Got a tip for us?
Let us know
Become a MacRumors Supporter for $50/year with no ads, ability to filter front page stories, and private forums.
(2019) 16" is HOT & NOISY with an external monitor! :(
- Thread starter Appledoesnotlisten
- Start date
- Sort by reaction score
I had a 2015 13" MBP which powered the same monitor over HDMI without any issues. Battery held up normally, the fans didnt raise the RPM and the temperature was OK. However, my 16" gets really hot, the battery drains really fast and the fans become audible sometimes. So yes, I would say it comes all down to the dGPU using 18+W (in my opinion) for no reason. There is no logical aspect as why a monitor should always be powered by the dGPU - and even if so, why the dGPU uses 18W. I don't get it how something like this passed Apple's QC. But then again, I have the cracking speaker problem with videos since day 1… Thanks Apple.
After some more testing and observation in the last few days, im starting to believe that the 18-20Watt radeon highside is not the main cause for the high fan speed but rather a sensor in the thunderbolt ports/area. Most likely the same sensor/component that also triggers the Kernel_Task throtteling process.
I use a 27“ Lg 4k monitor via usb-c cable (with 60watt powerdelivery) plugged into the right side of the macbook with open lid.
At this stage, the fans stay quit, even after 10-20 min of browsing in safari and watching a 1080p youtube video.
Now if i connect an ipad to the same side the monitor is connected (or charging cable, same outcome), while all the sensors in istat menus stay at around 60-65celsius, the fans slowly start to increase their rpm to 100%. (Ipad sidecar and charging with 12watt)
Using macsfancontrol or a similar app to change the default fancurve can prevent this rpm increase from happening but
after some hours of usage (~2hours in my case), the kernel_task process will eventually kick in and cripple the macbooks performance.
The only working solution other than a reboot is unplugging all cables. The fans will slowly go down to minimum speed and the kernel_task throttle process will stop aswell.
Before unplugging all cables, i closed all apps and let it sit in idle for a few minutes but the fanspeed wouldnˋt decrease despite all components temps being already at a cool ~50C.
I agree that the radeon high side of 18-20watts is annoying and most likely a driver bug but this „overheating thunderbolt port“ issue does look like the root cause of fanspeed related problems on the 16“ mbp.
I use a 27“ Lg 4k monitor via usb-c cable (with 60watt powerdelivery) plugged into the right side of the macbook with open lid.
At this stage, the fans stay quit, even after 10-20 min of browsing in safari and watching a 1080p youtube video.
Now if i connect an ipad to the same side the monitor is connected (or charging cable, same outcome), while all the sensors in istat menus stay at around 60-65celsius, the fans slowly start to increase their rpm to 100%. (Ipad sidecar and charging with 12watt)
Using macsfancontrol or a similar app to change the default fancurve can prevent this rpm increase from happening but
after some hours of usage (~2hours in my case), the kernel_task process will eventually kick in and cripple the macbooks performance.
The only working solution other than a reboot is unplugging all cables. The fans will slowly go down to minimum speed and the kernel_task throttle process will stop aswell.
Before unplugging all cables, i closed all apps and let it sit in idle for a few minutes but the fanspeed wouldnˋt decrease despite all components temps being already at a cool ~50C.
I agree that the radeon high side of 18-20watts is annoying and most likely a driver bug but this „overheating thunderbolt port“ issue does look like the root cause of fanspeed related problems on the 16“ mbp.
I asked the question on reddit as well (who uses the external display with the 16” and have no heating/loud fan issues) and several comment that there machine is running cool and quiet when connected to display. Makes me wonder maybe hardware does have something to it?
You can argue maybe they are simply not noticing it much, but mine was so hot, loud and laggy the moment I hooked it up there was no way to not notice it.
You can argue maybe they are simply not noticing it much, but mine was so hot, loud and laggy the moment I hooked it up there was no way to not notice it.
I'd say people aren't bothered by it or notice it. I can't believe it's a hardware issue that affects some machines, it's pretty specific.
yeah but when I hooked it up to a display the machine got so hot and laggy it wasn’t possible to actually use it. It’s not something that is possible not to notice.
So the very least it didn’t effect their machine in the same way. (And I didn’t do anything special, just chrome open)
I didn't get it to that extreme. I hooked mine up to my Acer 4K monitor via the Apple dongle and HDMI. The GPU went to 40w and the fans gradually span up, but I could still run Logic and play the Beck demo tune and do other stuff. The system didn't slow down at all.
For me it had a noticeable lag connected to simple 27” 1080 screen. I’m starting to wonder if the USB C to hdmi cable I used was part of the problem
Was it like that on day 1? Could it be heat damage from prolonged use with h an external display?
From day 1. It was sent back to Apple within then return period due to this issue
I initially had some 3rd party adapter and it felt laggy, but then i realised the adapter could only do 4K at 30hz so that's why it lagged. I bought the official apple adapter (after selling a kidney to afford it) which could do 4K at 60hz, and i get no lag.
This has been posted and discussed before a few times in this thread… Basically, it's worthless. And as many already said:
1. This is just fighting the symptoms - and with like every disease, just fighting the symptoms without curing the disease is a bad idea.
2. Throttling the system controlled fans leads to an increase in "kernel_task", which kicks in if the temperatures get too high and lowers the CPUs performance to reduce temperature. So basically, your device becomes slower because it can't control the fans anymore.
3. It is not really about fan noise, but the problem that the dGPU uses 18+W with an external display. And no fan control, no TG Pro / deactivating CPU boost can solve this.
I tested all his suggested workarounds… after like 2 hours the kernel_task throttles the performance really badly, one could think it is an 2010 Windows Laptop how slow it becomes. So in short: ignore this video.
