotspot at 90)

 

[SkyNetRising]

Temps under stress is CPU 89, GPU at 70-80 (hotspot at 90)

That is very high actually.
89C is max operating temperature for 7800X3D.
You're probably getting thermal shutdown.
Get a better cpu cooler.

 

[elux1337]

That is very high actually.
89C is max operating temperature for 7800X3D.
You're probably getting thermal shutdown.
Get a better cpu cooler.

I'm pretty sure it's not since stress testing doesn't shut it down. While gaming (where it crashes) it's in the 70's or low 80's.

 

[Dashhh]

I'm pretty sure it's not since stress testing doesn't shut it down. While gaming (where it crashes) it's in the 70's or low 80's.

What CPU cooler are you using? From the information provided, this does sound like a thermal issue. How's your case airflow? Any negative pressure? You might not be crashing during your stress testing, but note that thermal shutdowns are a triggered event. Your CPU might not be overheating during stress testing, but carrying some of that heat into gaming where you have a specific spike that your components deem unsafe can trigger shutdowns.

Unrelated, but I just had a user at work cross two bent CPU pins and I found it shorted a memory bank (they mentioned they didn't trust their CPU install and were nervous as it was their first build). There wasn't any logged shutdown events and the PC would just reboot. At first, I was convinced it was due to cooling because the temperatures I was getting throughout the system were horrible. I spent a wicked amount of time fixing their cooling, never thinking to check the actual CPU socket where I found the bent pins. Sometimes our issues are exactly as they appear.

 

[elux1337]

What CPU cooler are you using? From the information provided, this does sound like a thermal issue. How's your case airflow? Any negative pressure? You might not be crashing during your stress testing, but note that thermal shutdowns are a triggered event. Your CPU might not be overheating during stress testing, but carrying some of that heat into gaming where you have a specific spike that your components deem unsafe can trigger shutdowns.

Unrelated, but I just had a user at work cross two bent CPU pins and I found it shorted a memory bank (they mentioned they didn't trust their CPU install and were nervous as it was their first build). There wasn't any logged shutdown events and the PC would just reboot. At first, I was convinced it was due to cooling because the temperatures I was getting throughout the system were horrible. I spent a wicked amount of time fixing their cooling, never thinking to check the actual CPU socket where I found the bent pins. Sometimes our issues are exactly as they appear.

Thanks for the suggestions!
BeQuiet Pure Rock 2 is my cooler. I repasted and reseated it 2 times already, and the cooler should be enough. (150 TDP max)
I also don‘t have any thermal throttling.
Im close to 100% sure it‘s hardware related.
I just ordered a new motherboard and see if that helps.

Is there any log if the CPU shuts down due to thermal reasons?