MEMTEST hung-up(The screen has frozen unexpectedly)

[Chris Chang]

Hi,

We conducted tests using MemTest86 11.3 and encountered an issue where the screen display froze during testing. The log file shows the following details:


2025-05-30 19:10:00 - Running test #9 (Test 9 [Modulo 20, random pattern])
2025-05-30 19:10:00 - MtSupportRunAllTests - Setting random seed to 0x32BED6E1
2025-05-30 19:10:00 - MtSupportRunAllTests - Start time: 11448349 ms
2025-05-30 19:10:00 - Start memory range test (0x0 - 0x1080000000)

This issue did not occur when running MemTest86 11.2.
Attached is the log file for your reference.
Could you help investigate the possible cause or have the suggestion?

Thank you.

Chris

[David (PassMark)]

Was it random?
So if you run V11.3 again, does it stop at the same place (the start of test 9)?
Can you also post the log from V11.2 to compare, if you have it.

Other things of note from the log

CPU and RAM seems to be running extremely hot.
2025-05-30 17:23:04 - Current CPU temperature: 100C
2025-05-30 17:22:03 - DIMM1 temperature: 97.250C
While it varies a bit, thermal limit on RAM is typically more around 85C. Which you are well above.

There was some multi-threading warnings
2025-05-30 17:26:37 - WARNING - waited for 10s for CPU #12 to finish (BSP test time = 2077ms)
2025-05-30 17:26:37 - WARNING - waited for 10s for CPU #15 to finish (BSP test time = 2077ms)
There is some chance this is a BIOS bug, but could equally be a hardware fault when under load and high temperatures.

[Chris Chang]

The issue occurs randomly, but whenever it does happen, it is always during test #9.
Attached is the MemTest86 11.2 log file for your reference.
Thank you.

[David (PassMark)]

I don't think there was any change between 11.2 and 11.3 that would randomly cause a problem in test 9. If it runs without error some of the time, you might have got lucky with the 11.2 run.

Can you improve the cooling and drop the RAM temperatures by maybe 30C and then see if it is stable?