Random Shutdown on Macbook Pro (15" mid 2014)

I'm having this exact issue as well. It's already the second logic board in a row I'm having this problem with. Had the first one replaced. Honestly I have had nothing but trouble with this machine (MacBook Pro mid-2014, intel iris pro 5200). Everything is already replaced except for the base plate (which is just a piece of aluminium). Downloaded the app from https://realmacmods.com/product/macbook-pro-nocrash-utility/. Hope it works for me. After installation when you click the spinning gear icon in the top menu bar, it states 0% completed. This should stay like this right? Using the laptop for 15 minutes now. Going good so far. Thanks guys!

I think there are quite a lot of us and the problems occur since macOS 10.12 released.

I have installed the nocrashMBP app and it did help solve the problem by running a background python script to keep the CPU waking up. I think this can only be a temporary solution because it causes more battery consumptions.

Apple should acknowledge this problem even they don't wont to replace our Logic board, they should ask the developing team to update their OS kernel to deal with this special issue. We should find a way to tell them.

1) reboot with CMD+R pressed.

2) open Terminal

3) sudo csrutil disable

4) reboot with CMD+R pressed

5) sudo mv /System/Library/Extensions/AppleThunderboltNHI.kext

/System/Library/Extensions/AppleThunderboltNHI.kext.BAK

6) csrutil enable

7) reboot and forget about problem

https://outluch.wixsite.com/rmbp-crash

They don't acknowledge this because that would mean admitting that aaaallll those people who have already bought a new logicboard for €700, did it because a fault on their side, it's a problem which already exists for a couple of years. This is scandal material.

Thanks for the method. And I have tried your way.

For my mac, I still got no luck to get rid of it. 😐

BTW, if we do step 4) and then step 5), the terminal response "read-only and cannot be moved", I move those files in the Normal mode, not in the recovery mode.

Even if all of the affected logic boards have the same issue (a sudden shut off with no log activity when running on battery with a very low CPU load), it might reveal in a bit different conditions due to tiny hardware deviations across the affected units. My 15" MacBook Pro (Late-2013, Intel i7 2GHz, Iris Pro) started to shut off a week after upgrading to High Sierra (10.13.2). That first week of using High Sierra the laptop was AC-powered 99% of the time. The issue revealed multiple times during a meeting when running on battery for about 5 hours, the first time after the upgrade. iStat Menus registered the CPU frequency dropping below 900 MHz right before the shut-offs. I noticed that NoCrashMBP holds the frequency at nominal 2 GHz by putting a 20% load on a single core. I downgraded back to Sierra (10.12.6) and no shut-offs have been observed for a few days of working on battery, without NoCrashMBP or any other thicks. Even though iStat Menus still registers the CPU frequency dropping below 900 MHz, but the ambient frequency of the integrated GPU is stable at 750 Mhz on Sierra rather than 300 MHz on High Sierra, which could prevent shut-offs on Sierra due to a higher power consumption.

Since the issue seems to be somewhere in between the SoC hardware and power management software, Apple should either recall the defective logic boards or maintain a software workaround, even at the cost of reducing power efficiency for the owners of the affected laptops.

For us here, we have two MBP Mid2014 bought at different times and with different OS. When they both started to do the same thing (shutting down for no reasons) I started to wonder… if they BOTH do that, more then likely they ALL (or a LOT of them at least) do that. So back to the thread here, how do we get Apple to pay notice? Being the old US of A, I would ask, does anyone know a lawyer who would like to get Apple to pay notice? I am saying this thinking there has to be a LOT of us out there, and yes, Apple should just do a fix that does not kill the battery life and allows us to work on our MBP for what they are worth.

Yes, of course, more of us!!! There is a fix NoCrashMBP and it works (only $10) but it eats up battery time big time and also runs the MBP a little hot I found. Otherwise no crash withMBP on. So WE know the problem (it has to do with some low frequency thing on the CPU if I get it right) but APPLE can’t get around to even acknowledge the problem publicly and write a simple fix for it. Much easier telling all of us to buy new logic board and cross our fingers…

There are solution. You will lose ability to use ethernet via thunderbolt.

PS. Today I told apple about this solution via chat. They maybe will fix it soon.

1) reboot with CMD+R pressed.

2) open Terminal

3) sudo csrutil disable

4) reboot in normal mode

5) sudo mv /System/Library/Extensions/AppleThunderboltNHI.kext

/System/Library/Extensions/AppleThunderboltNHI.kext.BAK

6) reboot with CMD+R pressed

7) csrutil enable

8) reboot and forget about problem

I agree, NoCrashMBP helps to keep some load on the CPU package, and that eats up battery. However, I'm now curious about why Sierra works for me, while High Sierra shuts down randomly. I have noticed a distinct difference in resource/power management between these versions of macOS: Sierra keeps the frequency of the integrated Intel GPU at 750 MHz when idling, High Sierra maintains much lower GPU frequency of 250 MHz. The CPU frequency management looks very similar. That's easy to check with Intel Power Gadget.

