SER5 5700U crashes (CPU?)

ZTHawk

I did now several tests.
1) Changed to different RAM (3 different kits in total, also tested single RAM sticks)
2) Used different SSDs
    Always crashing, and lately even more.
3) Reverted back to 507 (seemed to help very little).
4) Reseted BIOS several times.

For easier testing installed Windows11 23H2 (which also crashed during the installation several times). With some luck I got a running Windows.
Testing with Prime95 under Windows revealed some points:
1) Blend => resulted in BlueScreen within ~10-20 seconds
2) Smallest FFT (high power) resulted in runtime > 10min (stopped it manually)
3) Small FFT (maximum power) resulted in runtime > 10min (stopped it manually)
4) Large FFT (stress memory controller and RAM) resulted in runtime ~10 seconds before bluescreen

During the test "Blend":
CPU ~65-70°C / ~0.99V (~0.78 after throttling) / ~3.5GHz (~2.6GHz after throttling)
Board Temp1 70°C / Temp2 50°C
SSD ~42°C

Test "Large FFT":
CPU almost no throttling ~3.2GHz
RAM ~ 20Gbps / 12Gbps (read / write)

As I tested several RAM kits, it cannot be the RAM itself.

I managed to run in mode "Blend" >30 min with following BIOS changes:
Advanced
        ACPI Settings
                Enable ACPI Auto Configuration => Enabled
        AMD PBS
                Allocate 100KB for Mp2 customize request => Disabled
                UCSI tunnel location => UCSI tunnel at EC RAM
        AMD CBS
                UMC Common Options
                        DDR4 Common Options
                                Common RAS
                                        DRAM Adress Command Parity Retry => Disabled
                                        ECC Configuration
                                                DRAM ECC Enable => Auto
                        Memory MBIST
                                MBIST Enable => Enabled
                NBIO Common Options
                        IOMMU => Disabled
                        GFX Configuration
                                UMA Frame buffer Size => 2G or 4G
                        SMU Common Options
                                SmartShift Control
                                        Sustained PowerLimit => 25000
                                        Fast PPT Limit => 25000
                                        Slow PPT Limit => 25000
Security
        Secure Boot
                Secure Boot => Enabled
                Secure Boot Mode => Standard

I have tested some variants but the most important seem to be:
Allocate 100KB for Mp2 customize request => Disabled
UCSI tunnel location => UCSI tunnel at EC RAM
DRAM Adress Command Parity Retry => Disabled
DRAM ECC Enable => Auto
UMA Frame buffer Size => 2G
Sustained PowerLimit => 25000
Fast PPT Limit => 25000
Slow PPT Limit => 25000

Maybe you have more ideas what could be the root cause.

tuiguang10

ZTHawk replied at 2025-02-15 03:57
I did now several tests.
1) Changed to different RAM (3 different kits in total, also tested single  ...

Hello  there，

Here are some additional steps and considerations that might help you identify the root cause of the crashes:
1. Power Supply Unit (PSU)
Symptoms: Random crashes, especially under load, can sometimes be attributed to an insufficient or failing PSU.
Action: If possible, try using a different PSU with adequate wattage and see if the issue persists.

2. Motherboard
Symptoms: Since you've already tested different RAM kits and SSDs, the motherboard could be the culprit, especially if the memory controller or other critical components are failing.
Action: Inspect the motherboard for any visible damage (e.g., bulging capacitors). If you have access to another compatible motherboard, try swapping it out to see if the problem persists.

3. CPU
Symptoms: CPU issues can cause instability, especially under load. The fact that the system crashes quickly with memory-intensive tests but runs longer with CPU-intensive tests might point to a CPU issue.

4. BIOS/UEFI Firmware
Symptoms: Sometimes, BIOS/UEFI firmware bugs can cause instability.

Please press delete key as soon as you turn on the PC, so that you can go to BIOS. Please send us a picture of the Main page.
We will check if you need to update the BIOS.

5. Cooling
Symptoms: Overheating can cause crashes, especially under load.

Action: Ensure that your CPU cooler is properly seated and that thermal paste is applied correctly. Check case airflow and ensure that all fans are functioning properly.

6. Driver Issues
Symptoms: Incompatible or corrupted drivers can cause instability.

Action: Ensure that all drivers, especially chipset and GPU drivers, are up to date. You can also try running the system in Safe Mode to see if the crashes persist.

7. Windows Installation
Symptoms: A corrupted Windows installation can cause crashes.
Action: Consider performing a clean installation of Windows on a different SSD to rule out any issues with the current installation.

