Crackling should not happen at all. What have you done before your tests? Any special sauce setup? I would do such tests on a pristine installation in its default state.
I agree, but at this point I think the finger is to point at ASUS and something with their firmware and how the laptop handles āstressā.
So my latest tests are on a fresh Ubuntu install, the only thing I did was add KDE (X11). Nothing else was changed. I just started using my laptop normally. I also tried updating the quantum after noticing crackling, but there was no change so I undid it (I created both a profile I think itās called so a separate file, and tried changing it via the pipewire config file).
Those changes have been undone, so I think itās less to do with pipewire (based on behavior and not seeing anything in pw-top or other places) and more to do with some weird laptop power distribution thing.
Itās worth mentioning that my specific laptop was a victim of ASUSā poor bios engineering and was really prone to crackling on Windows as well before a reddit user found the issue and it was such a big thing that it forced them to release an update.
Which begs the question how you confirmed that the change was applied. pw-top should be consulted for that.
You have installed said update, I take it?
Also, letās be clear that we are on the same page about what constitutes crackling. If it happens midstream, while the device cannot possibly enter a suspend/powersave mode, thatās what I call crackling (xruns). If it happens at playback start/stop, itās probably more related to suspend/powersave somewhere in the chain (USB, audio codec, pipewire node); often referred to as āpoppingā, when the analog output stage gets toggled on/off. To check if itās the latter, this should keep the chain alive:
pw-play - </dev/zero
(run it in a terminal and it will play perfect silence on the default sink/output)
Iāve been following this thread from the start, and honestly I think the discussion has been going in the wrong direction too focused on PipeWire and GPU load as a software problem. After digging deeper into it, Iām pretty convinced the root cause is something else entirely: a well-documented ACPI firmware bug sitting in the BIOS of 2021ā2024 ASUS ROG laptops, including the G18 line.
Thereās a solid community investigation documented on GitHub by a user called Zephkek who reverse-engineered the ACPI tables from multiple ROG laptops across generations. The short version: the BIOS firmware has a broken GPE handler (_L02) that keeps firing in a loop, and inside it thereās code calling Sleep() inside an interrupt context. The result is that a single CPU core gets held hostage for bursts of time. This isnāt a driver or OS issue switching to Linux doesnāt get you out of it either.
The chain roughly goes like this:
Firmware triggers ACPI GPE handler (_L02) every ~30ā60s
> High-priority interrupt polls dGPU power state unnecessarily
> Poorly optimized AML code seizes a single CPU core
> DPC latency spikes 1000µs+
> System-wide lag, audio crackling, frame drops
That nvidia-smi output showing 100% GPU load and 86°C during the crackling - thatās a symptom, not the cause. Whatās actually happening is the system scheduler getting disrupted by those ACPI interrupt storms, starving the audio threads of CPU time.
Update your BIOS
As of December 2025, ASUS shipped a UEFI update that rewrites the ECLV routine - removing the sleep/polling behavior and the self-rearming mechanism behind the stalls. Thatās the real fix, addressed at the source.
To do it: boot into Windows temporarily (or use a USB with the BIOS flasher utility), head to rog.asus.com, search for your ROG Strix G18 (G814), download the latest BIOS, and flash it. One thing worth knowing ASUS confirmed in September 2025 that flashing their BIOS updates wonāt void your warranty, so thereās no risk there.
The initial beta rollout in Sept/Oct 2025 only covered the Scar 15 and Zephyrus M16, but the wider rollout has been ongoing. Check your specific G814 support page, if the BIOS version is dated late 2025 or newer, it very likely includes the ECLV fix.
In the meantime: NVIDIA driver tweak
While youāre waiting on that, thereās an NVIDIA module parameter that can reduce unnecessary GPU power cycling on ROG laptops under Linux, which may take the edge off the symptoms. Create or edit /etc/modprobe.d/nvidia.conf and add:
options nvidia_drm modeset=1
options nvidia NVreg_EnableGpuFirmware=0 NVreg_EnableS0ixPowerManagement=1 NVreg_DynamicPowerManagement=0x02
NVreg_EnableGpuFirmware=0 disables GSP firmware on the NVIDIA side - on ASUS ROG laptops specifically this can interfere with power state transitions and make GPU power cycling erratic. Then run:
sudo update-initramfs -u
sudo reboot
It wonāt eliminate the ACPI stalls, but it can reduce how aggressively the GPU power states fluctuate, which seems to be one of the things that makes the ACPI handler fire more frequently under load.
On the external display lag
Thatās probably a separate issue with how Advanced Optimus / the MUX switch handles external output under Wayland. Itās been reported on several 2023 ROG models. Quite a few users mentioned the display stutter also improved after the BIOS fix, so Iād try that first before going down a separate rabbit hole for the display issue.
The problem almost certainly isnāt PipeWire or GPU thermals in isolation , itās a known firmware-level ACPI bug hitting the whole G18/Strix/Scar line. Update to the latest BIOS for your G814, slap those NVIDIA params in as a stopgap, and let us know how it goes.
Thank you! Iāll try these today. For the bios update, I was running version 333, so I believe post those fixes, but I actually see they put out a new version literally last week. Iāll try those and update this thread!
OK, I return with an update.
I updated the BIOS but nothing changed (running 334 now), I did the NVIDIA tweak and that seemed to have helped alongside setting the Nvidia X Server to Performance mode (I did switch between drivers and moved back to the latest -open one and the lag seems to have disappeared from my desktop.
The crackling continues but appears to be minimal. No changes to PipeWire seem to affect it so itās definitely something on ASUSā side. Honestly for now I am quite happy even with the crackling since it only happens when Iām in more demanding videogames/pushing the IO R/W to itās limit (which seems to be related to making the crackling worse whenever thereās a lot of read/write operations going).
I genuinely want to thank everyone in this thread for taking the time, and even if itās not a perfect solution, it works for now and itās a lot better than before. In case I find a proper fix to this and manage to 100% fix it I will post in this thread and hopefully help other people with ROG laptops. Iām just happy to be able to daily drive Ubuntu on my laptop for now.
Just a janitorial note, because this thread might be closed by then, you can always contact the moderators to reopen it.
Sounds good, thanks for letting me know! I see it auto closes in 2 days.
This topic was automatically closed after 30 days. New replies are no longer allowed.