Omarchy Hardware

Microsoft

Surface Laptop 4 AMD

Twenty-two exact-model owners establish a workable Surface Laptop 4 AMD under Linux, but none reports running Omarchy: touch and pen remain unsupported, the custom kernel matters, and suspend, audio, brightness and input have changed with kernel releases.

22exact-model reportsReport mix: three Arch Linux · one Arch-based · 18 other Linux

Bottom line

The AMD Surface Laptop 4 can be a capable Linux laptop, but the result depends on the linux-surface kernel and model-specific workarounds. The touch screen and pen are still unsupported, and owner reports show that input, brightness, audio, external displays and suspend can move forward or backward with kernel and cable changes. None of the 22 exact-model owners in this profile runs Omarchy, so an Omarchy installation remains an informed extrapolation from Arch, Arch-based and other Linux evidence.

Plan for the linux-surface kernel, keep a known-good kernel available and test the exact paths you depend on: encrypted-boot keyboard input, Fn keys, speaker and headphone output, lid wake, overnight standby and every display adapter. The 15-inch AMD model requires the custom kernel according to the project guide. Do not buy the AMD model for touch or pen use under Linux today.

What is reported

12 findings with cited evidence
  • Install and boot

    Owners completed Ubuntu and Arch-family installations, but the path is not stock-laptop simple. The 15-inch AMD model requires the linux-surface kernel, older graphics failures were worked around by disabling IOMMU, and one apparent kernel boot regression turned out to be an unmounted EFI partition that received the new initramfs in the wrong place.

    Expect an installation plus a hardware-enablement step, and preserve a working kernel. Disabling IOMMU is a security tradeoff, not a neutral compatibility switch.

  • Desktop and graphics

    Multiple owners reached a usable graphical desktop. Early AMD owners also reproduced graphics lockups under workloads such as video calls and map rendering when the IOMMU isolation workaround was enabled; disabling IOMMU stopped the crashes for one owner, while another still saw a later freeze on kernel 5.12.12.

    The desktop is usable, but old installation advice can create its own failure mode. Follow the current device guide instead of copying an undated kernel argument.

  • Internal display and brightness

    The internal panel works, but brightness control has changed with the kernel. One Ryzen 5 owner lacked adjustment in 2023; in 2025 two owners found that linux-surface 6.17.1 reset brightness to 100% after lid wake while 6.16.9 did not.

    A working panel does not establish working brightness keys or correct wake behavior. The project says the old acpi_backlight=vendor workaround is generally unnecessary from kernel 6.2 and can break control on newer kernels.

    Mixed resultsSource 1, Source 2, Source 3
  • Keyboard, hotkeys and backlight

    The keyboard and touchpad are supported with Linux 5.13 or newer or the linux-surface kernel. Owners confirmed an initramfs-module fix for missing keyboard input before encrypted-disk login. A separate 2025 Fn-key and focus regression received a BTN_0 filter that is now included in the linux-surface Arch package, but this profile has no owner retest after packaging.

    Normal desktop typing, pre-login typing and Fn keys are separate tests. The current project package contains the Fn-key fix; that is stronger than an open patch but still not the same as a post-update owner result.

  • Touchpad and pointing

    The precision touchpad is supported, although Arch and Ubuntu owners reported intermittent keyboard and touchpad freezes in 2023 and the project shipped a related fix in later 6.2 surface-kernel packages. The original owners did not return with a same-machine retest. The AMD touch screen and pen remain unsupported in the current feature matrix.

    The touchpad is a viable primary pointer; the touch screen and pen are not. Those are different component results, even though they share the pointing row.

  • Fingerprint and biometrics

    Microsoft sold the laptop with Windows Hello face-authentication hardware and no fingerprint reader. The retained exact-model Linux reports do not establish face login.

    Treat biometric login as untested here. A working camera does not by itself establish infrared authentication support.

    Evidence, no verdictSource
  • Wi-Fi and Bluetooth

    The current linux-surface matrix lists Wi-Fi and Bluetooth as supported, but the retained exact-model field reports do not isolate everyday wireless use as a tested result.

    The support matrix is positive project evidence, not a field report of a particular radio, access point or Bluetooth device under Omarchy.

    Evidence, no verdictSource 1, Source 2
  • Speakers, microphone and audio

    The current project matrix lists the speakers as supported, but a Ryzen 7 Linux Mint owner found both the internal speakers and headphone jack detected yet silent and supplied codec data for a model-specific quirk.

    Project support and a recent failure coexist. Test speaker, headphone and microphone paths separately on the exact kernel you plan to use.

    Mixed resultsSource 1, Source 2
  • Camera

    The current linux-surface matrix lists the camera as supported, and Microsoft specifies a 720p front camera plus infrared hardware. The retained field reports do not include an exact-model video-call or face-login result.

    Camera support is plausible but not field-verified in this profile; infrared login is a separate, unestablished capability.

