Why my Asus laptop kept rebooting after sleep and how to fix it
idiallo.com
idiallo.com
One of them was an HDMI audio driver, which wasn't completely a surprise given I connect to a KVM and I imagine something was weird about the connection (and I don't listen to audio through my monitor anyway)
The other was the realtek/intel driver for audio as well? (updated those) Never have had a problem again. Felt pretty good. Windows laptops are pretty awful about coming out of power save, but works like a champ now.
And generally, if you check online, you'll see plenty of reports about problems with MediaTek. I've learned my lesson - just avoid MediaTek in general. Intel adapters seem much less problematic and better supported overall.
This was recently demonstrated by this ultra wide range receiver hack which directly sampled the data from the modem ADCs.
I replaced my Asus just because of this problem.
Asus support just couldn’t help or didn’t care. They blamed the fact that I’d upgraded the RAM because it was easier to blame that than their one shitty component driver.
The worst thing was that because the notebook doesn’t shut down as expected it suddenly becomes a fire risk. I put mine in a laptop sleeve in a cupboard and it when into this loop, refused to shut down and got hotter and hotter. By the time I noticed it, it was literally smoking.
I switched to a Dell XPS15. It’s not without its issues either.
As a Mac user, I thought this could not exist. I mean, it is 2026 and a PC still does not have a decent collection of keyboard drivers.
It's annoying. Back in the mists of time when the laptop was new, it all worked fine. It's the updated Intel drivers that are broken.
I have the hope that Lenovo acquiring Phoenix will fix some of the awfulness.
There is a guide on Reddit, "How to enter advanced BIOS options on Gen 1 T15(g/p), P15(v), P17" which also applies to each generation of P16. It's possible a setting in there will finally cure your laptop's insomnia?
Back in the day the old thinkpads would just immediately shut down CPU and fan together and that was it.
Modern standby is completely crap.
(I also thought S0ix was supposed to fix all of this ...)
What I think I discovered after years of internet searching for it, is that it is related to the cpu sleep state that are available.
Before, when going to sleep, the cpu was really going to sleep. Kind of shutting the power off. In that state, the laptop would not wake up for random events at random times. But nowadays, the only cpu state that is offered by the cpu for sleep mode is one where the cpu is still running, but in very low power/activity mode. Like "simulating" the sleep. But so, from the motherboard point of view, the cpu is up and running.
And so, because of that, you could easily have random events that wake up the laptop when it is closed and then it will start to do nothing but heat up as closed and probably in a bag. Fan might spin, without an effect as in a confined space, but so the battery exhausting very fast. Until the laptop run out of battery.
For event I think that it might be something like a fake touch event on the touchpad because of movement of the laptop, screen touching when compressed or things like that; a little bit higher temperature sensor reading because of high outside temperature, or indeed a device driver that is not completely off like a wifi card that will trigger events.
Where it was the most confusing is that for the half a decade old laptop, it has both mode but recently linux distributions started to favor the cpu "fake sleep" as sleep mode by default instead of the "real sleep" that was used previously. Switching the config to "real sleep" appears to solve the problem but new laptops doesn't have this mode available anymore.
Yeah... maybe it's because i'm a video gamer, but i always save and save often.
If you can live without it, remove as much energy savings as possible on a laptop. It saves you from lots of troubles.
On my current iteration (Lenovo P1 with Intel AX211), the failure also requires some other device to try to talk to the laptop while it's being put to sleep. Anecdotally sleep tends to fail harder/more often on public networks, perhaps because there's more adversarial traffic scanning computers and waking the device up.
In the end, I really don't even understand why anyone would want their wifi driver responding to network packets when the device is asleep - given that these chips have DMA access via PCI to host memory (unless the IOMMU is configured, which is not a given) - this feature is equivalent to a persistent, always-on remote back door into your RAM, even when your machine is "off". Of course this requires an exploitable CVE in your wifi stack, but there is a constant stream of CVEs and obscure "remote management features", at least in the Intel drivers.
My solution in the end has been to just remove the wifi module from the motherboard (or in cases when it's soldered down, disable the driver by deleting it entirely) and then use an external USB wifi dongle. It's been almost a year now and no sleep issues, no more randomly heating up my backpack, and in public environments like e.g. hotels I don't worry as much because I unplug the wifi card from the laptop when I'm done with it. I had to go with the nuclear option because some updater somewhere is pretty eager about discovering the presence of the wifi device and "helpfully" re-enabling it, I suspect it's the Lenovo Vantage tool.
The external USB device is slightly inconvenient but at least I don't get woken up in the middle of the night by my laptop fan spinning up mysteriously anymore (OK not entirely true, the malware scan will trigger at night and spin up the fans but at least I know what's causing that...).
Perhaps it's time to ask an LLM of your choice to make one.
I suppose the code signing cert from the eMudhra fuck-up could work if you can prevent Windows from taking in the recovation, but that opens up a whole different can of worms.
Sometimes I would investigate but the monotony of tracing things in the Event Viewer and analysis was too much for my ADHD-addled brain. Enter Claude CLI and working through it with prompts plus allowing Claude to execute PowerShell to diagnose was painless and resolved the problem.
40+ years of laptop computing and sleep is still a problem !!!
All of the hotbagging nonsense where a laptop cooks itself in a bag seems to be caused by Windows trying to be clever and deciding there's some really important stuff it needs to do, so it wakes up the machine, does a bunch of stuff, and doesn't get back to sleep.
Windows, you're not as clever as you think you are. None of this work is urgent. Just stay asleep until I wake you up. Seriously, just don't.
But Windows isn't the only one to blame. Macbooks do the same (they're less bad because they boot so fast, turning them off completely isn't that painful).
I know because that machine was loud so I got woken up whenever it was doing that.
Microsoft of course decided that they had to copy this stupid decision, but using a timer interrupt to wake up and then sleep again was too easy, so now we have modern standby which means I must shut down my machine if I am not sure when I'll use it again, because the battery drains fast.
Now I can no longer leave my work laptop in standby over a long weekend because the battery will be completely dead.
... and that would be the day humanity knows AGI was achieved.
Even worse is the fact that support for S3 sleep is going away, so if that did work for you, you'll be pushed over to using S0 sleep, which also doesn't actually solve this problem to any significant degree.