MacOS Sierra / kernel_task problem

I installed Mac OS sierra yesterday night in my MacBook Air13" (4gb) I also use an external SSD (Samsung T3).


Today I could not do any task because of CPU and ram usages.


First, I thought this is because of new OS, but then I check activity monitor and saw kernel_task use around 50-70% percent of CPU.


I wonder, what it is and how I can fix this problem (I found some entries in the forum, but I am not sure this is the same problem or not).


Thank you.

Replies

I'm having the same problem on MacBook Pro (Retina, 15-inch, Mid 2015),

16 gb ram, 512 gb hd

AMD Radeon R9 M370X 2048 MB

Intel Iris Pro 1536 MB

I'm having exactly the same problem on my MacBook pro (Retina, 15-inch, Mid 2014),

OS Version: 10.12 (16A323)


When I plugin a second monitor using DP/DVI adapter (I tried multiple DP/DVI Adapters that worked before) after about 5-10 minutes kernel_task starts to eat up all CPU power until its unusable, heating up like crazy with 100% fan speed, if I plug out the monitor it returns to normal again.


Using an HDMI/DVI Adapter with the same monitor it works just fine.


I hope this is patched ASAP, I can only use one monitor in the time being.


What I already tried to resolve the issue:

- VRAM reset

- PRAM reset

- SMC reset

- http://blog.viktorpetersson.com/post/136535061619/how-to-fix-kerneltask-cpu-usage-on-el-capitan

Little update on my issue, this seems to have fixed it:

http://apple.stackexchange.com/a/164947/206764

I found a fix.....


  1. Disable kext by renaming it
    cd /System/Library/Extensions/IOPlatformPluginFamily.kext/Contents/Plugins sudo mv X86PlatformShim.kext X86PlatformShim.kext.disabled
  2. Clear kext cache (not sure if this is needed)
    touch /System/Library/Extensions/
  3. Restart

After installing OSX updates you may need to repeat the above procedure if the updates have re-created the

kext

Hi Guys, Same issue here, macbook pro late 2013, macOS Sierra 10.12.1, after connecting 2 thunderbolt displays 27 inch, kernel task goes 500%. So, after reading to many posts, I finally got it. Just bought this, https://www.amazon.com/Performance-Tool-W50063-Garage-Blower/dp/B00FJ5LBOA/ref=sr_1_1?ie=UTF8&qid=1477681670&sr=8-1&keywords=sopladora and open the macbook pro with this https://www.amazon.com/Mudder-Pentalobe-Screwdriver-Macbook-Retina/dp/B01FTO97Z0/ref=sr_1_2?ie=UTF8&qid=1477682286&sr=8-2-spons&keywords=macbook+pro+screwdriver+kit&psc=1. Then toons of dust all over the fans, now they seem to be like new, kernel_task 3.5% cpu, 2 thunderbolt displays connected and working with the retina too, connected to power and pretty hot on the left side of the mac for sure, but fans are super quiet and everythings seems to be working as new.


Enjoy.

Exactly the same problem here running 10.12.1 16B2555 on a Mid-2015 Retina Macbook Pro.


It becomes unusable with a secondary monitor.


Anyone found a solution that doesn't involve removing a (I'm assuming quite important) kext?

Same problem for me on Sierra 10.12.1 => kernel


Nom du modèle : MacBook Pro

Identifiant du modèle : MacBookPro11,3

Nom du processeur : Intel Core i7

Vitesse du processeur : 2,3 GHz

Nombre de processeurs : 1

Nombre total de cœurs : 4

Cache de niveau 2 (par cœur) : 256 Ko

Cache de niveau 3 : 6 Mo

Mémoire : 16 Go

+

Belkin thunderbolt Dock and Viewsonic 4K monitor


(Sorry French user 🙂)



