Become a MacRumors Supporter for $50/year with no ads, ability to filter front page stories, and private forums.

Fomalhaut

macrumors 68020
Original poster
Oct 6, 2020
2,022
1,747
I've noticed that my M4 Pro Mini seems to still be quite warm even after several hours in "sleep-mode".

Looking at iStat Menus, I can see that the CPU activity & temperature peaks quite often while sleeping, e.g.

1734774868485.png


Temperatures during those periods of activity rise to about 63C, but decrease when I start using the machine, which would indicate it is a background activity. I can see that Spotlight uses quite a lot of CPU over the previous 12 hours:

1734774975898.png


But I am surprised it is still indexing after a week of usage - I am only indexing the internal SSD, which has a handful of new apps installed. I have a dock and an external SSD mounted, but have disabled indexing for this drive.

Is there an easy way to determine exactly what processes are running during these periods of high CPU activity during sleep?

Thanks!
 
  • Like
Reactions: splifingate
Settings > Screen Time > App & Website Activity

Quite a neat settings app (personally, I can monitor the Studio, iPad and the iPhone).
 
Settings > Screen Time > App & Website Activity

Quite a neat settings app (personally, I can monitor the Studio, iPad and the iPhone).

Thanks, but unfortunately this does not show the CPU activity when my Mac Mini is supposedly "asleep". It only show the actual activity when I know it is awake because I am actively using it.

It's clear that while asleep there is a lot of CPU activity (about 25-40% in each hour while asleep), and this is what I want to find out. It may be Spotlight or Apple Photos indexing, but I can't see why this takes so long given there have been only very minimal changes in the last week - with zero new photos added to the library.
 
Shut off WiFi/unplug the Ethernet cable and put it to sleep to see if it’s still doing that. Pretty sure most of the activity will stop. Let us know what you find out.
 
  • Like
Reactions: Fomalhaut
Shut off WiFi/unplug the Ethernet cable and put it to sleep to see if it’s still doing that. Pretty sure most of the activity will stop. Let us know what you find out.

This is a good suggestion; I will try this overnight.
 
[UPDATE]

I was aware of the terminal commands for getting Power Management logs using "pmset -g log", but these require a fair bit of knowledge to fully decipher.

I found a very fully-featured UI called "Sleep Aid", which presumably uses the underlying logs or power management APIs to provide information on what's going on, e.g.

1734823413758.png


I will let this run for a couple of days to see what's happening when I'm in sleep mode.
 
  • Like
Reactions: EugW
I'm getting some interesting results from "Sleep Aid". Perhaps the most worrying is this one:

1734850300991.png


I'm getting CPU core temps of up to 65C *during sleep*, which explains why it feels quite warm to the touch just before waking. It is actually cooler when I'm using it than when it's sleeping, which just feels wrong.

Many of the "wake events" are: "wifibt DataFrame ARPT", which looks like either wifi or BlueTooth is doing something to wake the computer.

My next tests will be to:
(a) disable wifi and disconnect the Ethernet cable as suggested above
(b) Disconnect my CalDigit TB3+ dock, which has a number of peripherals attached (HDDs, SSDs, an audio interface)
(c) Disable Bluetooth (a bit problematic because I use the BT keyboard or mouse to wake the machine...and presumably I'd have to use the famous "hidden" power button to wake the machine.

I wish macOS had better options to for setting a usage profile for power. Even though this is a desktop, I don't need it to be doing any indexing or analysis while it's sleeping, or for such activities to take place during a single period while its asleep (e.g. first 30 minutes) rather than every 20-30 minutes throughout the entire sleep period.
 
"Disconnect my CalDigit TB3+ dock, which has a number of peripherals attached (HDDs, SSDs, an audio interface)"

I'll guess that the dock and attached items have something to do with it.

I have the simplest of solutions for you (but I'll guess you won't take it):
Shut the Mini down at night, and reboot in the morning.
I GUARANTEE you, there will be no "CPU usage" AT ALL if you do this.

Better yet, do what I do myself:
- Mac and all peripherals plugged into a power strip/surge suppressor
- At night, after shutting down the Mini, I reach down and flip the power strip to OFF
- Next morning, I turn the power strip back on and press the Mini's power on button.
Been running my Macs this way for 37 years (going on 38), works great for me.
 
  • Like
Reactions: niteflyr
"Disconnect my CalDigit TB3+ dock, which has a number of peripherals attached (HDDs, SSDs, an audio interface)"

I'll guess that the dock and attached items have something to do with it.

I have the simplest of solutions for you (but I'll guess you won't take it):
Shut the Mini down at night, and reboot in the morning.
I GUARANTEE you, there will be no "CPU usage" AT ALL if you do this.

