I installed Manjaro on a new Lenovo Thinkpad T14s 2-in-1 gen1. It works pretty well, apart from some minor display problems, but a major problem I could not solve is that the power consumption when in suspend mode is much too high.
When I close the lid, the laptop goes to suspend mode all right, the power led blinks and everything looks good. But it still drains a lot of battery - much more than I had on any previous machine. I did not measure it precisely, but the order of magnitude is several percents of battery per hour, where on my previous laptop I would hardly notice any difference after 24 hours of suspend.
Currently it seems to only supports s2idle:
$ cat /sys/power/mem_sleep
[s2idle]
I could not find any setting in the BIOS to change suspend mode.
I’m not sure where to go from there. Surely I’m missing something? I cannot believe they just killed the “real” suspend mode on this laptop without any similar alternative?
You cannot do much. This is the new normal, thanks to Microsoft which commanded the OEMs to switch to this sh…thing after releasing Win10.
It practically only turn offs the screen. In some cases it can put some components in lower power state, but that is heavily bios dependent. And this is closed source so all things ACPI etc. Will always be somewhat buggy under linux.
My experience is, it usually drains almost as fast as in wake but idle with screen turned off (with ssd).
If it however, drains more percentage in “sleep” as in wake idle and the laptop becomes hot, then there is some problem, ergo, something actively keeps the cpu busy. There was a recent topic with such symptoms but i do not remember if there was a specific solution and i think it was another laptop brand.
I had a Thinkpad S531 - long ago, probably around 2014 or so. Anyway, I can recall it because it drained the battery on suspend. I never solved the issue. Don’t know whether it is a Linux vs Windows thing - I bought the Thinkpad without OS and back then I ran openSUSE.
I feel I’m back twenty years ago, when suspending a laptop on Linux was an adventure in itself. I hoped that these times were over
I’ve been using Linux on Thinkpads for decades (literally) and never had such a problem. It seems “progress” is not always going in the right direction, I’m afraid…
Hmm, my 4 years old thinkpad P1 gen i4 (intel only, no NVidia) with Manjaro has a bios bug related to ACPI never solved by Lenovo, many complain on web regarding Lenovo for Linux on that machine…
This was probably done to give Windows-Victims™ the false impression that Windows was actually starting up, resuming, or waking really fast.
A bit like the TVs from the past which had “Instant-On”; this kept the CRT heater lit (and the valves, in the case of pre-transistorized or hybrid sets) - not necessarily at full power - and at least some of the circuitry running. I had one, but at least that “feature” could be bypassed. All that un-seen power drain!
Of course, M$ didn’t care, or maybe just didn’t realize, that this would seriously impact battery life.
Starting in the 90s, it was hard to find good laptop support. The Thinkpad T-series was always good there. Nearly guaranteed Linux support with IBM developers writing the kernel modules before the laptops release… Well into the new millenium!
Nowadays, you have to do your research, especially with some of the newer junk. It’s not like Lenovo ruined Thinkpads, it just seems nearly every manufacturer does it.
But it is easy to spot these budget laptops just by reading the specs. (The sleep modes you may have to dive a tad further.)
The kernels it: S2Idle
Microsoft calls it: Modernstandby!
Intel & others call it: S0ix
I call it something else entirely that would be censored.
We want devices that can only achieve zen in the happy Sleep (S3) state.
---Check S2idle path S0ix Residency---:
The system OS Kernel version is:
Linux may 6.12.77-1-MANJARO #1 SMP PREEMPT_DYNAMIC Fri, 13 Mar 2026 18:47:31 +0000 x86_64 GNU/Linux
---Check whether your system supports S0ix or not---:
Low Power S0 Idle is:1
Your system supports low power S0 idle capability.
---Check whether intel_pmc_core sysfs files exit---:
ls: cannot access '/sys/kernel/debug/pmc_core': No such file or directory
The pmc_core debug sysfs file is empty on your system.
Isolation suggestions:
Please check whether intel_pmc_core driver is loaded.
The intel_pmc_core sysfs missing will impact S0ix failure analyze.
---Judge PC10, S0ix residency available status---:
cat: /sys/kernel/debug/pmc_core/substate_residencies: No such file or directory
grep: /sys/kernel/debug/pmc_core/substate_residencies: No such file or directory
Test system does not support S0ix.y substate
Turbostat output:
15.011551 sec
CPU%c1 CPU%c6 CPU%c7 Pkg%pc2 Pkg%pc3 Pkg%pc6 Pk%pc10 SYS%LPI
1.50 80.69 16.00 91.61 0.00 0.00 0.00 0.00
1.02 96.99 0.00
0.58 97.89 0.00
0.93 97.59 0.00
1.13 97.48 0.00
1.30 97.21 0.00
0.84 97.75 0.00
0.88 97.43 0.00
0.70 97.87 0.00
2.67 0.00 94.88 91.61 0.00 0.00 0.00 0.00
2.67
1.31 0.00 97.13
1.31
3.00 94.27 0.00
2.62 93.81 0.00
CPU Core C7 residency after S2idle is: 16.00
CPU Package C-state 2 residency after S2idle is: 91.61
CPU Package C-state 3 residency after S2idle is: 0.00
CPU Package C-state 6 residency after S2idle is: 0.00
The system does not support the Pkg%pc8.
