Processors

[AmdPengu]

Dear community,

Two weeks ago I decided to build a secondary system and I have encountered an issue which is wrecking my brain. The issue is that the system out of no where reboots without me seeing a BSOD screen. Event viewer says its a kernel 41 automatic reboot issue which does not help me much further. These reboots happen during idle, general pc usage like browsing (youtube, twitch, forums) and also during gaming sessions at random. The funny thing is that these reboots DON'T happen when I am stress testing my system... I have experience with building my own systems and maintaining them. But I have no concrete idea why this is happening. I have suspicions it might have something to do with how my board handles the volts but i am not sure how to verify this. Hopefully someone could assist me in solving this nerve wrecking issue. Actions i have taken in hopes of resolving the issue:

Bios updates/downgrades:
- F14 AGESA 1.2.0.5
- F15 AGESA 1.2.0.7
- F16c (CURRENT) AGESA V2 1.2.0.8

- Two formats, one initial when building the PC and one after issues started happening.
- Several stress tests consisting of:
 - Prime95 small FFT (4 hours)
 - Karhu Ramtest (2 hours)
 - TM5 test with extreme@anta777 config (3 hours)
 - Cinebench r23 multi/single several runs (must been 10 or more at this point)
 - Aida PSU stress test enabling CPU, FPU, GPU & RAM (2 hours)

Used XMP profile @ 3600mhz, 1800FLCK, 1.35v
Used stock non XMP profile auto FLCK, auto volt
Currently using XMP profile @ 3200mhz, 1600FLCK, 1.35v
None of the times during the stresstests I encountered errors nor did the PC crash/reboot.

I also moved several parts from my main stable system into this one to see if it was potential hardware issues. Even after putting in several parts the issue still occurred in the same manner. The secondary system's specs are originally.
- W10 x64 fully updated
- Chipset drivers: updated via AMD site
- Ryzen 5600x
- B550 Aorus pro (v1)
- Asus Gtx 1070 (Put in my EVGA 3080 ftw3 from my main system to test - uninstalled drivers in safemode with DDU)
- Samsung 980 1tb nvme
- Corsair Vengeance LPX 3600mhz CL16 2x16gb (Put in my Ballistix 3600mhz CL16 4x8gb from my main system to test)
- Corsair VS series 550watt 80+ PSU (Ordered new: Corsair RMX 750watt)

[elstaci]

Once you replace the PSU with the new 750 Watt PSU report back if it continues to shutdown by itself.

In the past there was a setting for the PSU that would cause the computer to shut down when idle: Power Control Idle Control.

Normally when a computer shuts down by itself is due to:

1- Overheating issues (CPU, GPU or PSU)

2- Overclocking issues (CPU or GPU)

3 - Power issues (PSU)

4- Defective Hardware (Failing or intermittent)

5- Incompatible Hardware

6- Sometimes Driver issues.

Those are some of the main reasons that come to my mind at the moment.

[AmdPengu]

Hello thank you for replying. The odd thing is that during stress tests it never happens. Only when I browse, watch youtube, watch streams or go from idle to opening applications.

1) CPU does not get hotter than 68-70c degrees even under stressing, GPU (3080 currently) does not go over 75c. PSU (750watt) i assume isn't overheating as well should have enough air intake. The tower is 3-4cm above tiled floor.

2) CPU and GPU are running stock

3) Current 750watt is new and ive ran multiple stress tests at the same time to test power draw and has not shown signs of dips in 3,3v, 12v and or 5v line according to hwinfo64

4) Only part i haven't switched out are the 5600x and the motherboard. I am planning to put my 5800x3d in from my main rig to see if the issue still persists.

5) According to this website: https://pangoly.com/en/compatibility/gigabyte-b550-aorus-pro/ram
my ramkit: ' CMK32GX4M2D3600C16' is compatible

6) Currently running latest bios f16c AGESA V2 1.2.0.8, GPU drivers, latest onboard sound drivers and latest chipset drivers. Also windows 10 x64 is fully updated and registered.

[elstaci]

It doesn't sound like hardware issues since everything seems to work normally except that it intermittently shuts down by itself.

Unless you have a intermittent hardware failure. That is the hardware works fine but then fails for some reason causing the PC to shut down.

Like a intermittent CPU or Motherboard or PSU or GPU for instance

I would believe if you had incompatible hardware it would happen a lot more frequently or all the time especially under Stress tests.

