ASUS
ROG Zephyrus G14 GA403UV
Twenty-two exact-model reports show a capable Linux laptop whose NVIDIA switching, hybrid-mode brightness and sleep behavior still depend on the kernel and configuration.
Bottom line
Omarchy can run on the GA403UV, and one owner reports a stable daily setup. The larger Linux record shows why this is not yet a low-maintenance recommendation: exact owners have hit MUX black screens, recurring NVIDIA stutter or wakeups, hybrid-mode brightness failures, touchpad jitter and kernel-specific resume regressions. Stock Arch's current table reports video, sound, Wi-Fi, Bluetooth and power management working, with nvidia-open preferred to avoid crashes. That is a useful baseline, not proof of the complete Omarchy experience.
Existing owners have a credible path if they are willing to manage GPU state and keep a known-good kernel installed. Buyers should verify the full SKU because the soldered memory varies by market, then test Secure Boot, hybrid and integrated graphics, brightness, repeated sleep cycles, touchpad edges, external outputs, speakers and microphones during the return period. The RTX 4060 gives this machine real gaming and GPU-compute headroom, but integrated-only hardware remains the simpler choice when seamless unplugged-to-docked use matters more.
What is reported
12 findings with cited evidenceInstall and boot
Owners reached working Omarchy, Arch, Manjaro, CachyOS, EndeavourOS, Fedora, Ubuntu, Mint, NixOS and Bazzite systems. One Mint installation could not load its NVIDIA module with Secure Boot until the owner rebuilt and enrolled the signed module.
Installation itself is well established. A booted desktop does not settle the harder questions about GPU modes, brightness, sleep or ASUS controls, and Secure Boot adds a driver-signing check.
Desktop and graphics
Stock Arch now recommends nvidia-open to avoid crashes, while one Omarchy owner keeps separate kernels for changing GPU state and for NVIDIA work. Other exact reports include a MUX-mode black screen, game-only black screens after an update, severe desktop stutter, a dGPU that woke after lid cycles and an idle display deadlock.
Both GPUs are usable, but kernel, driver and mode changes can alter the result. Keep a working kernel available and test integrated, hybrid and discrete-only modes separately.
Internal display and brightness
Exact reports drive the OLED at its native resolution and 120 Hz. The first Arch entry needed the native ACPI backlight path, an Omarchy owner added a model-specific brightness script, and two Mint owners found hybrid mode targeting the inactive NVIDIA backlight instead of the panel's AMD path.
The panel works, but brightness can depend on GPU mode and configuration. Integrated-only mode worked in the Mint report; no retained exact-model Linux report validates variable refresh rate.
Keyboard lighting and ASUS controls
Earlier Arch, Ubuntu and NixOS reports found incomplete lighting, fan or charge controls. A later Omarchy owner controlled the keyboard backlight and bound the M4 key; asusctl 6.3.3 subsequently added GA403UV LED support upstream.
The support landing is meaningful, but it is project evidence rather than an exact-owner retest of every lighting mode, hotkey, fan curve and charge limit.
Touchpad and pointing
One CachyOS owner reported a jittery touchpad, lag and inactive outer edges while an external mouse remained normal.
A later suggestion changed pointer settings, but the commenter did not identify an exact model. There is no retained exact-machine confirmation that the touchpad problem was fixed.
Problems reportedSourceFingerprint and biometrics
The U.S. specification lists an infrared camera but no fingerprint reader. No retained exact-model Linux report validates face recognition.
Treat biometric login as unverified. Camera detection is not evidence that the infrared authentication path works.
Evidence, no verdictSourceWi-Fi and Bluetooth
Stock Arch marks Wi-Fi and Bluetooth working, and several exact-system inventories show the MT7922 connected and Bluetooth active. One Omarchy hibernation report still logged a firmware reload, brief disconnect and several-second freeze after each resume.
Normal connectivity has positive evidence. Resume recovery is less seamless, and the corpus does not measure throughput or long-session Bluetooth reliability.
Speakers, microphone and audio
Stock Arch marks sound working, and several other exact systems show active ALSA and PipeWire devices.
This establishes a sound path, not the quality or completeness of the six-speaker array, bass, microphones and headset jack. Those remain unisolated in the retained reports.
Camera
Several exact Linux inventories identify the Shinetech RGB camera and bind it to uvcvideo, but none records an application-level camera test or a working infrared stream.
The device is detected; ordinary video and infrared face authentication are not verified outcomes here.
Suspend and resume
Stock Arch marks power management working. One Ubuntu owner found Linux 6.17 resumed correctly but 7.0 consistently panicked after extended s2idle, while one Omarchy hibernation failure cleared only after changing systemd from platform to shutdown mode.
Sleep behavior is version-sensitive. Keep a known-good kernel, test repeated long sleeps and check graphics, network and audio after wake rather than stopping when the display returns.
External displays, docks and ports
One Arch owner found the internal Realtek microSD reader detected and bound but unable to mount media. A later Fedora owner could avoid an NVIDIA deadlock with integrated graphics, but that also disabled outputs wired to the discrete GPU.
