NVIDIA Driver Installation

Required: Kernel 6.10+ for Ryzen AI 9 HX 370 support.

uname -r

If encountering checksum errors during dnf update, clean the cache:

sudo dnf clean all
sudo dnf makecache

RPM Fusion provides NVIDIA drivers for Fedora. NVIDIA’s official CUDA repository does not yet support Fedora 43.

sudo dnf install https://download1.rpmfusion.org/free/fedora/rpmfusion-free-release-$(rpm -E %fedora).noarch.rpm https://download1.rpmfusion.org/nonfree/fedora/rpmfusion-nonfree-release-$(rpm -E %fedora).noarch.rpm -y

sudo dnf update -y

Wait for update completion.

Check available NVIDIA driver version:

dnf info akmod-nvidia

Confirm the version matches the current release for Fedora 43.

Install driver with CUDA support:

sudo dnf install akmod-nvidia xorg-x11-drv-nvidia-cuda -y

This installs the driver, CUDA libraries, and build dependencies (about 1 GB).

• akmod-nvidia - Automatic kernel module builder
• xorg-x11-drv-nvidia - NVIDIA driver (supports both X11 and Wayland)
• xorg-x11-drv-nvidia-cuda - CUDA libraries
• Build dependencies (gcc, kernel-devel, etc.)

Note: A MOK password prompt may not appear during installation. This is normal.

Force akmod to build NVIDIA kernel modules:

sudo akmods --force

This process may take 5-10 minutes.

Check that kernel modules were built:

ls /lib/modules/$(uname -r)/extra/nvidia/

All five kernel modules should be present.

sudo reboot

After reboot, open GNOME Software and note the MOK enrollment code. The driver is not yet active.

Reboot again:

sudo reboot

During boot, the MOK Management screen (blue screen) will appear:

1. Select “Enroll MOK”
2. Select “Continue”
3. Select “Yes”
4. Enter the MOK enrollment code from GNOME Software
5. Reboot

The system will boot normally after MOK enrollment.

After MOK enrollment, rebuild the kernel modules. They will now be signed with the enrolled key.

sudo akmods --force --rebuild

sudo reboot

The NVIDIA driver will now load correctly. GNOME Software should show the driver as installed (not pending).

Enable NVIDIA power services for better suspend/resume behavior and power management:

sudo systemctl enable nvidia-hibernate.service nvidia-suspend.service nvidia-resume.service

What these services do:

• nvidia-hibernate.service - Properly saves GPU state before hibernation
• nvidia-suspend.service - Manages GPU state during system suspend
• nvidia-resume.service - Restores GPU state after resume

These services prevent GPU state issues after suspend/resume cycles.

Important: Do NOT enable nvidia-powerd — mask it permanently

The nvidia-powerd.service manages NVIDIA Dynamic Boost, which shifts extra wattage (~5-15W) from the CPU to the GPU during heavy GPU loads. While useful on Intel-based laptops, it conflicts with AMD ATPX power management on the Zephyrus G16 and causes soft lockups and “GPU has fallen off the bus” errors.

On this laptop, GPU power is managed via ATPX (AMD-driven via ACPI). The NVIDIA suspend/hibernate/resume services and supergfxctl handle power states correctly without nvidia-powerd.

What you lose by disabling it: Minimal — a few FPS less during heavy GPU workloads. The ~5-15W Dynamic Boost is not worth the instability on AMD ATPX hardware.

Disable and mask permanently:

sudo systemctl disable nvidia-powerd.service
sudo systemctl stop nvidia-powerd.service
sudo systemctl mask nvidia-powerd.service

Masking creates a symlink to /dev/null, preventing any process — including NVIDIA driver updates via dnf — from re-enabling the service.

If you want to try re-enabling it later (e.g., after a kernel or driver update that may fix the ATPX conflict):

sudo systemctl unmask nvidia-powerd.service
sudo systemctl enable --now nvidia-powerd.service

Reference:

• NVIDIA Power Management Documentation

After reboot, check driver status:

nvidia-smi

You should see the NVIDIA driver and CUDA versions listed.

Confirm running Wayland (not X11):

echo $XDG_SESSION_TYPE

lsmod | grep nvidia

The NVIDIA modules are loaded and the driver is functional.