Maybe for you, but there are plenty of comments about the fan noise, so this may help some people where this is the main issue for them. And TBH is the dGPU using 18+W really an issue? OK so the machine gets hotter and the fans a little noiser, but is it a problem, or just an inconvenience?
I saw a video on Reddit but I can’t find it now. I saw somebody using airpods,
Running down each side of the keyboard part of the MacBook to turn off the screen. I think magnets had much to do with it
Will this replicate clamshell mode?
Running down each side of the keyboard part of the MacBook to turn off the screen. I think magnets had much to do with it
Will this replicate clamshell mode?
Well, if a performance decrease after like 2 hours is more convenient than a little bit of fan noise, go for it. But I doubt that anyone in this thread would sacrifice the perfomance and work with a laggy device, just because the fans get a little bit noisier.
And yes, 18W+ while being connected to an external monitor is an issue. The battery won't last more than 3.5 hours, the device gets really hot and the fans have to spin faster. However, if I just use an external monitor without charging the device, the fans aren't really bothering me. Oh and if you have the time, read the 150+ pages in Apples forum. Someone uses multiple 16" MBPs in his company and many of the devices started to show longterm malfunctions (rgb-lines along the scree, so probably a fried dGPU). This is probably due to the permanent high temperatures, just because the dGPU sits at like 18+W.
[automerge]1589119288[/automerge]
This will not work with the 16", as I have posted a few sites ago. The 16" uses a new "angle sensor", whereas older devices had a HAL sensor, that reacts to magnets (https://www.macrumors.com/2019/11/19/16-inch-macbook-pro-lid-angle-sensor/)
However, just closing the lid won't help either in a few cases: closing the lid while connected to the external monitor will shut the screen off. However, if you haven't set it up correctly (see the different findings in this thread, where different Hz matter), the dGPU will still use 18+W.
The only workaround on that video that actually did wonders for me was the Turbo Boost disabler. Two weeks on and no issues. The `kernel_task` issue persists with the TG Fan Pro software.
This workaround only works because its essentially suffocating the performance of the CPU in order for the GPU to breath. So to speak.
Yeah it's not acceptable to choke your CPU because your GPU is ovhereating.
For those who run connected to external displays without any issues - what exact cable and monitor are you using? I have a feeling the issue is to do with whether you're using display adapter or hdmi to drive the external display.
[automerge]1589175036[/automerge]
Haha thanks, unfortunately from reading up a bit about this it appears it's no ta driver fault but a limitation of the AMD GPUs, they have to run their memory speed on boil whenc onnectec to external displays through HDMI for some absurd reason.
That's why I'm very curious to know what other people are doing - if you don't have issues please tell us what your cable is, the type and brand of display!
*For the record I was banned from the Apple forums for pointing this out and saying it was a defect.
For those who run connected to external displays without any issues - what exact cable and monitor are you using? I have a feeling the issue is to do with whether you're using display adapter or hdmi to drive the external display.
[automerge]1589175036[/automerge]
Haha thanks, unfortunately from reading up a bit about this it appears it's no ta driver fault but a limitation of the AMD GPUs, they have to run their memory speed on boil whenc onnectec to external displays through HDMI for some absurd reason.
That's why I'm very curious to know what other people are doing - if you don't have issues please tell us what your cable is, the type and brand of display!
*For the record I was banned from the Apple forums for pointing this out and saying it was a defect.
Last edited:
@_meds
I don't think that the cable matters, as long as you don't use it in clamshell. Especially in the Apple forum different combinations were reported, all causing the same problem. I tested my device in our general office, where whe have all kind of monitors and cables: It didn't matter if I used my USB-C to HDMI Hub. USB-C to USB-C, USB-C to DisplayPort or whatever… Every time my device became hotter and it drained more power (I didn't have iStats installed back then, so I can't tell you any W from the Radeon, but the temperature readings in MacsFanControl and the fan spin increased).
However, using clamshell is not an option to me, so all the testing was done with the lid open. Seeing all these findings in this thread, my device would probably stay cool and silent in clamshell with the right settings, just like all the other 16" MBPs in here.
I don't think that the cable matters, as long as you don't use it in clamshell. Especially in the Apple forum different combinations were reported, all causing the same problem. I tested my device in our general office, where whe have all kind of monitors and cables: It didn't matter if I used my USB-C to HDMI Hub. USB-C to USB-C, USB-C to DisplayPort or whatever… Every time my device became hotter and it drained more power (I didn't have iStats installed back then, so I can't tell you any W from the Radeon, but the temperature readings in MacsFanControl and the fan spin increased).
However, using clamshell is not an option to me, so all the testing was done with the lid open. Seeing all these findings in this thread, my device would probably stay cool and silent in clamshell with the right settings, just like all the other 16" MBPs in here.
Power Nap is off here and still getting kernel panics when plugging in to charge+screen going to sleep.
Similarly Apple is telling me to wait for a fix. And that’s it. Even though in Australia under consumer law, this essentially says it’s not fit for purpose and a major fault (random kernal panic shutdown issue). This law allows consumers to request a replacement or refund. And since it’s possible software. Replacement won’t fix the issue (they’re not even offering any hardware remedy because they know it’s a software issue but it’s nowhere near ready).
I emailed Tim Cook nicely asking for a remedy on these issues. And I got a call from a “Exec Liaison” and the attitude was that that they pretty much don’t care about helping in any other way other than live with the software issue until there is a fix. Basically said feel free to take legal action as per my consumer rights but not interested in providing any remedy’s.
Thanks for sharing this - is it your understanding that Apple see it as a software problem and are intending to release a patch? I'll be receiving a top spec refurbed macbook pro 16" later this week, and want to make sure I test it out during the 14 day return window. I'll be connecting to an external display and I won't be using clamshell, will be interesting to see how loud the fan noise is.