PID COMMAND %CPU TIME #TH #WQ #PORT MEM PURG CMPRS PGRP

55564 Terminal 20.5 00:08.61 13 8 230 24M+ 0B 6580K- 55564

232 WindowServer 6.8 26:55.39 4 2 371- 134M 4096B 187M 232

131 hidd 4.2 03:06.73 7 3 233+ 2224K+ 0B 3664K 131

0 kernel_task 3.8 52:41.34 155/8 0 14 9530M 0B 0B 0

58412 top 3.6 00:00.53 1/1 0 22 2568K+ 0B 0B 58412

52701- Microsoft Ou 1.7 01:53.29 19 1 357 75M 0B 41M 52701

320 java 1.2 05:16.90 553 1 1138 223M 0B 105M 94

57093 Google Chrom 0.6 00:52.73 17 1 149 205M+ 0B 0B 56998

53030 com.apple.Am 0.4 00:02.17 4 2 64 500K 0B 1160K 53030

31261- Google Chrom 0.3 01:30.80 21 1 327 15M 0B 28M 31261

31293- Google Chrom 0.3 02:03.70 7 1 130 91M 0B 192M 31261

63- hasplmd 0.2 00:23.94 7 0 37 3612K 0B 3864K 63

56998 Google Chrom 0.1 00:43.20 46 3 534 105M 28K 0B 56998

16537 netbiosd 0.1 00:14.91 5 5 35 736K 0B 4776K 16537

hey everyone, I've finally pinned down that the secondary display was what was giving me the same problem. I thought it was the OWC Thunderbolt 2 dock, but it wasn't. It was the display attached to it. Now I'm hooked up to the display via direct HDMI and it gives me the same problem. The hub by itself doesn't give me a problem.


I want to share what might have fixed the problem for me:
I went to settings, and in energy saver I toggled automatic graphics switching to OFF. It worked, and my cpu load is close to idle. I turned it back again and it still doesn't spike. Maybe toggling that off and on triggers something?


Again, this worked for me, I hope it works for everyone else.

Model Name: MacBook Pro

Model Identifier: MacBookPro11,5

Processor Name: Intel Core i7

Processor Speed: 2.5 GHz

Number of Processors: 1

Total Number of Cores: 4

L2 Cache (per Core): 256 KB

L3 Cache: 6 MB

Memory: 16 GB


with two external monitors. My fan is cranking all the time now. MacOS also disabled and then re-enables one of the my screen periodically. Like it's redetecting it. Worked fine with 10.11.


Very disappointed in 10.12. This and other problems like Mail.app crashing all the time.

For me it worked excluding the internal os x discs from spotlight.

I have a partition with os x & one with data, it worked to exclude these drives within the prefs of spotlight (privacy tab).

Same issue here when connected to ThunderBolt monitor


Model Identifier: MacBookPro11,5, (Retina, 15-inch, Mid 2015)

Processor Name: Intel Core i7

Processor Speed: 2.8 GHz

Number of Processors: 1

Total Number of Cores: 4

L2 Cache (per Core): 256 KB

L3 Cache: 6 MB

Memory: 16 GB

Boot ROM Version: MBP114.0172.B09

SMC Version (system): 2.30f2


Hope that Apple finds a solution rapidly... This affects my work. I have to unplug the monitor to be alble to work...

Hello ,


Same issue with 2 attached monitors. Kernel_task is raising 400-500-1000% (system usage) causing system to be unusable.

I don't know how to re-produce it its random i can work all day on 2 monitors but sometimes it just cant have them for a 15 min. After disconnecting minidisply <> DVI connect everything is back to normal it tooks ~2-3 min to normalize.

OSX version: 10.12.1 (16B2555)

Model Identifier: MacBookPro11,5

2 monitors - U2412M (HDMI and MiniDiplsay > thunder)


We all wait some apple feedback for that because kernel_task is PID 0 which is the actual parent of all processes you cannot kill it or do something with it.


I tried couple of traces but didnt catch anything which is a clue.

I've managed to solve the problem by removing a couple of Google Chrome extensions. It was not a random guess. I've figured out that kernel task is consuming reasources only when I'm running a number of Chrome processes. It happend only after i've tried the suggestions from this and other forums.


Macbook Pro 13", 2015, El Capitan

Same problem here using a 2012 MacBook Pro Retina and 3 external Asus 1080p monitors (laptop closed). Kernal_task useage was skyrocketing to 400% or more and the system was crawling along. I also had problems booting the system with the monitors attached. Would constantly freeze unless I unplugged one, rebooted, then attached after the system logged in.


Solution was switching 2 of the monitors that were using Display Port-to-HDMI to Display Port-to-DVI (using adapters added to the HDMI and moving to the DVI connection on the monitor). The 3rd display connected via the Mac's HDMI output to the Asus HDMI input stayed the same.


Kernal_task plummetted to 4.0% and the system returned to normal. Booting was normal. Fan useage also dropped.


Curious why the DVI connection makes a difference.

It seems toggleing "Automatic graphics switching" does indeed reset something. When I toggle it the kernal_task process drops away almost immediately.


Interestingly when I moniter the raw temp values coming from the dedicated graphics there is no direct correlation between high temperatures and when the kernal_task begins to throttle the machine. It seems this is a bug.