In 2024, do you let your computer go to sleep or not ?

[MathWizard]

I have mine set to go to sleep after an hour, and it goes to sleep sometimes, so the 5V standby on the PSU is running, and the RAM on the mobo still gets power.

But I wonder side by side these days which PC would last longer, the one you let go to sleep, and or shut down all the time, vs the one you just leave on.

[mendip_discovery]

I still have spinny discs so I keep the main pc on 24x7. It also does some home server stuff as well.

My laptops go to sleep.

Having a smart phone these days I dont sit at the pc as much as I used to.

[PlainName]

Screen goes off after 10 mins (mainly to save replacement cost). Entire setup has power removed every night, so that's a bit more frugal than mere 'sleep'

[tom66]

I probably sleep my PC about 5-10 times per day, it's fine.  It is all SSD though.

There is probably very little difference in longevity.  In favour of sleeping would be lower thermal dissipation on components which extends their life.  In favour of keeping the PC on would be less chance of data corruption due to poorly-behaving programs, and less cyclic loads on the power supply.

I would argue the power bill favours sleeping the PC.  My PC uses about 150W in operation - at current UK on-peak rates that's 5 pence an hour.  So if I leave it running every night for a year (10 hours per night), it'll cost me £164 in electricity.  That's not insignificant, and I haven't put anywhere near that in replacement hardware in to my PC.  The last failure I had was a graphics card which was still covered under warranty.

[DiTBho]

laptop/SDD: sleep, wakeup on demand.
server/HDD: power off, wakeup on demand.
NAS/HDD: sleep, wakeup on demand, HDDs power off.

[Ranayna]

My desktop goes to sleep after 30 minutes.
When i first got my machine i had to disable sleep mode, because in 95% of cases it would not wake up anymore.
Since a windows re-install a year later (for unrelated reasons) it works totally fine.

But when i am done for the day, or will be away for longer periods, my machine gets shut down.

[rdl]

I set spinning drives to shut down in 15 mins. or so, displays maybe a hour, but never let the pc sleep. Always had too much trouble with it in the past.

[xrunner]

It sleeps but I don't let it automatically sleep. I put it to sleep myself when I leave for a while.

[DavidAlfa]

I use sleep after 5 minutes, never had any problem. Why waste energy if you're away?

[newbrain]

Main PC (NVME, SSD): sleeps after 15 minutes.
Media PC (SSD): sleeps after 60 minutes.
Second PC (NVME): sleeps after 15 minutes, when booted with Mint. Fedora gets very confused by sleep on this machine, so I need to shut it down.
TrueNAS NAS (HDD, SSD): on 24/7 - headless
Hyper-V server (HDD, SSD): on 24/7 - headless

[Sparks]

My main pc sleeps after 45 minutes. My laptop after 20 minutes. Would be less but I need to monitor things for work which updates every minute.

[soldar]

No sleep for me. Only the monitor turns off after like 15 minutes. Other than that the computer is running or I turn it off.

I have not tried sleep in many years but when I did with win XP i had some trouble. Now I am using Linux but I still do not use sleep.

[SiliconWizard]

I do use standby and practically never turn off computers anymore now, unless I'm away for more than a couple days. I don't let them go to standby by themselves ever, though, for the desktop computers, only for laptops. (I do for the displays.)
So I just put them to standby when I leave, instead of shutting them down. Avoids having to wait for rebooting and setting back your session as it was for getting back to work - it doesn't look like much, but always annoying.
Using standby also on Linux, absolutely zero problem here. But the typical number of cycles per day is 1 to 2 for desktop computers.

Unless you have many standby/resume cycles per day and you use spinning HDDs, it can only make your hardware last longer. (And even with HDDs, it would take a lot of cycles to even shorten their lifetime.)
If you have a dodgy PSU, it may shorten its lifetime a bit, maybe, but if it's a decent one, it can again only make things better compared to letting it on 24/7.
And of course there's your electricity bill.

[xrunner]

I do use standby and practically never turn off computers anymore now, unless I'm away for more than a couple days. I don't let them go to standby by themselves ever, though, for the desktop computers, only for laptops. (I do for the displays.)

A while back I was putting one to sleep and never turning it off - that was my plan. But I would see in the morning many times it was not sleeping but full on. I thought the sleep function was buggy or something ... wait it looks like the mouse has been moved? 

Oh kitty cat! He was checking things out at night and he moved the mouse. Since in the bios I had "wake on USB event" enabled he was able to turn it on. 

[SiliconWizard]

I do use standby and practically never turn off computers anymore now, unless I'm away for more than a couple days. I don't let them go to standby by themselves ever, though, for the desktop computers, only for laptops. (I do for the displays.)

A while back I was putting one to sleep and never turning it off - that was my plan. But I would see in the morning many times it was not sleeping but full on. I thought the sleep function was buggy or something ... wait it looks like the mouse has been moved? 

Oh kitty cat! He was checking things out at night and he moved the mouse. Since in the bios I had "wake on USB event" enabled he was able to turn it on. 