    Evidence, no verdictSource 1, Source 2
  • Suspend and resume

    Early owners reported blank-screen resume and no-wake failures. Two owners said kernel 5.14.9 fixed suspend power draw in October 2021, and another later reported working wake through March 2023. Fedora and Arch reports then documented severe heat or drain again; one Fedora owner restored sleep by rolling back to 6.0.12.

    Basic wake and low-drain standby have both worked on particular kernels, but neither is settled across updates. The current project matrix still marks suspend and hibernation unknown and S0ix partial.

  • External displays and docks

    Owners reported missing high-resolution and ultrawide modes and external screens turning off with the lid. In the strongest follow-up, a mini-DisplayPort-to-DisplayPort cable delivered full resolution after mini-DisplayPort-to-HDMI paths failed; another lid-close thread traced its result to unreliable adapters.

    USB-C video exists, but adapter, cable, resolution and lid state all matter. Test the complete display chain instead of treating one working monitor as proof that every dock path is sound.

    Mixed resultsSource 1, Source 2, Source 3
  • Battery, power and thermals

    Four owners reported severe closed-lid drain, including a full battery lost in two to three hours and another depleted overnight; a fifth exact owner said the powered-off drain described in one thread did not occur on their machine. Two owners reproduced 90 C operation and delayed or unexpectedly weak fan response, while one later found the fan eventually engaged under sustained load. A Manjaro update also produced a zero-percent battery reading on one machine.

    Awake battery runtime, standby endurance, powered-off drain, battery reporting and thermal control are separate questions. The measured Windows runtime shows the hardware's potential, but it does not resolve the Linux lid-close reports.

Changes over time

19 dated events
  1. S0ix resume work begins

    Patches improved basic resume, while lid events and post-resume power state still needed work.

  2. Disabling IOMMU stops graphics lockups

    An AMD owner found that force_isolation triggered graphics crashes and that disabling IOMMU restored stability; the project guide later adopted the safer of the two compatibility paths.

  3. Boot failure traced to an unmounted EFI partition

    Mounting the boot partition through fstab put the initramfs in the correct place and restored boot; the kernel update itself was not the root cause.

  4. Ubuntu installation reaches a usable desktop

    An AMD owner booted the installer with nomodeset, completed installation and applied the linux-surface workarounds.

  5. Kernel 5.14.9 fixes suspend draw for two owners

    Two exact-model owners report that the new surface kernel restores low-draw suspend; one still sees a full charge after two hours asleep. Hibernation remains a separate, untested path in those reports.

  6. Battery status falls to zero after an update

    A Manjaro owner lost usable battery reporting after an otherwise routine update; the thread does not establish a final fix.

  7. Owner reports suspend working after an update

    The owner of two AMD Surface Laptop 4 machines said suspend and wake worked after updating and reaffirmed the result in March 2023, while noting that touch still did not work.

  8. Display cable change restores full resolution

    A mini-DisplayPort-to-DisplayPort connection drove all three screens at full resolution after mini-DisplayPort-to-HDMI paths failed.

  9. Encrypted-boot keyboard modules confirmed

    Adding the Surface HID, aggregator and pin-controller modules to the initramfs made the built-in keyboard available before login.

  10. Surface-kernel packages add an input-freeze fix

    Project follow-up in the owner thread identifies a related fix in the 6.2.13-2 and 6.2.14 surface-kernel packages; neither the Arch nor Ubuntu owner returned with a same-machine confirmation.

  11. Overnight closed-lid drain returns on Fedora

    Two exact-model owners report rapid sleep drain after later updates, including a full battery depleted overnight.

  12. Rolling back to 6.0.12 restores sleep

    A Fedora owner with rapid sleep drain says several kernel rollbacks were required before sleep worked again on 6.0.12.

  13. Docked lid behavior depends on the adapter

    External screens initially went dark when the lid closed, but the owner later traced the result to unreliable display adapters rather than a settled model-wide limit.

  14. Fan and processor-frequency failure reported

    One owner saw high temperatures without expected fan response and, at other times, a low-frequency state with the fan running hard.

  15. Detected but silent internal audio reported

    The speakers and headphone jack appeared to Linux but produced no output on a Ryzen 7 configuration.

  16. Fn-key regression reaches multiple owners

    Two exact AMD-model owners report stuck Fn input and broken focus behavior. Reloading the Surface HID module helps temporarily, and one owner tests a BTN_0 event filter.

  17. Newer kernel resets brightness after lid wake

    Two owners reported the 100% wake behavior with linux-surface 6.17.1 while 6.16.9 remained a working comparison.

  18. BTN_0 fix enters the linux-surface package

    The project packages the exact-model-tested hid-surface filter in its Arch kernel. The retained owner thread has no post-package retest, so code state and field outcome remain separate.

  19. Optional Omarchy linux-surface path proposed

    An open pull request adds a linux-surface kernel option to the Omarchy installer. Testing named in the pull request covers a Surface Book 2, not this model.