FW16 - 7700S Bazzite GPU ring timeout causing newer games freeze/OS stays responsive (RESOLVED with RMA)

Community Support
, ,
1

I recently bought a refurbished 7700S to get into PC gaming. I’m still very new to Linux, so bear with me on this. For context I did use AI to help me troubleshoot a little.

**I want to determine if this is a software or hardware issue before my return window closes on the refurb 7700s
**
Issue: GPU ring timeout causing game freezes. Only the game freezes, OS remains responsive.

  • Framework Laptop 16

  • Official 180w FW brick/cable

  • CPU: AMD Ryzen 7840HS

  • GPU: AMD RX 7700S

  • Bazzite 44.20260608 (Fedora 44 based)

  • Kernel: 7.0.9-ogc3.2.fc44.x86_64

  • Mesa: 26.1.0

amdgpu 0000:03:00.0: smu driver if version = 0x00000035, smu fw if version = 0x00000040
amdgpu 0000:03:00.0: SMU driver if version not matched
amdgpu 0000:03:00.0: ring gfx_0.0.0 timeout, signaled seq=50521, emitted seq=50523
amdgpu 0000:03:00.0: device wedged, but recovered through reset

Impacted games so far: Cyberpunk 2077, Starfield, both DX12/Vulkan titles via Proton?)

Not affected: FEAR, Serious Sam (3, 4, Siberian Mayhem), Final Fantasy XIII, Poly Bridge 3, Warframe (tbd, but I think this one is safe)

Workarounds attempted:

  • amdgpu.runpm=0 — did not resolve
  • Multiple Proton versions including GE-Proton — no improvement
  • BIOS version: 04.04 (beta)

It seems similar to Managing AMDGPU power state causes crash · Issue #190 · FrameworkComputer/SoftwareFirmwareIssueTracker · GitHub

Resolved via RMA. Support sent me a replacement and it works without issue. What a very strange/odd situation though. It bugs me not to know what exactly these games were triggering on a hardware level that would cause this issue.

2

Because the OS stays responsive and amdgpu reports a gfx ring timeout/reset, I’d try to separate “7700S bad” from “current Bazzite/Mesa/kernel path bad” before the return window closes. If you can reproduce it quickly, test one known case after booting a plain Fedora/Kinoite live or a different installed kernel, with no extra launch options except the default Proton. Also save the exact window around the reset with journalctl -k -b and filter for amdgpu/ring/reset/wedged/smu. If it follows the card across a stock-ish distro/kernel and the same power brick, that is stronger RMA evidence than Proton tweaks.

3

That was a great suggestion. I was able to reproduce it on Windows… sadly I think it’s a bad 7700s :frowning: I’ll open a case with support

4

Make sure to give our support a shout, if you haven’t already. Ask to speak to the Linux team and they’ll help you weed out the issue.

5

Resolved. I had to RMA the dGPU. The replacement came today and doesn’t exhibit the same issues. Thanks support!