Open GNOME Software (white bag icon):

• Navigate to “Installed”
• Search for “NVIDIA Linux Graphics Driver”
• Status should show “Installed” (not “Pending”)
• “Uninstall” button is visible

This confirms the system recognizes the driver as properly installed.

Adding certain kernel parameters can improve NVIDIA driver performance, especially for Wayland sessions and dual-GPU setups.

Step 1: Add recommended kernel parameters

sudo grubby --update-kernel=ALL --args="rd.driver.blacklist=nouveau modprobe.blacklist=nouveau nvidia-drm.modeset=1 nvidia-drm.fbdev=1 nvidia.NVreg_PreserveVideoMemoryAllocations=1"

Step 2: Verify parameters were added

sudo grubby --info=ALL | grep args

Expected output should include the added kernel parameters.

Step 3: Reboot to apply changes

sudo reboot

What these parameters do:

• rd.driver.blacklist=nouveau - Prevents the open-source Nouveau driver from loading during early boot (initramfs)
• modprobe.blacklist=nouveau - Prevents Nouveau from loading after boot
• nvidia-drm.modeset=1 - Enables NVIDIA kernel mode setting for better Wayland support and performance
• nvidia-drm.fbdev=1 - Makes NVIDIA use its framebuffer via the kernel DRM framework instead of a generic framebuffer. Improves handoff between console and Wayland/GNOME and prevents race conditions during suspend/resume on hybrid GPU laptops
• nvidia.NVreg_PreserveVideoMemoryAllocations=1 - Preserves VRAM allocations during suspend/resume instead of releasing and rebuilding them. Prevents corrupted VRAM after resume, which can cause soft lockups

Why blacklist Nouveau:

• The proprietary NVIDIA driver and Nouveau cannot coexist
• Blacklisting prevents conflicts and ensures the proprietary driver is always used
• If the NVIDIA driver fails, you can remove these parameters from GRUB to fall back to Nouveau

Benefits:

• Better Wayland performance and stability
• Prevents driver conflicts during boot
• Improved external monitor support
• More stable suspend/resume cycles on hybrid GPU setups
• Smoother graphics performance in general

Note: These parameters are optional but recommended for optimal performance.

Re-enable graphical boot splash:

Fedora uses rhgb (Red Hat Graphical Boot) and quiet to show a Plymouth splash screen during boot instead of scrolling kernel text. If you removed these while debugging NVIDIA or boot issues, re-add them:

sudo grubby --update-kernel=ALL --args="rhgb quiet"

The default Plymouth theme (bgrt) shows the ASUS/BIOS manufacturer logo. To debug boot issues in the future, you can temporarily remove them:

sudo grubby --update-kernel=ALL --remove-args="rhgb quiet"

References:

• NVIDIA Driver Modesetting - Arch Wiki
• Understanding nvidia-drm.modeset=1 - NVIDIA Developer Forums
• NVIDIA Power Management Documentation

The GA605WV ships with a Sharp LQ160R1JW02 16" 2560x1600 240Hz display. ASUS factory-calibrates each panel and provides color profiles via their ASUS System Control Interface. On Windows, these are automatically applied by Armoury Crate/GameVisual. On Linux, we must install them manually.

These color profiles were extracted from ASUS Windows driver packages and optimized for GNOME Color Management.

Install the color profiles:

The ICC color profiles are located in the /icc-profiles/ directory of this repository. Clone the repository or manually download the profiles and copy them to ~/.local/share/icc:

mkdir -p ~/.local/share/icc

# If you've already cloned the repository:
cp /icc-profiles/*.icm ~/.local/share/icc/

# Or download the specific profiles you need from the repository

Activate Native profile in GNOME:

1. Open Settings → Color Management
2. Select Built-In Screen
3. Click Add Profile
4. Select Native
5. Click Add

Note: If GNOME Settings shows old technical names (e.g., “ASUS GA605WV 1002 104D158E CMDEF” instead of “Native”), close Settings and reopen, or log out/in to refresh the color cache.

Available color profiles:

Recommendation:

Use Native for best color accuracy. This profile contains factory calibration specific to the Sharp LQ160R1JW02 panel in this laptop. The other profiles (DCI-P3, Display P3, sRGB) are generic colorspaces without panel-specific corrections.

Note: The _1002_ in the filename refers to the AMD iGPU (Vendor ID 0x1002), which drives the internal eDP display on this hybrid GPU laptop.