Cats can be an issue with it.
Note that not all BIOSes and mice act the same relative to waking up. With some mice, just moving them around is enough to trigger a wake-up event, while with others, you need to click on a button to do so. I try using mice which only do the latter. But that's something you won't know until you try...

[xrunner]

It's a Logitech (which is the only brand I have) and they have an "OFF" switch on the bottom so I tried turning it off and that did the trick ... until he learns to step on the KBD to over-ride. 

[SiliconWizard]

I'm sure the cat finds that amusing and does it on purpose once he's found out a couple times...
You should be able to disable wake-up from the keyboard (or any USB device) using the device manager on Windows.
On Linux, it's a matter of writing a udev rule.

[David Hess]

My newer Windows 10 systems work with sleep and even hibernation mode, but sometimes the mouse or network do not recover and I have to reboot to restore them, so I do not bother even on my laptop unless I have a specific reason to turn sleep on while traveling.

[Halcyon]

Mostly, no.

[Psi]

Sleep working well depends on your motherboard, some motherboards have zero issues with it, others have issues occasionally when coming out of sleep, and some are just terrible.
If you have issues with sleep you can use hibernate instead and it should work flawlessly.
With an SSD the speed to hibernate and resume is fine. It's only bad on spinning disks or if you have a butt load of ram.


With spinning disks it's best (for longevity) to
leave them spinning rather than letting them power down and up on activity + timeout.
either disable the spin-down timeout entirely or set it to something much larger, like an hour or two instead of the default, which is usually like 8-15 minutes.
The starting/stopping wears the motor, bearings and loads the motor controller with lots of inrush starting currents, and the thermal cycling isn't good on other parts either.
The feature for power up/down on activity was added for environmental reasons so they can quote a lower average power consumption.

[Ranayna]

Sleep working well depends on your motherboard, some motherboards have zero issues with it, others have issues occasionally when coming out of sleep, and some are just terrible.

Drivers can also be a huge issue.
I have an AMD system, CPU and GPU were relatively new when i got them, and the drivers apparently were not all done yet
The system went to sleep jut fine, but it could not wake up.

You hit the button, lights turned on and fans spun up, but apart from that nothing happened and i had to force power off the system.
I disabled sleep when i was not able to find a fix and got used to sleep not working.

About a year ago, i reinstalled windows, with all new freshly downloaded drivers. I forgot to disable sleep again and was already annoyed when i noticed the system asleep. But to my surprise it woke up without a hitch. And it has done now everytime, not a single crash happened.
I have not made other changes to the system when i reinstalled.

Regarding systems waking up unintetionally: My recommendation would be to disable "wake by" for almost everything. Especially wake on USB can be very finnicky with some peripherals. Give me an old school press of the power button everytime

[Jeroen3]

My pc starts doing Folding@Home when idle in winter because it keeps the room warm.
I have it set to never go to sleep, if I go afk it's probably doing something that I want it to finish.

[David Hess]

Another reason that I do not use sleep on my laptop is that it cold boots faster than it wakes up from sleep mode.  The only reason to use sleep mode is if I want to keep applications open.

As an aside, my laptop and modern desktops cold boot faster than my Android phone.  My laptop is especially fast.

[PlainName]

With spinning disks it's best (for longevity) to leave them spinning rather than letting them power down and up on activity + timeout.
The starting/stopping is what wears the motor and bearings, and the thermal cycling isn't good on other parts either.

Not disputing that but do you have a reference for it? A simple thought experiment suggests it isn't true for all cases (5 mins run time every 20 years, for example) so there should be some point where best practice changes from turning off to leaving it on.

[DiTBho]

Unless you have many standby/resume cycles per day and you use spinning HDDs, it can only make your hardware last longer. (And even with HDDs, it would take a lot of cycles to even shorten their lifetime.)

umm 

[rsjsouza]

With spinning disks it's best (for longevity) to leave them spinning rather than letting them power down and up on activity + timeout.
The starting/stopping is what wears the motor and bearings, and the thermal cycling isn't good on other parts either.

Not disputing that but do you have a reference for it? A simple thought experiment suggests it isn't true for all cases (5 mins run time every 20 years, for example) so there should be some point where best practice changes from turning off to leaving it on.

I would say it is heavily dependent on the use case. A computer that powers on/off several times a day might be a problem, as the startup current on HDD motors is much higher than nominal, according to a resident expert that used to work on this industry. However, this is certainly not true for very lightly used servers - our NAS sees very little activity (accessed once every few days) and there is no scenario where leaving it powered will extend its life, as bearings will be the dominant wear factor.

[coromonadalix]

my personal computer  is going into sleep / hybernation every day,  mostly build silent with almost nothing who spin,  nvme / ssd based

it's been 2 years now ... no problems, and i'm on a ups,  only a few times the computer has fully shutted down ...

starts in 8 seconds .... not fast enough loll   

Firefox with multiple panes always opened,   "tab saver pluggin"  "Firemin application"  just in case
Outlook ready to go

few hibernation / sleep   settings and optimization,  that's all

[EPAIII]

Sleep mode is not off.

