(2019) 16" is HOT & NOISY with an external monitor! :(

[am2am]

A 2 min export is way too short to actually get all the cooling system and chassis up to temperature. That's when you're gonna get throttling. Leaving apart the fact that some users may be doing workflows that don't involve the GPU much at all (like a lot of the Adobe Creative Suite, which is poorly optimized to the newer GPUs), all that power/heat is effectively wasted.

I agree - with the workflow not requiring GPU this might be a problem and a waste .. but if you don't need GPU in your workflow MBP16 with dGPU may not be the best machine for your needs.

2 mins is short indeed but still an indicator.
I have been working with and without external monitor on real projects that took much longer to render/export - did not measure it but have not noticed any significant difference in the performance.
[automerge]1588978280[/automerge]

Can you please try running cinebench on loop via min test duration AND luxmark GPUs stress test (loops) at the same time for 10 minutes?

Do this with just the internal display and then with the external display and lid open which triggers 18W Radeon at idle.

While the tests are running, watch Intel power gadget to see if both frequency and power/PKG are able to sustain a stable level (and not what the levels are). If possible, a screen recording showing Intel power gadget for a minute, after running the stress tests for a few minutes would be great.

This should test and confirm (or disprove) that even though hot and noisy, performance is not lost (I don't believe this will be true).

Sorry - not a big fan of benchmarks, not using them normally so not sure if I run them correctly (first time I use Luxmark). I did not run the test 10mins. it was closer to 5 mins in every scenario.

Every time I started with Luxmark - selected stress test and verified that GPU power is jumping to approx 60W+, then I started Cinebench 20 and started intel power gadget.
In both scenarios both fans reached the maximum and stays like this till the end of the test for few mins.

Below you see the screenshots nearly at the end of each test (first screenshot:MBP16 internal monitor only, second screenshot: internal + external benq 4k)

Not an expert so unable to interpret the results, but non-scientific experience was very much similar in both cases.

 

[mrmachine79]

Sorry - not a big fan of benchmarks, not using them normally so not sure if I run them correctly (first time I use Luxmark). I did not run the test 10mins. it was closer to 5 mins in every scenario.

Every time I started with Luxmark - selected stress test and verified that GPU power is jumping to approx 60W+, then I started Cinebench 20 and started intel power gadget.
In both scenarios both fans reached the maximum and stays like this till the end of the test for few mins.

Below you see the screenshots nearly at the end of each test (first screenshot:MBP16 internal monitor only, second screenshot: internal + external benq 4k)

Not an expert so unable to interpret the results, but non-scientific experience was very much similar in both cases.

View attachment 913350 View attachment 913351

Thanks. That is helpful. Sustained 30-35W for power and 1.7-1.8ghz (below advertised base) frequency does seem low, but consistent.

This leads me to believe that GPU performance may not be reduced, or may be only marginally reduced with GPU workloads with external displays and open lid.

BUT, I believe CPU workloads are affected with an external display vs internal only, even though not really "using" the GPU.

Could you please try one more set of tests. Just cinebench (no luxmark) for at least 5 minutes on loop, first with internal only, then with both displays?

I think you should see sustained 50-60W and above base frequency, maybe over 3ghz.

You could also run the usual Luxmark “selected only” test with just the dGPU selected, with internal only and then dual display. That should confirm if GPU performance is also affected and to what degree.

EDIT: Note also that CPU temp is only 80c. It's not a thermal throttle issue (CPU too hot) but either an overall system temp issue (fans already maxed, slowing down CPU is the only other thing the system can do) OR a power throttle issue (if dGPU is using 50-60W, not enough power left for CPU).

 

[cptcaveman]

I returned my 5 day old AUD$6000 (USD$3900) dollar machine due to this issue today. I will update to either a revised MBP 16 that has this issue fixed later in the year, or to a fully loaded MBP 14 when it comes out. Who knows I might be lucky enough to get 4k internal display at the same time. I simply can't believe that something as stupid as this is an issue for open lid users. I can buy laptops for literally 1/10th the cost that I can plug into a display and don't jump 30 degrees C at idle. Lift your game Apple, if you charge such a premium price you need to deliver perfection!

 

[trojans10]

This thread is scaring me - I'm using an ultrafine 5k. Am I better with the 2020 13? or the 2019 16? I ordered both to test out.

 

[Quu]

This thread is scaring me - I'm using an ultrafine 5k. Am I better with the 2020 13? or the 2019 16? I ordered both to test out.

Keep in mind forums like this will attract those with problems. People who are perfectly happy with the machine aren't going to come to this thread in as large number as those who are googling the problem they have and coming across this.

Personally mines fine, isn't loud open or closed with an external. But it sits over a foot away from me when plugged into my external monitor. I also don't have any performance issues when a monitor is connected (I don't believe many do but I saw a few people saying they had issues of this nature).

 

[am2am]

Keep in mind forums like this will attract those with problems. People who are perfectly happy with the machine aren't going to come to this thread in as large number as those who are googling the problem they have and coming across this.