Background:

The profiles were found through analysis of ASUS Windows driver packages. The ASUS CDN URL structure:

https://dlcdn-rogboxbu1.asus.com/pub/ASUS/APService/Gaming/SYS/ROGS/{id}-{code}-{hash}.zip

For the GA605WV, this is: 20016-BWVQPK-01624c1cdd5a3c05252bad472fab1240.zip

The profiles contain factory color corrections specific to the Sharp LQ160R1JW02 panel (Panel ID: 104D158E) used in this laptop model.

Technical Details:

The profiles in this repository are pre-processed with custom ICC metadata ‘desc’ tags so they appear with readable names directly in GNOME Color Management. For users interested in how such modifications work, you can implement similar ICC ‘desc’ tag manipulation yourself using Python’s PIL/ImageCms.

Problem: System freezes or crashes when using external monitors via Thunderbolt/USB-C, especially when connecting/disconnecting displays. Logs show AMD GPU errors:

amdgpu 0000:66:00.0: amdgpu: MES failed to respond to msg=RESET
amdgpu 0000:66:00.0: amdgpu: Ring gfx_0.0.0 reset failed
amdgpu 0000:66:00.0: amdgpu: GPU reset begin!

Cause: This laptop has dual GPUs (AMD Radeon 890M integrated + NVIDIA RTX 4060 discrete). The AMD GPU’s PSR (Panel Self Refresh) feature has a bug causing crashes with external Thunderbolt monitors.

Solution: Disable AMD PSR by adding a kernel parameter:

sudo grubby --update-kernel=ALL --args="amdgpu.dcdebugmask=0x600"

Verify it was added:

sudo grubby --info=ALL | grep args

Reboot:

sudo reboot

What this does:

• amdgpu.dcdebugmask=0x600 disables PSR (Panel Self Refresh) on the AMD GPU
• PSR is a power-saving feature where the display refreshes itself without GPU involvement
• The PSR implementation has bugs with Thunderbolt/USB-C external monitors

Trade-offs:

• Pro: Stable system with external monitors
• Con: Slightly higher power consumption (PSR disabled)

Verification: Monitor for AMD GPU errors while using external displays:

sudo journalctl -f -k | grep -i amdgpu

If no amdgpu: [drm] *ERROR* messages appear, the fix is working.

Reference:

• Fedora Discussion: Zephyrus G16 External Monitor Crashes

What’s happening: System freezes completely during VS Code use. Kernel 6.18.x/6.19.x have critical amdgpu driver bugs. VS Code hardware acceleration triggers AMD Radeon 890M page fault → complete freeze.

Fix: Add to ~/.config/Code/User/settings.json:

{
    "disable-hardware-acceleration": true
}

Next steps: Restart VS Code. System stays stable, VS Code slightly slower but perfectly usable.

Sources:

• VS Code Issue #238088
• Framework: Critical amdgpu bugs kernel 6.18.x

What’s happening: System freezes or crashes during Brave Browser use, even with minimal workload (a few tabs). This is the same underlying issue as the VS Code crash: Chromium-based applications with hardware acceleration trigger AMD Radeon 890M page faults on kernel 6.18.x/6.19.x.

Typical crash sequence in logs:

amdgpu: [gfxhub] page fault (src_id:0 ring:24 vmid:2)
amdgpu: Faulty UTCL2 client ID: SQC (data)
amdgpu: ring gfx_0.0.0 timeout, signaled seq=302899, emitted seq=302901
amdgpu: GPU reset begin!

After GPU reset, gnome-shell crashes (Signal 6 ABRT) because it detects a context reset.

Fix: Open Brave Browser and go to brave://settings/system. Turn off “Use hardware acceleration when available”.

Alternatively via terminal:

sed -i 's/"hardware_acceleration_mode_previous":true/"hardware_acceleration_mode_previous":false/' ~/.config/BraveSoftware/Brave-Browser/Local\ State

Or start Brave with the --disable-gpu flag:

brave-browser-stable --disable-gpu

Next steps: Restart Brave. Verify via brave://gpu that GPU acceleration is disabled. System stays stable, Brave is slightly slower on heavy pages but perfectly usable.

