I’m fairly new to the Fedora ecosystem and I’m looking for help with a graphics card issue.
System details
Laptop: Alienware m18 R1
GPU: NVIDIA GeForce RTX 4080 Max-Q (Mobile)
OS: Fedora
Drivers: NVIDIA proprietary drivers (installed and reinstalled following multiple guides)
Problem
Whenever the GPU is put under heavy load, the system crashes to a black screen. For example, this happens immediately when starting Cyberpunk 2077. Other, less demanding games like Stronghold Crusader run without any issues.
Once the black screen appears, the laptop no longer responds to any input. The only way to recover is to perform a hard reset.
What I’ve tried
Reinstalling the NVIDIA proprietary drivers multiple times
Following various guides and suggested fixes found online
Consulting search engines and LLMs for troubleshooting steps
Unfortunately, none of these attempts have stabilized the system.
Question
Does anyone have ideas or suggestions on what else I could try to diagnose or fix this issue?
Please post the output of nvidia-smi and inxi -Fzxx. Post both of those outputs as preformatted text so it retains the same formatting as seen on-screen. This is accomplished by Copy & Paste the text here then highlight the text and click the </> button on the toolbar just above the text entry window.
The nvidia-smi output shows that gnome-shell is using the nvidia gpu.
The only thing I see strange is that the GPU is showing 100% usage with only 174MiB of GRAM used and only 4W power usage.
You also have 2 moniors, 1 is 4k (3840x2160) @ 16x9 and the other is 2k? (2560x1600) @ 16x10. Theoretically the differing aspect ratios there could be a factor as well as the fact the built-in screen is using DP and the external screen is using HDMI.
Possibly try
setting both to the same aspect ratio
setting both to the same refresh rate
connecting the external monitor to DP instead of HDMI
Maybe someone who has the same GPU NVIDIA AD104M [GeForce RTX 4080 Max-Q / Mobile] can assist.
My desktop has this for the nvidia-smi command.
$ nvidia-smi
Wed Jan 7 16:51:02 2026
+-----------------------------------------------------------------------------------------+
| NVIDIA-SMI 580.119.02 Driver Version: 580.119.02 CUDA Version: 13.0 |
+-----------------------------------------+------------------------+----------------------+
| GPU Name Persistence-M | Bus-Id Disp.A | Volatile Uncorr. ECC |
| Fan Temp Perf Pwr:Usage/Cap | Memory-Usage | GPU-Util Compute M. |
| | | MIG M. |
|=========================================+========================+======================|
| 0 NVIDIA GeForce RTX 3050 Off | 00000000:06:00.0 On | N/A |
| 0% 38C P8 9W / 130W | 512MiB / 8192MiB | 1% Default |
| | | N/A |
+-----------------------------------------+------------------------+----------------------+
+-----------------------------------------------------------------------------------------+
| Processes: |
| GPU GI CI PID Type Process name GPU Memory |
| ID ID Usage |
|=========================================================================================|
| 0 N/A N/A 4098 G ...c/gnome-remote-desktop-daemon 1MiB |
| 0 N/A N/A 4165 G /usr/bin/gnome-shell 221MiB |
| 0 N/A N/A 4332 G /usr/bin/Xwayland 4MiB |
| 0 N/A N/A 4479 C+G /usr/bin/ptyxis 28MiB |
| 0 N/A N/A 6838 G ...rack-uuid=3190708988185955192 111MiB |
+-----------------------------------------------------------------------------------------+
Note that even with several items using the GPU it shows only 1% active.
Please also show the output of dnf list --installed \*nvidia\*
I ran the command again and this time it reported 0%, so it doesn’t seem to behave the same way every time.
I’m using both the laptop’s integrated display and an external monitor. The black screen issue occurs even when only a single monitor is connected. I tested both the internal display and the external monitor individually, with the same result. I also tried connecting the external monitor via USB-C, but the behavior is unchanged.
After today’s kernel update, there was a small improvement: the game now runs for a bit longer (around 5 minutes) before crashing. Unfortunately, the crash still happens.