The low-risk graphics workaround conflicts with docked use. No retained exact-model report confirms USB-C display output or USB4 docking, and the card-reader report remains unresolved.
Battery, power and thermals
One Omarchy owner observed 7 to 10 watts with the dGPU off and 14 to 25 watts with it on. Other reports include idle fans, a frozen battery reading, intermittent charge-threshold support, stale AC/DC control state and firmware- or lid-triggered NVIDIA wakeups. The Ubuntu charge-threshold owner reported the control working after moving to Ubuntu 24.04 and Linux 6.10.
Turning off the dGPU can materially reduce draw, but recurring wakeups can erode the gain. These are owner observations rather than controlled Omarchy endurance tests; ASUS-specific power-limit support also depends on a sufficiently new kernel.
Changes over time
23 dated events- Model-specific ASUS controls requested
An early Arch report found ineffective keyboard and matrix lighting and fans turning at no load.
- First stock Arch entry records graphics setup caveats
The first ArchWiki row marked the main hardware paths working but documented Optimus trouble and a native-backlight setting. Later revisions required the NVIDIA 555 generation and additional Wayland configuration; the current row prefers nvidia-open to avoid crashes.
- Fan-curve control fails after initially working
A Fedora owner found model-specific controls and fan curves, then reported that the fan-curve path stopped applying and fans ran below the configured zero-rpm threshold.
- Control gaps persist on another Linux stack
An Ubuntu owner reported missing keyboard-brightness, charge-limit and fan-curve controls on the exact model.
- Discrete-only MUX mode produces a black screen
A Manjaro owner could log in behind a black internal display after switching to the hardware MUX's discrete-GPU mode. The thread did not record an exact-machine fix.
- Battery status stops updating
A Bazzite owner reported that the displayed percentage and charging state froze instead of tracking the battery.
- Owner works around dGPU wake after lid cycles
An Arch owner found the dGPU powering on after closing and reopening the lid in Integrated mode. A suspend service that cycled through VFIO restored Integrated mode but left persistent notifications.
- Long-session NVIDIA stutter reported
An EndeavourOS owner reported that the desktop periodically fell to about 5 frames per second after several hours and recovered only after rebooting.
- Notification dismissal changes dGPU state
A Fedora owner found that clearing a ROG Control Center notification toggled the dGPU instead of merely dismissing the message. Duplicate filings describe the same owner and incident and count once.
- CachyOS update breaks video in games
After an update, one owner got black video with continuing audio in several games. A WineD3D fallback restored one title.
- Later CachyOS update restores game video
The same owner reported normal game rendering after a larger update that included the kernel, NVIDIA driver and Mesa. The responsible package was not isolated.
- Internal microSD reader remains unusable
An Arch linux-g14 owner found the Realtek reader detected and bound to its driver, but inserted media would not mount. The report remains unresolved.
- Touchpad jitter and dead edges reported
One exact-model CachyOS owner reported lag, jitter and inactive outer touchpad areas while an external mouse worked normally.
- Two-kernel Omarchy setup reaches stable daily use
One owner documented a stable setup that uses linux-g14 to change dGPU state and a mainstream kernel for normal work, games and compute. The same report added working brightness and keyboard paths plus dGPU-on and dGPU-off power readings.
- Hybrid brightness targets the wrong backlight
A Mint owner found that ordinary controls changed the NVIDIA backlight while the internal panel followed the AMD path. Integrated-only brightness and direct AMD control worked.
- Signed NVIDIA-module rebuild restores Secure Boot path
A second Mint owner restored on-demand NVIDIA use under Secure Boot by rebuilding and enrolling the open module. Hybrid brightness remained a separate problem.
- asusctl adds GA403UV LED support
The upstream 6.3.3 release named the model in its LED-support change. This is a capability landing, not a complete owner retest.
- Custom ACPI workaround suppresses dGPU wakeups
One owner reported long-term success after modifying the machine's ACPI table. The change is invasive, firmware-specific and documented here as evidence rather than a recommendation.
- Linux 7.0 resume regression reported
An Ubuntu owner found Linux 6.17 resumed correctly but 7.0 consistently panicked or froze after extended s2idle on the same machine.
- Control center saves settings to the wrong power profile
An Arch owner found that the control-center UI did not refresh after an AC/DC change and could save an unsupported NVIDIA Dynamic Boost value to the battery profile. Rapid closure did not document a shipped fix.
- Idle NVIDIA display deadlock reported
A Fedora owner reproduced an unrecoverable display failure after idle. Integrated-only mode avoided it but also disabled outputs wired to the NVIDIA GPU.
- Omarchy hibernation workaround reported
Changing systemd from platform to shutdown hibernation avoided the reported half-resume failure.
- NVIDIA battery wakeups documented on Omarchy
One owner traced periodic dGPU wakeups to firmware notifications and an open NVIDIA driver issue.