To rule out a Driver issue, enable Microsoft Verifier and see if by chance it shows a driver as the issue when it shuts down: https://learn.microsoft.com/en-us/windows-hardware/drivers/devtest/driver-verifier

Here is Gigabyte Support download page for your Motherboard: https://www.gigabyte.com/Motherboard/B550-AORUS-PRO-rev-10/support#support-dl

I don't see your Corsair RAM Number listed under the "VERMEER" Processors that covers the Ryzen 5800X. Here are all the 3600Mhz Corsair RAM listed:

Of course this list might be out-dated and your RAM hasn't be tested yet by Gigabyte or it was found your RAM was incompatible.

I tried to check at Corsair but your Motherboard isn't listed so I couldn't check for compatibility with your motherboard.

To rule out bad RAM run MEMTEST86 or Windows Memory Diagnostic to check your ram.

to rule incompatibility just install one RAM Stick of 16GB in DIMM Slot# 2 and see if the computer continues to shut down after running the RAM Diagnostic test.

Also run CHKDSK C: /f/r/x to check your Windows drive for errors.

Run SFC /scannow to check your Windows Installation.

Also do a CLEAR CMOS to put your BIOS back to factory Default in case it is a BIOS setting your changed that is causing the issue.

I would open a AMD Support ticket under Processors and  Warranty to see if they believe you need to RMA your processor or not: https://www.amd.com/en/support/contact-email-form

I would also open a Gigabyte Support Warranty ticket for your Motherboard to see if they feel your Motherboard needs to be RMAed to be checked or repaired or replaced.

You might want to open a Corsair Support Ticket to see if your RAM is compatible with your Gigabyte Motherboard or not.

NOTE: You have a very tough troubleshooting issue since everything works fine under various Stress tests except that it intermittently shuts down.

Thus you need to start methodically eliminating hardware and software issues by running Hardware diagnostic tests as an example.

[AmdPenguTwo]

Thank you very much for taking time in writing such comprehensive reply. I tried to reply to your first one but the website prevented me in doing so due to too many replies in an X amount of time.

regarding your first reply:
Power Control Idle Control has three options in my bios and currently has Auto selected. Should i try a different option?
1) Typical current idle
2) Low current idle
3) Auto

Regarding your second reply:
To rule out bad RAM run MEMTEST86 or Windows Memory Diagnostic to check your ram.
Memtest86 is a program which I recall using many years ago. I recall using it while booting it from an usb through bios.

- Do you still recommend this method and how long do you recommend me running memtest86 for? Ever since those days I moved on to Karhu's Ramtest and more recently TM5 ramtest with the 'extreme anta777 config'.
- I assume i should test 1 dimm in A2 slot at one time. If this is the case what ram configuration would you recommend me testing the ram at?
  * Everything default/auto?
  * at max compatible speed for the CPU which would be: 3200mhz @ 16-19-19-36?
  * at 2133mhz @ 15-15-15-36 as advertised on?:
https://www.corsair.com/us/en/p/memory/cmk32gx4m2d3600c16/vengeancea-lpx-32gb-2-x-16gb-ddr4-dram-360...
 
I shall follow your advice in running windows drive for errors, checking windows installation for errors and contacting three manufacturers (AMD, Gigabyte & Corsair). Again thank you very much for taking time and replying so thoroughly. I shall try to update the thread when I get new info and hopefully the thread can end on a positive note.

[elstaci]

I would change the BIOS setting to "Typical Current Idle" just for troubleshooting purposes. Some PSU can't handle the "Low Current Idle" setting and shuts the computer down in the past.

The newer high quality PSU have no problems with the "Low Current Idle" setting. That is why I just said to put it on 'Typical' for testing purposes only.

Windows Diagnostic would be the easiest one to run. Enable it and restart the computer and it will automatically start the Windows Diagnostic program. MEMTEST64 is the one most popular programs that people use because it one of the best RAM testing program for free.

IF using MEMTEST test have it go through 3-5 passes to make sure. Generally if you have a bad RAM Module it would show up before the first pass finishes.

If you are satisfied with the other RAM Testing software use that since you are familiar with it. I would install all the RAM Sticks on your Motherboard and have it checked all at once.