Even power off mode is not really off.

If you want a computer to be really shut off, you need to connect it through a power strip and turn that power strip off.

Even that does not work for a lap-top. I think there you would need to actually remove the battery.

[rsjsouza]

No need to remove batteries. In a laptop you put it down by issuing a "Shut Down", not an S0 "Modern Sleep Mode" which can apply updates and bring up the computer without your knowledge. i have been using Windows 10 and 11 for quite some time now and this effectively prevents any computer activity.

[tom66]

Is motor failure of HDDs a realistic risk?  It's my understanding HDD failures typically occur for a few reasons:

a) Head touches disk surface due to excessive vibration/physical shock/bearing failure;
b) Too many sectors become irretrievable, leading to breakdown of the filesystem, possibly due to (a);
c) Disk controller failure for unknown reasons, ESD/power supply issues, random silicon death etc.

I haven't heard of motors failing.  Typically, one would expect a brushless motor to survive hundreds of thousands of hours.  The bearing is a risk, but I think it should be well established now how to make these bearings last a very long time given the precision elsewhere used in hard drive manufacture.

Spinup/spindown shouldn't correlate strongly with HDD life in "normal" applications (shutting down disks on idle) IMO.

[DiTBho]

Is motor failure of HDDs a realistic risk?  It's my understanding HDD failures typically occur for a few reasons:

a) Head touches disk surface due to excessive vibration/physical shock/bearing failure;
b) Too many sectors become irretrievable, leading to breakdown of the filesystem, possibly due to (a);
c) Disk controller failure for unknown reasons, ESD/power supply issues, random silicon death etc.

I haven't heard of motors failing.  Typically, one would expect a brushless motor to survive hundreds of thousands of hours.  The bearing is a risk, but I think it should be well established now how to make these bearings last a very long time given the precision elsewhere used in hard drive manufacture.

Spinup/spindown shouldn't correlate strongly with HDD life in "normal" applications (shutting down disks on idle) IMO.

Yup, exactly as I have already written something detailed in two topics here on this forum and have been ignored 

[Psi]

Motor failure is the motor driver IC, rather than the motor itself.  The motors do die, but motor controller death is far more common.

With spinning disks it's best (for longevity) to leave them spinning rather than letting them power down and up on activity + timeout.
The starting/stopping is what wears the motor and bearings, and the thermal cycling isn't good on other parts either.

Not disputing that but do you have a reference for it? A simple thought experiment suggests it isn't true for all cases (5 mins run time every 20 years, for example) so there should be some point where best practice changes from turning off to leaving it on.

That is true,  but in a typical use case when you have the feature enabled you end up with many more on/off cycles than you would have had with the feature off.
Perhaps i should have added "for a typical use case" to my previous comment.
There is also the annoyance of the ~2 second delay when you click on the drive in windows explorer and it locks up waiting for the drive to finish starting and become available for access.

Sorry I dont have any reference, 

[PlainName]

Aren't most things prone to failure when powered up? So not turning them off makes sense on that basis.

But... electrolytics don't have a finite life, so leaving them on may wear them out. And, probably, any issue with those would be noticeable when starting up whereas the same fault wouldn't cause a shut down once working. Similarly, disks may be borked but limping along, and it's only when you try to cold start them that they actually show how bad it is. So perhaps it's not that clear at all.

[rdl]

With spinning disk drives, how long they last is surely related to how they are used. I have a server for archival purposes, it can be days between times when it is accessed. Sometimes even as long as a month. There is no way I can believe that leaving the drives running constantly will make them last longer than if they are shut down when not needed.

[mcarp]

5 computers on 24/7, 2 of them since 2008.  those 2008 machines have both had 2 psu replacements.  laptop had a wall wart replacement, its but up for about 10 years. battery died, not replaced.  2 i7s just installed christmas of 2022. no drive or other components have gone bad except for 1 network adapter built in got flakey and i put in an expansion card to replace that one. also on 24/7 for the past 25 years a theater 5.1 channel amplifier and speakers. oh, i had 2 17 inch very very old acer monitors that went bad, caps replaced, both went bad again, but those were whew, er, before the 2008 computers.

[Psi]

With spinning disk drives, how long they last is surely related to how they are used. I have a server for archival purposes, it can be days between times when it is accessed. Sometimes even as long as a month. There is no way I can believe that leaving the drives running constantly will make them last longer than if they are shut down when not needed.

Yep, how many times the data is being accessed each day is the key.

They could be doing spin up/down/up/down/up/down 100 times a day, or once a day. etc.. depending on what your doing.
My only real point was that excessive spin-up spin-down is not good, and IMHO the default time to spin-down is way to short, especially on some drives like WD green's.

[coromonadalix]

just got an  usb3 to 8 drives storage, 8x power switch for all hdd if needed,  and a main power switch 
thermally controlled fan,  spin very low if no spinning drives and will ramp up if needed

the spinning drives after some time +/-15 min,  will spin down automatically,  you access them  and they will spin up ...   

but for sure,  when all 8 are not used, i do a full shutdown.