Better yet, do what I do myself:
- Mac and all peripherals plugged into a power strip/surge suppressor
- At night, after shutting down the Mini, I reach down and flip the power strip to OFF
- Next morning, I turn the power strip back on and press the Mini's power on button.
Been running my Macs this way for 37 years (going on 38), works great for me.
Well, this is obviously a "work-around", but doesn't really answer the question about what is going on when macOS sleeps, and doesn't suit people who have a lot of concurrent tasks on the go that need to be picked up over multiple days.

A daily shutdown/restart cycle is the computer equivalent of a "clean desk policy", where everything has to be put back into their respective files/folders/drawers/cupboards. On the one hand you start fresh each day with the set of documents that you need, but it can incur a substantial overhead just setting up everything again the next morning. If you only do one thing at a time, this might be OK. If you have 50 web-tabs, a dozen documents and different apps open , then it's not very efficient.
 
I found some good information from "Sleep Aid" over the last day.

I disabled a number of "keep awake" flags using Sleep Aid, as follows:

1734937724219.png



This *almost* stops all activity during sleep:

1734937793454.png


There are still several Power Management events that wake up the Mac to attempt "mDNSResponderMaintenance", but these last no longer than 1.5 minutes roughly every 1-2 hours.

I can presumably toggle to same switches using "pmset" flags.

My Dock and external drives were still connected throughout this period, so these don't seem to be the culprit.
 
  • Like
Reactions: VicVendrell
I found some good information from "Sleep Aid" over the last day.

I disabled a number of "keep awake" flags using Sleep Aid, as follows:

View attachment 2465020


This *almost* stops all activity during sleep:

View attachment 2465021

There are still several Power Management events that wake up the Mac to attempt "mDNSResponderMaintenance", but these last no longer than 1.5 minutes roughly every 1-2 hours.

I can presumably toggle to same switches using "pmset" flags.

My Dock and external drives were still connected throughout this period, so these don't seem to be the culprit.
Is there a way to disable WiFi and network access while sleeping from the macOS settings? I don’t like my Mac doing things and communicating with the internet while sleeping… maybe some terminal command? Anything more native than a third party app…
 
  • Like
Reactions: apostolosdt
I will share my experience with sleep on MacMini M4pro 24/512.40 degrees processor is set on the device whether it is set to auto mode or low power. It is still at 40.degrees in sleep. The only difference is in low power where it turns off the fan in sleep. Why that is, I have no idea. Best conditions for component life I'm guessing. In sleep at night the consumption is 4.7 W. I have no storage connected. I wonder if the M4 also has it so that it runs the fan in sleep.

In real life, the ASD monitor and MBPro M1pro 16" are icy to the touch in the morning, but the mini is warmer. But it's cold in the cold sense. I've had it for about 14 days, and I actually picked it up because it was the only one in stock here (Czech Republic) with a 10% discount. Everywhere waiting for other versions 1 month. I would have been happy with the basic version, but I wanted it for Christmas.
 
Last edited:
The percentage of CPU is based on the maximum percentage occupancy at the current clock speed.

Whilst sleeping the clock speed is very low.
Actually… Well, here’s one explanation:
Unfortunately Apple doesn’t appear to define what the % CPU column means, other than stating that it’s “processor capability”. As far as I can tell, the figure given is the total of ‘active residency’ for all cores, meaning the percentage of core cycles that aren’t spent in idle. So if a core were completely idle, active residency would be 0%; when no idle cycles were recorded, it would be 100%.

This is the total for all CPU cores, so if your Mac has eight cores, the scale maximum for % CPU is 800%. Furthermore, when an Intel core uses hyperthreading to boost its performance, scale maximum for % CPU doubles: with eight cores in full hyperthreading, that comes to a total of 1600%.
Since Apple Silicon doesn’t use logical cores (e.g., Hyper-Threading), it’s just 100% per physical core.

It's clear that while asleep there is a lot of CPU activity (about 25-40% in each hour while asleep), a
So, if iStat (or whichever utility) is reporting the percentage the same as macOS, the UNIX way, then that usage is really 40% of 1200% — assuming the 12-core M4 Pro — or a total usage of just 3.3333333333% of the entire CPU’s capability.

By the way, there’ve been discussions about this before:

 
Actually… Well, here’s one explanation:

Since Apple Silicon doesn’t use logical cores (e.g., Hyper-Threading), it’s just 100% per physical core.


So, if iStat (or whichever utility) is reporting the percentage the same as macOS, the UNIX way, then that usage is really 40% of 1200% — assuming the 12-core M4 Pro — or a total usage of just 3.3333333333% of the entire CPU’s capability.

I think you mean "entire system's capability". It is 25-40% of one (clock-reduced) CPU which I agree is a fraction of the system.

However, I still find 3.33% system usage quite high. Normally I expect to see 99.5% of CPU power available on average when idle. I certainly don't expect to see >3% used when asleep. When asleep, I expect the system to pause every process except just those needed to wake the computer. Otherwise it's not sleep mode it's conscience sedation mode.
 