I once had RAM issues on my PC with 4 sticks of Corsair 4GB RAM Sticks. Ran MEMTEST64 and almost immediately it was showing thousands of errors when it should have been Zero errors. Removed 2 sticks of RAM and ran MEMTEST64 again thus being able to find the bad RAM Stick.

RAM Testing just rules out bad RAM MEMORY Hardware but not if it is incompatible or not.

Once you check for bad RAM and everything checks good then just install one RAM Stick of 16GB and see if the PC continue to shut down intermittently. Sometime incompatible RAM will work with one stick but not two stick in Dual Channel configuration.

You can run the RAM at its native frequency since the computer did boot up without problems with that settings in BIOS.

All the other tests are easy to run through elevated Command Prompt or Powershell and shouldn't take very long to run except maybe Chkdsk if you have a large SSD.

Since you have a Samsung SSD you can run Samsung's own SSD Testing software rather then using WIndows CHKDSK test. It probably would be better for your SSD.

Those test is just to eliminate possible bad hardware or corrupt Windows installation just to rule those out.

Windows Verifier is to rule out any corrupt or bad or incompatible Drivers that might be causing the shut down issues.

You seemed to know a lot about computers so I let you decide which tests to run or use since you have done a lot of troubleshooting already.

Opening Support tickets might help to find out possible hardware issues.

It could be something as simple as some incompatible hardware attached to the PC or Motherboard or some loose cable on the motherboard or hardware or a Windows settings or defective cable or even an intermittent short on the motherboard to ground.

Sorry for the long replies but your situation is not easy to troubleshoot thus you need to eliminate as much as possible thus all the tests and troubleshooting procedures.

Anyways, I agree with your last statement " hopefully the thread can end on a positive note."

[AmdPengu]

Thank you again for another elaborate reply Elstaci. Until today the two components I haven't switched were the CPU and Motherboard. Mainly because the 5800x3d in my main system is behind a closed loop. However today I had time and motivation to switch out the 5600x for my main system's 5800x3d. The system has been nearly running for two hours without any hiccups. I am hopeful but skeptical at the same time.

Currently I am running a single dimm in A2 with auto settings. Soon I will put in the 2nd dimm in B2 at manufacture advertised speeds (3600mhz CL16). When all goes well i'll maybe even fool around with FLCK and timings too.

I have another motherboard and a ram set (QVL compatible) on the way. I will keep the thread updated. Hopefully I can cancel those two orders and the thread can end on a positive note :).

[elstaci]

It is possible you may have a defective 5600X processor since the 5800X3D is so far working without shutting down.

I would definitely open a AMD SUPPORT - Warranty ticket AFTER you install the full set of RAM at its native speed and if it doesn't shut down anymore.

You can try and put the 5600X in your other working PC and see if the shut down issue starts up again. If it does then you just verified you have a defective AMD processor and would start a RMA process with AMD Warranty.

Good luck. Please post back if the Shut down issues have stopped and if AMD SUPPORT replied back to you concerning your 5600X processor.

[BobC0728]

I also think you have a bad 5600x.  I had two bad 5900x chips.  Drove me crazy for months. Third time is the charm for me!  I think AMD is letting chips out of the factory that should probably end up in the scrap heap....imho.

[AmdPengu · Accepted solution]

I can pretty much confirm the 5600x is faulty. Let's see how AMD will handle the RMA process. Maybe I get lucky and receive a 5600x3d :)? Ever since I slotted in my main system's 5800x3d everything has been performing to expectations. Sad that it's a faulty 5600x but at least the problem is known now, good enough ending i'd say. Thank you again @elstaci for thinking with me.

[elstaci]

Your type of intermittent shut down issues are very difficult to troubleshoot since it could be various hardware or software (Driver) causing it.

Luckily you had two similar systems in which you can test out hardware on, rather than like most Users, need to RMA the hardware and hope it is bad.

Anyways, Hey good Troubleshooting and glad you finally was able to narrow it down to a defective CPU.

Take care.

[AmdPengu]

Hello Elstaci,

You are correct, I was very fortunate to have quick access to spare parts in order to speed up the troubleshooting process. I have since received a replacement 5600x from the retailer and can confirm it was 100% a CPU issue. The system has been running as expected with the original intended parts minus the defective CPU.

[elstaci]

In most cases, like yours, the Users would have RMAed both the Motherboard and CPU since  they wouldn't have the means to check each hardware individually.

Anyways, Good Troubleshooting, plus I will be referencing this thread in the future for others with similar issues.