Question GPD P2 Max 8100Y died in a weird way - could someone help diagnose it?

Aug 28, 2022
1
0
10
0
Hello,

I resorted to writing about it here because I am getting desperate - my laptop died and nobody would repair it, even if I want to pay. This will be a long post so bear with me please.

I bought a GPD P2 Max in 2019. It worked mostly OK for ~2.5 years but then it seemed to have failed catastrophically (during a windows update - but it does not matter that much when, as the problem seems to be with hardware).

After this failure, the only thing that was happening after turning it on was: the fan spun up for about 3 seconds, then turned off. And it kept spinning up and down ad infinitum until I pressed the power button. I tried to talk to GPD tech support directly (in Hong Kong); they would only respond meaningfully on Aliexpress. They wanted $150 for the privilege of me sending it to them to take a look, so I started looking for a computer repair shop locally. Everyone would turn me away, as the mobo plans were not available to them. This took about 1 month. The laptop's behaviour did not change. I running it off AC only, even disassemblin and repasting the CPU - nothing had any effect.

Fast forward 4 months: I wanted to try another repair shop and switched it on. Miraculously, it switched on. Alas, after 15 min it crashed again, with a BSOD. Here it is:

I tried switching it on a few times more. More crashes ensued. The final one looked like this:

After this, the computer would not boot again. Same problem - the fan spinning up, then down, until I switched off the laptop. I removed the NVME SSD and tried to boot another PC from it. Turns out it was completely fine. I updated Windows on the drive without any problems. The P2 Max would not boot, either with or without the drive.
Fast forward 1 month. The laptop boots again, about 5-10 min each time. I found myself in another country for work, temporarily. I asked around and they told me at a computer shop they would repair it. So I brought it but they won't take it without the SSD (which I removed as it has my data on it). Reluctantly I reinserted the SSD and went there again. Then, after hearing that I won't give them my password for the Windows account and will create a guest account for them. They lost patience with me and told me they wouldn't repair it. So I finally tried to troubleshoot myself. The SSD was not necessary there at all, as the crashes happen if the laptop boots using a Linux liveUSB:

Above you can see the Ubuntu dash being moved to the right a bit. Such graphical artifacts happen often when it crashes. In Windows, I observed a pattern of red dots, subtly visible here on the top:

I ran a few passes of Windows Memory Diagnostics. The first time it crashed before seeing any errors:

Then it crashed at 37% of the first pass. After this, at 37% it did not crash but detected errors at 37/38%:

and continued until the end of the first pass, after which it crashed again:

Which got me thinking whether it might be bad RAM. 37/38% of 16GB is ~6 GB. I tried restricting Windows memory usage using bcdedit /truncatememory, limiting it to 6 GB (4.2 GB reported in Windows) and then ~5.8 GB, but the crashes were still happening:

Then I tried running Memtest86. Firstly, version 9.4 crashed before finding any errors:

Then I ran 9.5 beta a few times. Firstly with the laptop cover open and a fan blowing directly onto the bottom of the laptop for cooling: The first errors were detected at Test 5:

and the test stopped/crashed at Test 7:

Then I closed the chassis and ran it again. This time it crashed earlier:

Finally, I ran the test and put the laptop in the fridge. Here is the result:


I browsed the memtest86 log files to possibly find the faulty areas of memory. But there was no consistent pattern. Here are some examples of first detected errors from v9.5 first test:
Code:
2022-08-28 13:52:18 - [MEM ERROR - Data] Test: 5, CPU: 0, Address: 1149EABCC, Expected: 6590689A, Actual: 6510689A
2022-08-28 13:52:18 - [MEM ERROR - Data] Test: 5, CPU: 0, Address: 1149EABD4, Expected: 6590689A, Actual: 6510689A
2022-08-28 13:52:18 - [MEM ERROR - Data] Test: 5, CPU: 0, Address: 1149EABDC, Expected: 6590689A, Actual: 6510689A
2022-08-28 13:52:18 - [MEM ERROR - Data] Test: 5, CPU: 0, Address: 1149EABE4, Expected: 6590689A, Actual: 6510689A
2022-08-28 13:52:18 - [MEM ERROR - Data] Test: 5, CPU: 0, Address: 1149EABEC, Expected: 6590689A, Actual: 6510689A
Second test:
Code:
2022-08-28 14:20:30 - [MEM ERROR - Data] Test: 3, CPU: 2, Address: 44A275E7C, Expected: 00000000, Actual: 00800000
2022-08-28 14:20:30 - [MEM ERROR - Data] Test: 3, CPU: 2, Address: 44A275E74, Expected: 00000000, Actual: 00800000
2022-08-28 14:20:30 - [MEM ERROR - Data] Test: 3, CPU: 2, Address: 44A275E6C, Expected: 00000000, Actual: 00800000
2022-08-28 14:20:30 - [MEM ERROR - Data] Test: 3, CPU: 2, Address: 44A275E64, Expected: 00000000, Actual: 00800000
2022-08-28 14:20:30 - [MEM ERROR - Data] Test: 3, CPU: 2, Address: 44A275E5C, Expected: 00000000, Actual: 00800000
Third test:
Code:
2022-08-28 14:30:55 - [MEM ERROR - Data] Test: 1, CPU: 0, Address: 409C740, Expected: 000000000409C740, Actual: 008000000409C740
2022-08-28 14:30:55 - [MEM ERROR - Data] Test: 1, CPU: 0, Address: 409C748, Expected: 000000000409C748, Actual: 008000000409C748
2022-08-28 14:30:56 - [MEM ERROR - Data] Test: 1, CPU: 0, Address: 409C750, Expected: 000000000409C750, Actual: 008000000409C750
2022-08-28 14:30:56 - [MEM ERROR - Data] Test: 1, CPU: 0, Address: 409C758, Expected: 000000000409C758, Actual: 008000000409C758
2022-08-28 14:30:56 - [MEM ERROR - Data] Test: 1, CPU: 0, Address: 409C760, Expected: 000000000409C760, Actual: 008000000409C760
The laptop at least started booting again more consistently. I am at a loss about what can be the cause. I looked at the motherboard and can't find anything burnt. The RAM is obviously soldered so I can't swap it. BIOS chip shorts? A CPU problem? Can anyone recommend a repair service that could fix this? I won't be sending it to GPD as they told me that for $150 they can fix it and it they find out they can't, the motherboard costs $500. Which is obviously a joke, for this amount I might as well but a new (and better) ultraportable.

EDIT: I looked at the logs again and the errors keep happening only on CPU 0 and CPU 2. These are probably the first threads of both of the cores. Is there any way to disable first half of the threads in a multithreaded processor?
EDIT2: after the fridge treatment the laptop stopped booting actually - we're back to the fan on/off situation. Now I will need to wait a few weeks for it to boot again, unless there is a solution.
EDIT3/UPDATE: I am kinda disappointed that nobody is answering. I have disassembled the laptop completely and put the motherboard next to a hot electric infrared heater. I kept it there for about 5 seconds, 3 times. I did it twice, once for (what I think is) the RAM chip, and once for the CPU. Then I reassembled the laptop and it booted nicely. I performed another memtest86 and this time it ran for more than 1 hour. The first run completed successfully and the second run crashed near the beginning. So to me it seems that there is a cracked cold solder under the CPU. Is there a way to fix this?

If you've survived this long post, thank you and if you have some good ideas what to do, please share!
 
Last edited:
Thread starter Similar threads Forum Replies Date
D Laptop Tech Support 1
T Laptop Tech Support 1
A Laptop Tech Support 0
W Laptop Tech Support 0
B Laptop Tech Support 0
D Laptop Tech Support 0
Michael@Jensen@WK Laptop Tech Support 0
Hamza ben ali Laptop Tech Support 6
R Laptop Tech Support 0
C Laptop Tech Support 2
Zekie112 Laptop Tech Support 4
Kavya Janaki Laptop Tech Support 1
M Laptop Tech Support 4
A Laptop Tech Support 0
vitalApocalypse Laptop Tech Support 1
S Laptop Tech Support 0
Arthurdensley Laptop Tech Support 1
Kelvin Mii Laptop Tech Support 0
Zenalt Laptop Tech Support 0
C Laptop Tech Support 1

ASK THE COMMUNITY