  • Haha
Reactions: Populus
I think you mean "entire system's capability". It is 25-40% of one (clock-reduced) CPU which I agree is a fraction of the system.

However, I still find 3.33% system usage quite high.
To be slightly more thoroughly fair, the CPU usage is ~2.1% to ~3.3% (25/1200 and 40/1200).

Anyway…

I get your argument. Sleep state should be extremely minimal. To that point, I see my Mac mini uses about that same ~3-4% CPU usage (as reported by Activity Monitor for combined System and User) when idle (i.e., no mouse or other user input whatsoever). In other words, “Sleep” seems more of low-power standby state vs sleep. Perhaps Apple should consider a terminology change — I know that won’t happen.

To throw my hat in the ring regarding Mac sleep issues… My recent experience was after adding another external SSD drive. When I had all four SSDs, two OWC 1M2 and two compact USB 3.x NVMe drives, connected to the rear ports, the mini would not sleep properly (including stay very warm) with improper, unexpected disk ejects. When I disconnect (or rather connected to one of the front ports) one of the USB 3.x enclosures, the Mac went back to properly sleeping err standby.

Since then, yes, it does the typical(?) Power Nap behavior of awakening briefly — maybe <5 minutes — every (I guess) 30 minutes or so to perform whatever routine tasks as apparent by the external drive lights being active.

Would it be better if these routines were less? Probably. However, I don’t feel the normal “sleep" behavior concerning or annoying enough to do a deep dive.
 
To be slightly more thoroughly fair, the CPU usage is ~2.1% to ~3.3% (25/1200 and 40/1200).

Anyway…

I get your argument. Sleep state should be extremely minimal. To that point, I see my Mac mini uses about that same ~3-4% CPU usage (as reported by Activity Monitor for combined System and User) when idle (i.e., no mouse or other user input whatsoever).

For reasons unrelated to this I recently did a test across Mojave, Catalina, and Monterey and CPU time was >99.5% at idle on a 4-core Intel (i.e. a small fraction of an M4...) once the system has settled after boot (which in some cases required >25 minutes). Sonoma and Sequoia tests to come.

However, note you can't use Activity Monitor to measure this because Activity Monitor interferes with its own measurements. What I used was:

top -u -s 60 -S -l [B]#[/B] > top.log

where # > 2 (1st snapshot is always unreliable and 2nd one varied but 3+ showed the steady-state).

In other words, “Sleep” seems more of low-power standby state vs sleep. Perhaps Apple should consider a terminology change — I know that won’t happen.

Yeah I won't hold my breadth either...

To throw my hat in the ring regarding Mac sleep issues… My recent experience was after adding another external SSD drive. When I had all four SSDs, two OWC 1M2 and two compact USB 3.x NVMe drives, connected to the rear ports, the mini would not sleep properly (including stay very warm) with improper, unexpected disk ejects. When I disconnect (or rather connected to one of the front ports) one of the USB 3.x enclosures, the Mac went back to properly sleeping err standby.

Since then, yes, it does the typical(?) Power Nap behavior of awakening briefly — maybe <5 minutes — every (I guess) 30 minutes or so to perform whatever routine tasks as apparent by the external drive lights being active.

Note I have Power Nap disabled on my computer but my impression from this thread is that is not sufficient with the latest OS/hardware to keep this newborn in deep sleep throughout the night?

Would it be better if these routines were less? Probably. However, I don’t feel the normal “sleep" behavior concerning or annoying enough to do a deep dive.

I just want my computer to gets its full 8 hours beauty sleep! Just kidding but I would prefer the computer to stay asleep if I put it to sleep. If the computer doesn't have its own bedroom and has peripherals connected, coming out of sleep could cause those peripherals to make noise. Also like the OP I would prefer my computer not go on the Internet or update itself without my knowledge.

Understand others prefer the fully-automated, "appliance-like" experience, but I would say there at least a) needs to be full transparency about what it is doing when you're not looking and b) it should be possible to disable/control. I certainly wouldn't expect it to do anything while asleep that would cause it to get hot.

P.S.Just to be clear that I fully appreciate that as UNIX-based system, Macs need to do some periodic maintenence. I just don't think that should override sleep. And at this point surprised this maintenence puts any material load on a system that must be 50x faster than the first systems to run Mac OS X.
 
My M4 Pro seems to wake up as well after some time. I tried Sleep Aid, though I didn't really understand much on how to use it. I ended up turning off all the check boxes in the settings, and the system didn't wake up, but then the app kept giving me warnings about keeping stuff turned off on exit, so I ended up reverting back to default settings. I'm assuming it's something with my network settings or Apple Mail app.
 
Register on MacRumors! This sidebar will go away, and you'll see fewer ads.