CPU Package C-state 10 residency after S2idle is: 0.00
S0ix residency after S2idle is: 0.00
cat: /sys/kernel/debug/pmc_core/substate_residencies: No such file or directory
(standard_in) 1: syntax error
Your system achieved PC2 residency: 91.61, but no PC8 residency during S2idle:
---Debug no PC8 residency scenario---:
modprobe cpufreq_stats failed
Loaded 0 prior measurements
Cannot load from file /var/cache/powertop/saved_parameters.powertop
File will be loaded after taking minimum number of measurement(s) with battery only
RAPL device for cpu 0
RAPL Using PowerCap Sysfs : Domain Mask d
RAPL device for cpu 0
RAPL Using PowerCap Sysfs : Domain Mask d
Devfreq not enabled
glob returned GLOB_ABORTED
Cannot load from file /var/cache/powertop/saved_parameters.powertop
File will be loaded after taking minimum number of measurement(s) with battery only
Leaving PowerTOP
Turbostat output:
14.884573 sec
CPU%c1 CPU%c6 CPU%c7 Pkg%pc2 Pkg%pc3 Pkg%pc6 Pk%pc10 SYS%LPI
1.34 81.09 16.00 92.28 0.00 0.00 0.00 0.00
0.45 98.08 0.00
0.76 97.29 0.00
0.71 97.92 0.00
0.60 98.02 0.00
1.02 97.55 0.00
0.86 97.77 0.00
0.75 97.73 0.00
0.45 98.15 0.00
1.97 0.00 96.04 92.28 0.00 0.00 0.00 0.00
1.97
2.34 0.00 95.97
2.34
2.29 95.58 0.00
2.19 95.02 0.00
awk: cmd. line:1: (FILENAME=- FNR=1) fatal: attempt to access field -1
(standard_in) 1: syntax error
Your CPU Core C7 residency is available: 16.00
(standard_in) 1: syntax error
Your system CPU Core C7, GFX RC6, PCIe Device D state and Link PM state are OK,
but still did not achieve PC8 after powertop --auto-tune
lsmod says intel_pmc_coreis loaded.
Apart from that I struggle to make sense of this information. Does that suggest you all any idea?
On the spec page you can see the processor. So you have one of these 4:
Intel® Core™ Ultra 5 225U Processor (E-cores up to 3.80 GHz P-cores up to 4.80 GHz with Turbo Boost, 12 Cores, 14 Threads, 12 MB Cache)
Intel® Core™ Ultra 5 235U Processor with vPro® (E-cores up to 4.10 GHz P-cores up to 4.90 GHz with Turbo Boost, 12 Cores, 14 Threads, 12 MB Cache)
Intel® Core™ Ultra 7 255U Processor (E-cores up to 4.20 GHz, P-cores up to 5.20 GHz with Turbo Boost, 12 Cores, 14 Threads, 12 MB Cache)
Intel® Core™ Ultra 7 265U Processor with vPro® (E-cores up to 4.20 GHz, P-cores up to 5.30 GHz with Turbo Boost, 12 Cores, 14 Threads, 12 MB Cache)
It doesn’t really matter which you have because, you can stop reading after Intel® Core™ Ultra.
None of them support S3.
In some cases it is harder to find out, since it does come down to the manufacturer. But this one is easy to spot.
Easy to spot… if you know that suspend capabilities depend only on the processor model, and you know that Intel Core Ultra don’t support S3. Once again, AFAICT Intel’s specs (Sorry, I’m not allowed to include the link in my post) don’t mention anything about power states, S2idle, suspend or S0ix.
Now I know. Thanks for your explanations!
So S2idle it will be. Now the Arch wiki says that this state provides “identical energy savings but a drastically reduced wake-up time”.
I’ll have to investigate further to see if anything can be done to achieve this…
I’m sure I’ve looked at every post in this topic, but I can’t seem to find the detailed system information that should be provided when requesting support in this forum.
Did you run it as root? ls can’t access that folder on my system without root. (and the instructions for that tool say to run it as root)
Your system only has s2idle as an option, which unfortunately is not a true sleep state (System Sleep States — The Linux Kernel documentation). So to solve your problem of too much power consumption, I really think your only option on this hardware is what I suggested earlier.
Which will use zero power and by the look of your hardware should be plenty fast enough since you have a NVMe drive. You’ll probably want to increase your swap partition though to make it at least the size of RAM.
That is the workaround i use. If i know i will pause for an hour or two, i clap the screen, a.k.a. pseudo standby and lock. If i have too many windows open and know i will resume the work session next day - hibernation. The OS pops up in less than a minute, including the VM if one is open.
In all other cases - shutdown. Boot time is about 20 seconds for Manjaro.
I do not believe hardware with S3 is produced anymore, since at least couple of years, thanks to the monopoly of Microslop. When they say, “if you want to sell laptops with Win, you make it s2idle only” that pretty much settles it for the major oems.
If anyone knows how is it by the pure linux manufacturers, like system76 of Framebook, i would gladly take ideas for my next machine. But the major OEMs are s2idle only (i have a lenovo now).
A newer kernel would also be a good thing to try, especially for something like this. (And being a newer system, architecture, etc.) The latest LTS kernel is 6.18 (but try even newer).