8. Event Viewer
Symptoms: Windows Event Viewer can provide more detailed information about what is causing the crashes.

Action: Check the Event Viewer logs (especially under "System" and "Application") for any critical errors or warnings that occur around the time of the crashes.

9. Stress Testing
Symptoms: Stress testing can help isolate the issue.

Action: Use tools like MemTest86 to test the RAM thoroughly, even though you've already tested different kits. Also, consider using OCCT or AIDA64 to stress test different components (CPU, GPU, PSU) to see if you can isolate the issue.

10. Peripheral Devices
Symptoms: Sometimes, peripheral devices can cause instability.

Action: Disconnect all non-essential peripherals (e.g., USB devices, external drives) and see if the system stabilizes. If it does, reconnect them one by one to identify the problematic device.

11. Voltage and Clock Settings
Symptoms: Incorrect voltage or clock settings can cause instability.

Action: Ensure that all voltage and clock settings in the BIOS are set to their default values. Avoid overclocking until the system is stable.

12. Check for Short Circuits
Symptoms: A short circuit somewhere in the system can cause instability.

Action: Inspect the case and motherboard for any loose screws or metal contacts that might be causing a short. Ensure that the motherboard is properly mounted and not touching the case inappropriately.

13. Check for BIOS/UEFI Settings
Symptoms: Certain BIOS/UEFI settings can cause instability.

Action: Reset the BIOS/UEFI to default settings. Please repeatedly press delete key as soon as you turn on the PC, go to BIOS. On Save&Exit page, please click Restore defaults, press enter to confirm, then press F4 to save.

14. Check for Firmware Updates
Symptoms: Firmware issues can cause instability.
Action: Check for firmware updates for your SSD, GPU, and other components.

15. Check for Windows Updates
Symptoms: Sometimes, Windows updates can resolve underlying issues.
Action: Ensure that your Windows installation is fully up to date.

If none of these steps resolve the issue, it might be worth considering professional diagnostics, especially if the problem lies with the motherboard or CPU, which are harder to test without spare parts.


Have a nice day!

ZTHawk

4. BIOS is as already said 507. But tested 508, 512 too. (In a previous post I have send an image already, it was 508 back then)
5. Cooling is fine (temps are at ~70° under load). I cannot test the thermal paste (but as the temps are stable, I do not think it is an issue)
6. N/A as also happened during windows installation and Unix systems (windows is just easier to test for me)
7. N/A as this stick is creating correct installations. The crashes were during CPU intensive action (FAN was spinning up)
8. Nothing that points to a
9. Memtest86 was fine, but I also used different RAM kits, so it is deffinetly not the RAM (other systems are running fine with tested RAM)
10. N/A because crashes occur also during installation and other systems are fine with them.
11. Can you name some of those settings?
12. Everything looks fine
13. As already written I have tested defaults several times. Not working. Only with mentioned changes to BIOS (see previous post) I achieved a "stable" system for now
14. N/A as tested with several SSDs. No other components are installed (it is a SER5 5700U)
15. N/A as crashed during installation and Unix systems

Remaining points:
1. I have no other PSU available that matches this one
2. Have not seens anything suspisious. If I had another mainboard then it would also contain a new CPU. The Mainboard is basically the Mini PC itself.
3. This would mean to replace the complete Mini PC (like point 2)

I did another test, and it seems that this BIOS setting is the most important one:
Advanced => AMD CBS => UMC Common Options => Memory MBIST => MBIST Enable => Enabled

Having everything set to DEFAULT in BIOS and only enabling MBIST, did make the system much more stable. I could not intentionally crash it yet. Though I received a crash when I closed prime95. (It did not crash when I tried to repeat it).

ZTHawk

Another observation today:
1) PC is off (for some time)
2) Turn on and prime95 results in bluescreen ~10-20s
3) After reboot prime95 crashed (windows not)
4) Started prime95 again and windows crashed (weired looking screen, not bluescreen)
5) After reboot prime95 is running fine

Seen this now several times (step 3 and/or 4 did not always occurred).

Behaviour is like an engine that needs to get warm first.

tuiguang10

ZTHawk replied at 2025-02-16 18:12
Another observation today:
1) PC is off (for some time)
2) Turn on and prime95 results in bluescreen ...

Hello there,

According to your description, it seems that your machine's motherboard is broken.

It is suggested that you can contact this after-sales mailbox to send the machine back for repair: support-pc@bee-link.com

Please send an email with the SN picture and purchase history of your machine, explaining your machine problem and our repair suggestions.

Our staff will provide you with the corresponding after-sales address.

Have a nice day!