3

My MPB has had this problem for years every time the ambient temperature gets a bit warmer. It starts in the spring and gets ridiculous during the summer.

The MBP clocks down its processor to 0.8 GHz, making it pretty much unusable. This happens as soon as some process-demanding apps (like NodeJS or Unity) are running on the machine.

However, the CPU temperature is just fine, between 60 and 80 degrees Celsius. It doesn't look like a thermal problem. Thus, I don't understand why this happens?

I tried every hint I found on the internet, including resetting SMC and NRAM, I even re-installed macOS.

Because this has happened from day one, I have been at the Apple Store many times, and the MBP now has its 4th logic board (!) Still, the problem remains.

What else might cause this strange phenomenon? I'm at a loss!

  • 1
    It's not just the CPU temp that can cause a throttle down, there are other temp sensor that feed input to the SMC. Use a tool like iStat Menu to get all the temp readings. Another question would be where are you using this MBP that seasonal changes create the issue? Also, what are the fans doing? – Allan Apr 19 '18 at 17:59
  • "Where" is not so much interesting as at which ambient temperature (AC or not?); also: how many peripherals are attached, does this happen on battery or plugged in? In any case, follow Allan's request and post more sensor readings as well. Since I do not use Unity or NodeJS much: any other "demanding apps" running? Last system install time, caches cleared? – LаngLаngС Apr 19 '18 at 19:56
  • Regarding the where and the temperature: It even happens in German spring time, which means about 25°C, which is not that much. The fans are running at full speed as soon as I start the Mac (takes about 20secs until they reach their max). But I now have a suspicion: It gets much worse if a monitor or dock is attached via Thunderbolt. If I detach it, the CPU speed goes back to normal after a while, but the fans are still running loud. – waldgeist Apr 20 '18 at 18:24
  • Mine keeps doing it. All the temperatures are fine. I have a 2016 model. It's super frustrating. I hate this laptop. imgur.com/UcbAjna – Tim Harper Sep 19 '19 at 17:45
  • There are two reasons for 0.8Ghz from my experience: 1) external monitor 2) broken battery. In my case first time it was battery, so Apple replaced it for free. In second case it was external monitor and the only workaround I found is to use AirPlay instead of HDMI cable, after that no throttling anymore. – imike Mar 27 at 18:08
2

Just to show that adding to a two year old question is worthwhile, imike's comment from only a month ago, was spot on for me (although waldgeist himself had already begun to suspect the same in his own comment from only a day after his original post.)

So I have this very same MacBook Pro model, and was seeing the very same issue. And that very cause -- an external HDMI-connected monitor -- is at play.

The following shows two instances of it happening. In each case: something has caused the frequency to flatline; I remove the HDMI cable from the Mac; and within a few seconds things return to normal.

HDMI impact on MacBook Pro Core Freq.

I still don't know how the external monitor (Samsung U32J59x) is having this effect, but it's a step forward to at least know it is.

ADDENDUM

Red Herring alert. Sadly, whatever this is, it turns out it is not specific to HDMI. I have now seen it with an old Apple Cinema Display. I've even seen it while driving an iPad as an extra display using Duet. In each case, detaching the external device solves the problem and the frequency returns to normal. But in fact, I've even seen it occasionally with no external display at all. For example, it will sometimes happen during a video call (with Zoom, at least). There, I can solve the problem by turning off the video channel in my call (i.e. using only the audio channel).

So it may be as benwiggy suggested in a comment, something to do with graphics in general and the GPU. In other words, it's not an HDMI monitor, or even a monitor at all, but some function of an increase in graphics demand (which of course adding an external monitor would probably be an example). Push demand high enough and the CPU crawls into a corner and curls up; then remove the demand and the poor wee thing comes out of fibrillation and recovers. Sounds credible I guess.

One other point for anyone else chasing this thing. I can't be sure--maybe I'm now just fixated on it --but I could swear the problem is getting worse, week by week. If I do plug in an external display, it will happen, and fairly quickly. Five, maybe ten minutes at most, and then the frequency falls off a cliff and performance becomes barely usable. Then, detach the display and it will recover. It's less obvious when it happens when there is no display attached -- i.e. during a Zoom call -- but it's still highly likely.

So, what aspect of the machine could be deteriorating week by week and so could create this getting-worse effect? Well, the battery for one. And I recall that a battery issue was mentioned in a related thread. Also, I did have a power issue with this machine, about a year ago. While debugging that, I did get a couple of warnings that my battery needed replacing. In the end the problem turned out to be a PSU issue, and getting a new one did seem to solve it. And so I put the battery warnings down to mere bad-PSU-induced delusion on the part of the Mac's battery monitoring system. But it's perfectly possible there was indeed a battery issue beginning to appear, in addition to the main PSU one.

So I'd say that may be a worthwhile lead to pursue.

But for me: well I just didn't have time to keep debugging and so I went for a much more certain solution. A nice, shiny, brand new 16" MacbookPro! :-)

| improve this answer | |
  • I note that you have the model with a dedicated GPU. I wonder if that's a factor. I have the 2014 15" MBP just with Intel graphics, and I use an external monitor very occasionally , but have not seen this issue. – benwiggy Apr 24 at 9:35
  • I can't see any obvious change in GPU activity when the problem arises. However, there may be something in what you say; see my addendum to my answer. – tkp May 10 at 19:13

You must log in to answer this question.

Not the answer you're looking for? Browse other questions tagged .