kernel_task is using 100% CPU when waking from sleep

But is 600MB out of 8GB a lot or I shouldn't be worried at all?

Doesn't strike me as being high.

What is actually the problem you are trying to solve?

I don't want to waste CPU or Memory, that's all.

Memory is only wasted if it is not being used.

Stop worrying and just use your Mac.

Radaz, I've got an MBA with 8GB, and my kernel_task stands at 930 MB. So your number sounds good, at least relative to mine.

I've had a unusual number of kernal_task ramp ups with my mid 2009 macbook pro laptop. Enough so that the screen turns blank and then I wait and then the login screen appears. This begins to happen so frequently that my laptop becomes unusable. I tried to uninstall some old sofware and it still continued to happen. Tried to go into safe mode and its still persisted.

I changed my hard drive with ML (Old backup) and it seems to work fine. (For now) This all really started to happen after I spilled tea into my touchpad and my touch pad and hard drive got fried. Replaced my hard drive and my touchpad.and backed up my system from my time machine drive. I have exactly the same 1 TB hybrid Toshiba drive and it started flipping out as it overheated. Kernel ramps ups and logouts, and black screens.

I did a PVRAM reset and a SMC reset and things still didn't work. Finally replaced the drive with my older ML Lion install and it was having the same problems. But now after my machine was off for a day, it seems to work fine with my old ML os backup. I have After Effects running in the background. and Firefox and some other apps and my fan doesn't ramp up that often and everything seems back to normal.

I do want to get back to my 1 TB Toshiba Hybrid drive that I purchased that has Maverick on it. Because i'm on a 250gb drive with Mountain Lion installed on. So what shoudl I do?

Thanks for any help!

i Have exactly the same problem. Macbook Pro 13" mid 2009.

I installed the OS completly new and blank.

No Restore from Backup, no installed third party. I made only the updates from Apple for the OS, safari, and so on.

Before i did the updates, everything was fine (10.8.0 bfor the updates). Then the problem shows up (after i did the update to 10.8.5)

so dear apple: This Problem is so long unsolved. I've a lot of trouble with apple in the last time. And i really nead your devices (cause i dont wanna use windows) but if u don't get the problem solved, i have to switch. sad but true! Cause i need a device, where i can work with. Not a device where i have to work for that this thing will run!!!

And trust me if I'll change, a lot of other people in Germany, Frankonia will change too. Cause they're my customer. and if i say, u couldn't by a mac, cause they're not able to solve the issues, then they'll switch too.

think about it. I'll be waiting for your mail or answer!!!

Hey sid0186,

I don't think Apple reads these forums, so if you want to be heard use the feedback form on their website. I cannot speak for your laptop and 10.8.5, but I know that for my mid 2013 macbook air the 10.9.2 update stopped the issue. Have you tried installing 10.9?

Hey Jan,

sorry for my bad english at first.

I'd installed mac os 10.9.2 before. An there was still the problem. cause i didn't had the problem with the audio jack. It was the Powerplug.

So now i'm back on 10.8.5. Same problem I solved it on my own way.

That's the easy way. But only do this if your familar to the hidden files!!!

Firstly, you’re going to need the model identifier of your pesky Mac:

$ system_profiler -detailLevel mini | grep "Model Identifier:"

Model Identifier: MacBookAir2,1

The kext we need to modify is IOPlatformPluginFamily.kext, you can verify it is running by-

sh-3.2# kextstat | grep IOPlatformPluginFamily

67    3 0xffffff7f81229000 0x7000     0x7000     com.apple.driver.IOPlatformPluginFamily (5.1.0d17) <8 7 6 5 4 3>

Within this kext will be another further kext- ACPI_SMC_PlatformPlugin.kext, in which each recent Macintosh model is listed with relevant instructions that the kernel uses to determine how and when to invoke the control. (Make sure you switch to root now)

sh-3.2# cd /System/Library/Extensions/IOPlatformPluginFamily.kext/

sh-3.2# cd Contents/PlugIns/ACPI_SMC_PlatformPlugin.kext/

sh-3.2# cd Contents/Resources/

sh-3.2# ls | wc -l

49

So, there’s 49 different profiles listed in this kext as of the writing of this. We simply need to move the ‘plist’ file for the model identifier we discovered earlier out of that directory, so in my case I need to move file ‘MacBookAir2_1.plist’ out of this directory (to somewhere safe), but replace my model identifier with your specific one.

sh-3.2# mv MacBookAir2_1.plist /Users/<your username>/

You can simply reboot now for the changes to take effect. We could have removed the entire kext but it’s much safer to remove the plist file for the specific model as (to be honest) I don’t know what the rest of that kernel module does. I’ve been using this ‘fix’ for a few days now and not noticed any problems at all- overall it’s much quicker and I don’t have to worry about it running out of steam. What I would recommend is that you be a bit more careful about the placement of your Mac, you don’t want to be covering the exhaust. The Mac should protect itself if it *does* reach max temperature but I would just be a little more careful.

Thanks for trying to help me out 😉

Greetz Sid

I had the same issue, unplugging my headphones helped. That was pretty ugly, please have this fixed.

Report bugs here:

https://developer.apple.com/bug-reporting/

Hi,

have the same issue. MacBook Pro Retina. I have a 27" display attached (via Thunderbolt). after i plugged in a second display (21", 1080p, via HDMI) this issue came across!!! Hope this problem is gone with Yosemite...

Same issue for me.

MBPr mid-2012, OS X 10.9.4 Mavericks, 8GB RAM, but crucially...

