P870tm issue with clevo rtx 3080

[Maro97]

Hello everyone 

The laptop suddenly loses power while I’m playing games, as if I had unplugged the charger from the laptop while the battery is not installed. What could be causing this problem?

[Maro97]

The image changed to look like this after I installed the graphics card driver in Windows.

[Khenglish]

Your card has broken solder connections between the core and vram. It either needs to be repaired or replaced.

[Scruffy]

can be bad mxm slot ( is taht the one with vapor chamber ?)  or is this one where u put other vga like 980 mxm and so on ?

 

before conclusion tell us  if  u have more notebook or just one...

 

notebook pwoer off can be cpu overclock..... try a stable multiplyer for 4.2ghz ....

[Maro97]

52 minutes ago, Scruffy said:

can be bad mxm slot ( is taht the one with vapor chamber ?)  or is this one where u put other vga like 980 mxm and so on ?

 

before conclusion tell us  if  u have more notebook or just one...

 

notebook pwoer off can be cpu overclock..... try a stable multiplyer for 4.2ghz ....

 

 

[Maro97]

What do you guys think? Could this be a problem with something else besides the graphics card?

[Maro97]

45 minutes ago, Scruffy said:

can be bad mxm slot ( is taht the one with vapor chamber ?)  or is this one where u put other vga like 980 mxm and so on ?

 

before conclusion tell us  if  u have more notebook or just one...

 

notebook pwoer off can be cpu overclock..... try a stable multiplyer for 4.2ghz ....

I have a clevo p870tm with a modded vapour chamber heatsink with the 4 arms adapter to support rtx 3080 it’s a second gpu which i bought from the same seller on ebay the first one worked without no issues before i cracked it 

[Scruffy]

can be bad chip, but so many of them went bad ? maybe core problem....

 

try reballing it.

[Maro97]

19 minutes ago, Scruffy said:

can be bad chip, but so many of them went bad ? maybe core problem....

 

try reballing it.

 

[janepa]

2 hours ago, Maro97 said:

can be bad chip, but so many of them went bad ? maybe core problem....

 

try reballing it.

 

Hi, this might sound a little weird, but I’ve already encountered something similar a few times and when I wasn’t completely sure if reballing would help, I put the card in an electric oven for 10 minutes at 180 °C. Maybe the artifacts will disappear then. But it’s also possible that one of the memory chips is defective. The GPU core temperature might be very low, but that doesn’t mean the VRAM temperature is the same.

 

[Scruffy]

never do reflow... it only stress cpu, vram and all other component. and it's not a proper reballing.

 

send card tto someone that has the right equipment to do a proper job

 

hard part is to be able to test the vga .. if they dont have an mxm notebook to try after they do rework and reballing.

 

3080 available on ebay for like 330€.

 

 

[Maro97]

Another issue with rtx 3080 

The screw connecting the modded vapor chamber heatsink to the adapter was installed without the size-reducing nut, which caused the screw to be the wrong size. Of course, the screw scratched the PCB and caused a short circuit on the GPU.

Therefore, I strongly recommend always installing the nut and making sure it is properly in place before tightening the screw.

[Maro97]

The short circuit that happened on the MXM card has been fixed. But a short occurred on the 1.3-volt rail that powers the VRAM, which caused some of the VRAM modules to malfunction. I tested it using MATS. This is the result I got. Can someone tell me how many VRAM chips need to be replaced?

 

 

 

 

 

[Scruffy]

from the second picture, looks like some trace is broken. always use ipa to make the pcb and trace more visible. if u can make a better picture with ipa on it...

[Maro97]

1 hour ago, Scruffy said:

from the second picture, looks like some trace is broken. always use ipa to make the pcb and trace more visible. if u can make a better picture with ipa on it...

My friend, the repair process has already been completed, and I have the MATS test report as the result.

The card had a short circuit, and I already explained everything about it. My question is: based on the MATS report, how many VRAM chips need to be replaced?

 

[harkaz]

@Maro97 Most likely all chips that report errors (6/8). Start with the 2 that appear to be physically damaged and report most errors then retest.