PC Processors

[rabidsmurf]

A few days ago I replaced the motherboard/CPU/RAM and added an additional M.2 SSD in an otherwise completely stable system. I have no reason to suspect the issue is caused by GPU/PSU as these were rock solid for several years until the above changes.

Specs:

AORUS x570 Master (BIOS F32) - New - Settings are default except for XMP profile.
Ryzen 7 5800x - New
eVGA GTX1080ti FE (now 4 years old, never a single problem)
Corsair 2x16gb 3600 (XMP): CMW32GX4M2Z36000C18 - New
Power Supply: Corsair ax850i (2-3 years old, never a single problem).
HP ex920 1TB M.2 SSD - Boot drive, about a year old, no SMART errors according to disk checkup.
WD SN850 Pci-e 4 2TB M.2 SSD - New,  used for games library, no SMART errors according to disk checkup.
Windows 10 Pro - 20H2 - Fresh install, all updates applied.

I am getting random lockups of the operating system, usually during light workloads (browsing/email/youtube, etc). Task Manager can be opened but values are not updating, explorer process cannot be restarted, forced to hold power to restart. I can play games like POE/Witcher 3/Cyberpunk seemingly for hours with no issues.

Upon restart the only indication of a problem I can find is a "Kernel Power event ID 41" in the system event log.

I initially had the incorrect RAM (optimized for intel), I exchanged at the store for the Ryzen optimized RAM - No Change.
I initially flashed f33j (AGESA 1.2.*) beta bios onto motherboard, no realizing it was a beta BIOS - Crash/freeze happened every few hours, sometimes within minutes of each other
I initially had AMD Chipset drivers from Gigabytes website which were a few months older than listed on AMD's website - Updated to the newer ones.
I re-flashed to F32 last stable full release bios (AGESA 1.1.0.0) - Crash/freeze seems less frequent, I actually went 32~ hours uptime with no issues, varying workloads/gaming and idle time, but today it happened again.

I do not have HWINFO installed, although I do use CPUID HWMonitor and CPU-Z for general monitoring. The last crash happened within 30 minutes of opening HWMonitor, and I am reasonably sure it's been open for every crash. I am suspicious of this, but at the same time I've used this tool for years so I could just be grasping at straws.

I have been leaning towards AGESA/BIOS stability issues, and returning the motherboard to try an ASUS board instead, however after reading these forums I am now wondering if it's the processor, I have seen only one WHEA-Logger cache hierarchy error in the logs. Most of the time it is only the generic Kernel power error.

I have 10 days to return the motherboard to the store, the processor was final sale so RMA only. I am looking for ideas or other things to check before I take these measures.

[mo44]

Seen quite a few people in various forums complaining about HWMonitor/CPU-Z/CAM causing Windows to freeze but that is still pretty anecdotal, I've been having the same issue myself and it also seems dubious to me but perhaps these applications are just triggers for some problem in a low-level driver or BIOS.

Would love to see this addressed but it's hard to say if the issue is with AMD, CPUID, or Microsoft.

[cyberloner]

disable turbo boost and try stability..........

if stable.... wait new bios that drop the 4850boost to 4800boost.......

my aorus now boost at 4800mhz........ all motherboard push too much... maybe they use golden sample chip to build the bios

[layer8]

I have the same issue and I have a very similar setup except:

- that I am running Linux

- Radeon RX 5700 XT 

- some other brands (not important I think)

- except for the GPU and SSDs, all brand new components

So I doubt it is a Windows problem.

Doesn't happen while gaming but on usual desktop workloads, browsing or downloading games. Didn't have any problem before migrating the new PC.

A friend has the same mainboard and no issues I guess (didn't ask but didn't complain). I think it is a firmware problem because afaik he did not update BIOS nor did he change anything in BIOS. I did update to F34 and changed some settings. Maybe I downgrade again.

[layer8]

I managed to solve it.

tl;dr : I resettet BIOS to default settings (maybe not important, read further why) and downgraded to F33.

First of all I checked if the SSDs are still ok. They are only 2 years old so they were ok as expected.

Then updated BIOS F34 again, to make sure nothing went wrong in the first place.

Still freezes and lags.

Then I resetted all BIOS settings to default.

Still freezes and lags.

Then I downgraded to F33.

No more freezes and lags! I will restore my preferred settings step by step the next days and see if I notice any problems.

Why I didn't update to F35e? Well, afaik a letter suffix means it is not stable yet and I don't want to be gigabytes beta tester. They already caused enough problems for me...

By the way, I have revision 1.2 for the case it its important.

[DRDOS]

A friend of mine experienced exactly the same issue on a similar system setup recently, with the latest BIOS version.
The only effective fix we found was a disabled factory overclocking option (Turbo Boost or whatever that temporarily increases a CPU clock rate to over 4 GHz).

I suspected probably a motherboard's Vcore DC-DC converter had not enough of extra power capability to maintain a stable voltage in an overclocked mode for 95 watts 8-core CPU. So maybe a load-line calibration option could fix this issue then, but we didn't try this out yet.