The increased GPU frequency on Sierra could be accompanied by a higher voltage of the whole CPU/GPU package, which prevents it from shutting down. On High Sierra can maintain a higher voltage by putting some load onto CPU when rinning NoCrashMBP.

Though I don't understand how it can be related to the Thunderbolt Ethernet driver.

Is the Thunderbolt Ethernet driver affected only in High Sierra? Isn't the cure just a side-effect of turning off a mandatory driver? E.g. there might be an infinite loop checking some condition related to the Ethernet driver, which could put some load on CPU and thus prevent a shutdown. Did you observe any shut offs on Sierra?

I've rolled back to Sierra, and it works fine for me now. I'd like to try your solution, but it would take a whole day to upgrade to High Sierra then downgrade back if the proposed solution won't work.

It isn't a problem related to the operating system. I've had this problem on 3 different ones. Although after a clean install the laptop maybe worked for half an hour instead of 5 minutes. After that it's business as usual. For me using a web browser seems to speed up the problem. Sometimes I could use the computer for some time and the problem just occurred after opening any webbrowser. But with NocrashMbp it doens't the problem is gone.

Well, after a few days of successful use of Sierra, my MBP 15" Late '13 started shutting off. So I'm confirming the issue is irrelevant to the macOS revision. I'm successfully running High Sierra now with the Thunderbolt driver disabled, as recommended by @outluch3.

I'm thinking the "fix" is just a side-effect of disabling the extension, which can stop working on any future OS update. I have no Thunderbolt peripherals and the normal current of the Thunderbolt controller is 0A reported by iStat Menus. When AppleThunderboltNHI.kext is removed, the current is stable at 0.28A, which might indicate some erroneous activity in the Thunderbolt driver. I noticed that the total system power consumption has also increased with the extension removal: it's about 9.5W when no significant activity is observed by Activity Monitor, CPU frequency is 800 MHz, GPU frequency is 750 MHz, and the screen brightness is minimal (1 dot). When the extension is enabled, the total power consumption is about 7.5W.

It means that the Thunderbolt workaround might be similar to what NoCrashMBP is doing - increasing the power consumption by about 20%. It might be an indication of a faulty power regulator in the affected MBP models.

Hi Everybody,

Me too I have the same issue with MacBook Pro (Retina, 15-inch, Mid 2014).

Bootcamp Windows 10 is working great, no issue.

(This alone, tells me it is not a hardware issue, but rather a MacOS software issue)

When I connect the Thunderbolt / Ethernet connector, it is working great, no issue.

(it does need to have an Ethernet connection, just the connector is enough)

The last release of High Sierra 10.13.3, does not solve it. Still the issue.

Here is a food for thought:

1 - When we do a clean install of El Capitan, and check the content of the file:

/System/Library/Extensions/IO80211Family.kext/Contents/PlugIns/AirPortBrcm4360.k ext/Contents/Info.plist

we have the lines :

<key>IONameMatch</key>

<array>

<string>pci14e4,43ba</string>

<string>pci14e4,43a3</string>

<string>pci14e4,43a0</string>

<string>pci14e4,4331</string>

<string>pci14e4,4353</string>

</array>

2 - When we check the content of the same file on High Sierra :

<key>IONameMatch</key>

<array>

<string>pci14e4,4331</string>

<string>pci14e4,4353</string>

</array>

3 - When we start on Bootcamp / Windows 10 and check the version of the Wifi Hardware we see

Device PCI\VEN_14E4&DEV_43A0&SUBSYS_0134106B&REV_03\80E600FFFF00000100

which is pci14e4,43A0

4 - One possible interpretation of this :

The MacOS Wifi driver on High Sierra, does not manage the hardware pci14e4,43a0

but only pci14e4,4331, pci14e4,4353

which could explain the crashes.

What do you think ?

Thanks guys.

I have the same problem on my MacBook Pro late 2013 2,6 GHz.

Problem started about one month ago.

Deactivating AppleThunderboltNHI.kext worked, but now thunderbolt drivers are disabled.

Anyway I am still able to attach a display using apple mini display to vga adapter plugged in the thunderbolt port.

Problems started after upgrading to High Sierra, I than downgraded to Sierra but nothig changed.

This is my third (and last) MacBook pro with harware problems, first two were GPU related (NVIDIA).

I'm really disappointed, Im a web developer, seems that these machines are not suited to work hard...

Really, really disappointed...