...Running 2 external displays: Asus 23" via MDP-VGA cable, and NEC 19" via MDP-DVI cable.

Headphone jack has no impact. Memory "cleaning" has no impact. Going onto battery rather than AC power has no impact. Shutting down Parallels (Windows 7 Ultimate, using 3 CPUs and 2GB) has not impact. I've gone through multiple OS updates. I *think* I've reset PRAM and/or SMC during the past 12+ months that CPU usage has become ridiculous, and if so, those haven't had an impact -- nothing helps.

Well, one thing helps: unplugging both monitors and going back to using only the MBPr's own screen via the integrated graphics, rather than the 3 (or 2) monitors that the discrete graphics card has been driving.

Sometimes when I then plug back in the external monitors, CPU usage stays low. Sometimes is immediately spikes right back up. And sometimes it stays low for 10-30 minutes, then spikes back up.

Note that my kernel_task CPU % in Activity Monitor CPU % table is way above 100%. I often get 1500%, although more often it's 350% to 600%. Either way, the graph shows 100%.

Another weird factor is the timing of the spike. Typically everything is just fine until early afternoon, but then the bulk of the afternoon has kernel_task spiking high. The computer is almost unusable when this is happening. I just have to move to 1 screen mode, and wait it out.

I *believe* that the afternoon is slightly warmer than the morning and evening, and there's a critical point at which the fan kicks in at max, kernel_task spikes in order to prevent real processes from taxing the CPU, and the MBPr goes into self-preservation cool-down mode. But this is of course a ridiculously draconian solution to a 5 degree ambient temperature increase. And, I've monitored the various temperature sensors on the CPUs, near the battery, etc. (using Temperature Monitor app) and the temps barely increase. So I figure something is coded improperly. But I bet it is really, really difficult to isolate, based on the reasons people give for the problem in this thread!

Good luck to all of us.

UPDATE

I tested out some different multi-monitor configurations, and have at least found a partial solution.

Scenario 1: MBPr + Asus 23" VGA + NEC 19" DVI - fails, kernel_task out of control

Scenario 2: MBPr + Asus 23" VGA + NEC 19" VGA - fails, kernel_task out of control

Scenario 3: MBPr + Asus 23" VGA + NEC 19" #2 (different unit of identical monitor, as I have 2 of them) DVI - fails, kernel_task out of control

Scenario 4: MBPr + Asus 23" VGA + NEC 19" #2 VGA - fails, kernel_task out of control

Scenario 5: MBPr + Asus 23" VGA + Acer 23" DVI - fails, kernel_task out of control

Scenario 6: MBPr + Asus 23" VGA + Acer 23" VGA - fails, kernel_task out of control

Scenario 7: MBPr + Asus 23" VGA - success, kernel_task near 0% CPU usage

Scenario 8: MBPr + NEC 19" #2" VGA - success, kernel_task near 0% CPU usage

As you can see, the problem appears to be simultaneously using 3 monitors. It is fine with just 2.

Very frustrating since the NVIDIA GeForce GT 650M 1024 MB graphics card is quite good. But one of the following may be happening with 3 monitors:

a) the GPU can heat up dramatically, requiring kernel_task to run a fake process (like System Idle Process in Windows) to prevent the CPU from overheating (this does not make sense)

b) the GPU requires a lot more resources, requiring kernel_task to run a sub-task that helps the third monitor show video (this kind of makes sense, but the fact that this only happens in the afternoon is unexplained)

c) above some level of system resource use, a critical point is reached and kernel_task goes crazy (pretty hard to diagnose this one...)

d) some process happens to be running in the background, and in combination with the third monitor being present, it causes kernel_task to go crazy (again, hard to diagnose -- and I typically have the same stuff open all the time. I even went into Windows [running under Parallels] to make sure that no processes such as hard drive defragmentation or indexing ever run)

BTW, if kernel_task IS going crazy, and I shut down and restart, kernel_task will again spike within minutes of startup -- usually, but not always, even if I have no apps loaded (other than the system processes and menubar crap that always runs, but that leaves almost no memory footprint or CPU usage).

So, this is easy to replicate, in the sense that it happens consistently. But it is very, very hard to isolate the cause. I hope Apple can figure this out. I hate having to drop back to 2 monitors every afternoon. Ridiculous.

Hi, Same CPU problem kernel task 300%. Macbook Pro 9,2 mid 2012

RAM 4 Gb

HD 500 GB

i5

New installation Mavericks 10.9.4

**** Interrupt distribution ****

CPU 0:

Vector 0x46(SMC): 7.33 interrupts/sec

Vector 0x56(HDEF/EHC1): 2.97 interrupts/sec

Vector 0x72(IGPU): 323.07 interrupts/sec

Vector 0x73(XHC1): 9.51 interrupts/sec

Vector 0x76(SATA): 7.73 interrupts/sec

Vector 0x78(ARPT): 35.85 interrupts/sec

Vector 0xdd(TMR): 995.75 interrupts/sec

Vector 0xde(IPI): 259.29 interrupts/sec

CPU 1:

Vector 0xdd(TMR): 325.45 interrupts/sec

Vector 0xde(IPI): 4195.35 interrupts/sec

CPU 2:

Vector 0xdd(TMR): 233.54 interrupts/sec

Vector 0xde(IPI): 271.96 interrupts/sec

CPU 3:

Vector 0xdd(TMR): 164.61 interrupts/sec

Vector 0xde(IPI): 4036.88 interrupts/sec

I try unload/load AppleHDA and no work, remove plist also no work.

Pls Helpme.