Background: Brave, VS Code, and other Chromium-based applications (Chrome, Edge, Electron apps) use GPU shader compilation via Mesa. On kernel 6.18.x, the amdgpu driver has a bug in the Shader Queue Controller (SQC) memory access, causing page faults that trigger a full GPU reset. The fix is to disable hardware acceleration per application until a kernel/Mesa update resolves the issue.

Sources:

• Framework: Critical amdgpu bugs kernel 6.18.x

What’s happening: System freezes with an NVIDIA soft lockup, even without active GPU use. Kernel logs show:

watchdog: BUG: soft lockup - CPU#23 stuck for 62s!
NVRM: Xid (PCI:0000:65:00): 79, pid=<...>, GPU has fallen off the bus

This can occur due to a combination of factors on hybrid GPU laptops:

• nvidia-powerd conflicts with AMD ATPX power management
• NVIDIA dGPU power state transitions fail
• Corrupted VRAM after suspend/resume cycles

Additional symptom: Reboot hang (black screen, backlights stay on)

The system appears to shut down but never completes the hardware reset — the screen goes black but keyboard and screen backlights remain on. This occurs when nvidia-powerd interferes with ACPI power state transitions during shutdown/reboot.

Root cause: supergfxd starts nvidia-powerd behind your back

Even when nvidia-powerd is disabled via systemctl disable, supergfxd (the GPU switching daemon from asusctl) directly calls systemctl start nvidia-powerd.service during GPU mode switches. This bypasses the disabled state and re-activates the conflict with ATPX.

How this was diagnosed:

Checking the logs of the hung boot reveals supergfxd starting nvidia-powerd:

journalctl -b -1 --no-pager | grep -iE "nvidia.*powerd|supergfxd"

Key evidence:

supergfxd: [DEBUG supergfxctl] Did CommandArgs { inner: ["start", "nvidia-powerd.service"] }
nvidia-powerd: ERROR! Client (presumably SBIOS) has requested to disable Dynamic Boost DC controller

The SBIOS error confirms the firmware rejected Dynamic Boost, but nvidia-powerd was already running and interfering with power state management. Checking the shutdown sequence:

journalctl -b -1 --reverse | head -20

Shows the hardware watchdog failed to stop, confirming the ACPI reboot never completed:

watchdog: watchdog0: watchdog did not stop!

Fix:

1. Disable and mask nvidia-powerd (masking is essential — disable alone is not enough because supergfxd bypasses it):

sudo systemctl disable nvidia-powerd.service
sudo systemctl stop nvidia-powerd.service
sudo systemctl mask nvidia-powerd.service

2. Add kernel parameters for more stable NVIDIA power management:

sudo grubby --update-kernel=ALL --args="nvidia-drm.fbdev=1 nvidia.NVreg_PreserveVideoMemoryAllocations=1"

3. Reboot:

sudo reboot

Next steps: System is more stable after these changes. The NVIDIA dGPU is still properly managed via ATPX (AMD-driven power switching) without nvidia-powerd interfering. The mask creates a symlink to /dev/null, ensuring no process — including supergfxd and NVIDIA driver updates — can re-enable the service.

Background: On laptops with AMD iGPU + NVIDIA dGPU, the ATPX framework (via ACPI) controls which GPU is active. nvidia-powerd tries to make power decisions independently, which conflicts with ATPX. The NVreg_PreserveVideoMemoryAllocations=1 parameter prevents VRAM from being lost during power transitions, and nvidia-drm.fbdev=1 provides cleaner framebuffer handoff.

Check if NVIDIA modules are loaded:

lsmod | grep nvidia

Check system logs for errors:

sudo journalctl -b | grep nvidia

Rebuild kernel modules:

sudo akmods --force --rebuild
sudo reboot

If you receive the error modprobe: ERROR: could not insert 'nvidia': Key was rejected by service, the kernel modules were built before MOK enrollment completed.

Solution:

# Rebuild modules after MOK enrollment
sudo akmods --force --rebuild

# Reboot
sudo reboot

To reset MOK if needed:

sudo mokutil --reset

Reboot and attempt enrollment again.

Check session type:

echo $XDG_SESSION_TYPE

If output is x11, ensure Wayland is enabled in GDM:

sudo nano /etc/gdm/custom.conf

Verify this line is present and not commented:

WaylandEnable=true

Reboot after changes.

Ensure kernel headers match running kernel:

sudo dnf install kernel-devel-$(uname -r)

Force rebuild:

sudo akmods --force