Graphics Artifacting/Glitching dependent on movement on screen and power settings

Hey everyone.
I’ve been trying to debug this odd issue for the last couple of months or so. I can’t pinpoint the exact time it started, because three potentially relevant things happened:

  1. I updated Fedora 41 to 42
  2. I repasted to PTM7950. I’ve done so dozens of times before for friends/clients computers, routine operation and nothing hardware-wise seemed to have gone wrong.
  3. I updated the bios to 3.06 (could have gotten the exact version wrong).

At some point after some of these operations, the following issue started appearing:
The screen has varying degrees of artifacting everywhere. BIOS, boot screens, installers, windows, installed fedora, live usbs, everywhere. The differences are as follows:

  • In everywhere but Fedora, the artifacts present themselves as black rectangular flickers/blocks, as well as something reminiscent of very bad screen tearing.
  • In Fedora, the artifacts are less severe, without tearing, and like rectangular pink noise instead.
  • The severity of artifacts seems to be directly connected to how much is moving on screen, especially the mouse. If I don’t touch anything in fedora there are basically no artifacts on screen, while moving a mouse in the windows USB installer renders the screen basically unusable
  • Using the “power saver” mode in ppd in fedora eliminates 99.9% of any artifacts, which is why I didn’t notice it immediately. Stepping up to “balanced” introduces them back, with “performance” having even more.
  • Artifacts are more common when the laptop is plugged in
  • Artifacts happen exclusively on the internal screen, never on an external, which is why I still have hope that this is a software issue.
  • I do use a 7900xt over USB4, and that works completely fine.

Here are the debugging steps I’ve tried:

  • Taking off and reinstalling the cooler in case it was a mounting pressure issue, no effect.
  • Reseating the display cable
  • Reinstalling Windows 11. Couldn’t actually physically get through the installer because the installer was unusable because either the artifacting was so bad or it just didn’t display an image. When I did manage to get to a new install, installing Adrenalin drivers seemed to help the issue a lot, until an unrelated thing broke the install and I gave up on windows for the time being.
  • Booting with one stick of RAM (tried both), booting without SSD, no effect.
  • Trying fixes from this thread: Rolling back mesa to version 24, amdgpu.sg_display=0, amdgpu.dcdebugmask=0x10, amdgpu.dcdebugmask=0x400, amdgpu.dcdebugmask=0x410, all with no effect.
  • Upgrading BIOS to 3.09, no effect.

So… that’s about all the information I have gathered on this issue so far. I’m still holding on to the hope that it is a software issue as getting a replacement is currently problematic due to me being away from a supported shipping country for quite some time. Unfortunately, I have ran out of ideas on what I can do, and am hoping someone here does. Will be waiting for your replies!

  • Which OS?
    Fedora 41-42, Windows 11, Doesn’t seem to matter
  • Which Framework product and which generation?
    Framework 13, 7640u, 2.8k display, this 64gb RAM kit, Samsung 2tb nvme ssd

Did you check your thermal paste application? Graphic glitches in the UEFI/BIOS and boot screens are indicative of a problem with the gpu, not a software issue (there’s no driver loaded at that point - just a simple frame buffer). I’ve seen glitching like that with a few GPUs, both discrete and integrated, and it was either a hardware failure (memory or component failure) or because of a problem with the TIM application.
Fingers crossed that’s what the problem is.

Yes, I did, the application is fine… as well as all temperatures are expectedly good for ptm. Moreover, I remembered that I played multiple 3d games on the iGPU, with the internal screen, with no glitches there.

Edit: double-checked.

1 Like

Would you be able to catch a video showing the glitching artifacts?
Upload it somewhere and link to it here.

Here, I cobbled together a quick compilation.

That is a hardware fault.
It might be a fault with the mainboard or the display.
Note, I am just another user like you.
You need to raise a support ticket with the framework website to get them to help / replace the parts.

If it is then it is very unfortunate. The laptop is pretty much fully functional, especially docked… And I am currently outside of the countries supported by Framework to ship to. If there is a chance this can be fixed at home, then I’d take it… Otherwise I guess I have to go on an impromptu holiday to Europe if I need to replace the mainboard/display…

With the display glitching.
Have you checked the cable connections both ends.
One end on the main board, the other end at the back of the display.

Yes, checked both.

That looks like a signal integrity issue. Could be the cable, could be a trace on the motherboard, could be the electronics on the display itself, could be the connectors, the only thing that jumped out at me was the extra paste around the CPU. That can cause components to heat up unnecessarily and potentially put some of the caps and resistors out of spec.

You might try starting there and see if that has an impact. Otherwise, there are a number of places it could be but it looks like a hardware fault for sure since it happens literally after booting and on the BIOS screens which should not be taxing at all.