RAM fail

[Sharon]

Hi Teams

Could you help me confirm this is CPU error or hardware design error?
https://drive.google.com/drive/folde...usp=drive_link
Thanks.

Sharon

[David (PassMark)]

There are two errors.

The video RAM failure is bad, but still might not be noticed by an end user most of the time.
The main RAM failure is very bad. Likely the machine will be unstable, have strange crashes and misbehavior.

This is one would hope not a design error. As every PC of this type would fail in the same way.

Errors were

LOG NOTE: 2026-01-16 00:39:14, 2D Graphics, GPU Intel(R) Graphics, Verification failure of GPU buffer contents (Offset: 187621848, Read: 0x0054000000000000, Expected: 0x0000000000000000)

SERIOUS: 2026-01-16 00:58:38, RAM, Error verifying data in RAM

LOG NOTE: 2026-01-16 00:58:38, Memory (RAM), Process 2, Data verify error: virtual address 0x00000000A8EF9668, read 0xafffffffffffffff, expected 0xffffffffffffffff

LOG NOTE: 2026-01-16 00:58:38, Memory (RAM), Process 2, Current pattern: 64-bit Ones (11111111), bits in error 0x5000000000000000

[Sharon]

Thanks for your explaination.

From our verification results:

• IRMT test: PASS
• MemTest86: PASS
• Only the Burn-in test reports RAM test failure

Intel RMT (Rank Margin Tool)

Since IRMT and MemTest86 both pass, we would like to better understand why the failure is only observable during the Burn-in test.
Could you please help provide some insight into the differences between these tests? Thank you for your support.

[David (PassMark)]

There are various possibilities

1. It's somewhat random. And retesting with MemTest86 might show a fault
2. It is load related in some way. This might mean
   1. Temperatures are higher in Windows, compared to BIOS (MemTest86), especially if GPU is in heavy use.
   2. The EMI is higher with more hardware in use
   3. Power / voltage levels less stable with more hardware in use
3. There is an unexpected interaction between the iGPU and CPU (hardware fault or bug). e.g. electrical corruption of the data bus when both are in use.
4. There is a memory allocate bug in video device driver. For example the same block of RAM is allocated both to main RAM and iGPU video RAM. This would lead to RAM corruption on both sides. This is a software bug and not a hardware fault.
5. Similar type of bug but in some other random kernel level device driver. You can have bugs that overwrite the memory of other processes. Causing corruption.
6. When in Windows you have some overclocking or power control software that isn't running in BIOS. So CPU / RAM might be running at higher or lower clock speeds compared to when in BIOS. Switching between speeds might also be more common.

None of this is easy to debug however. I'd swap out the RAM and then CPU to start with. It might be a one off hardware fault.