Personally mines fine, isn't loud open or closed with an external. But it sits over a foot away from me when plugged into my external monitor. I also don't have any performance issues when a monitor is connected (I don't believe many do but I saw a few people saying they had issues of this nature).

Same message from me. I'm not here regularly - jumping from time to time out of curiosity because I do NOT have problems described in this post. Yes I have 18W when external display is connected but it is not impacting my work as explained few posts earlier. I can connect external display and with light work (several tabs in safari, YouTube 1080, emails) my fans are inaudible idling at minimum.

 

[am2am]

Thanks. That is helpful. Sustained 30-35W for power and 1.7-1.8ghz (below advertised base) frequency does seem low, but consistent.

This leads me to believe that GPU performance may not be reduced, or may be only marginally reduced with GPU workloads with external displays and open lid.

BUT, I believe CPU workloads are affected with an external display vs internal only, even though not really "using" the GPU.

Could you please try one more set of tests. Just cinebench (no luxmark) for at least 5 minutes on loop, first with internal only, then with both displays?

I think you should see sustained 50-60W and above base frequency, maybe over 3ghz.

You could also run the usual Luxmark “selected only” test with just the dGPU selected, with internal only and then dual display. That should confirm if GPU performance is also affected and to what degree.

EDIT: Note also that CPU temp is only 80c. It's not a thermal throttle issue (CPU too hot) but either an overall system temp issue (fans already maxed, slowing down CPU is the only other thing the system can do) OR a power throttle issue (if dGPU is using 50-60W, not enough power left for CPU).

Disclaimer: I will NOT run more tests. Don't have time for it
Here are the results - screenshots of power gadget made consistently at the end of 300s tests
(my configuration for the reference: i9 2.4, GPU 5500 with 8GB, 2T SSD - purchased Nov last year)

1. Cinebench 20 - internal display only.
I would say as expected - solid CPU performance above advertised 2.4Ghz (3,09 avg), power draw ~60W


2. Cinebench 20 - both internal & external 4k
Lower average frequency (still above advertised), power draw ~ 47W. As mentioned in one of the pervious posts this is not a surprise - if you are not using dGPU the situation is impacting your overall performance.

In my case this is only hypothetical impact - real life I am not seeing performance drop. I connect external monitor only when I am working on graphics intensive applications with dGPU used heavily. As shown before in such scenario performance is nearly the same with and without external monitor.
I do recognise that it may impact others with workload not requiring dGPU.



3. Luxmark - internal display only.
Radeon started with ~60W (I have istats opened) at the beginning of the test, after approx 1m30s stabilised at around 40W consistent draw.



4. Luxmark - internal & external displays.
Quite honestly I see no difference. Radeon started with similar ~60W and after a while stabilised around 37-38W consistent draw.


Overall - I am really happy with MBP16.
Have purchased MBP15 in May 2019 but returned after a week to add more SSD storage. Before I've placed new order rumours on 16" surfaced and I decided to wait. Couldn't be happier.
This is really powerhorse with great capabilities - the best macbook pro in years in my opinion.

Catalina is a different story .... worst OSX release in years.
I see some light in the tunnel - they are improving slowly.
I was reporting before that Radeon draws consistent 5W as soon as more than 1 user is logged in (no application needed performance GPU, 1 user: Radeon High side ~0.03W, with more than one user:~5W). I still see it but now it only draws ~2W. Improvement. Maybe they will fix 18W draws with external monitor as well.

 

[MrGimper]

Same message from me. I'm not here regularly - jumping from time to time out of curiosity because I do NOT have problems described in this post. Yes I have 18W when external display is connected but it is not impacting my work as explained few posts earlier. I can connect external display and with light work (several tabs in safari, YouTube 1080, emails) my fans are inaudible idling at minimum.

So as a summary, and correct me if I'm wrong, but your machine behaves the same as others when using external monitors.... but the difference is that you don't consider it a problem when others do? I'm not judging, I'm just making sure that the hardware is doing the same, but your mindset is different, that's all.

 

[mrmachine79]

3. Luxmark - internal display only.
Radeon started with ~60W (I have istats opened) at the beginning of the test, after approx 1m30s stabilised at around 40W consistent draw.

Thanks very much for running those tests. Confirms my suspicion that CPU performance is degraded by this issue. True that many people won't notice because the machine is not grinding to a halt. This result is way better than what I see on my 2019 15" where it has 15W additional power draw and can grind to a halt with medium usage.

I should have clarified earlier, for the luxmark tests I was interested in the luxmark test result number to determine if GPU performance is degraded by the additional power draw. Intel power gadget CPU data doesn't help with that.

Without re-running those two tests, did you happen to note the luxmark test result numbers?

 

[am2am]

o as a summary, and correct me if I'm wrong, but your machine behaves the same as others when using external monitors.... but the difference is that you don't consider it a problem when others do? I'm not judging, I'm just making sure that the hardware is doing the same, but your mindset is different, that's all.

The title of this thread is: 16" is HOT & NOISY with external monitor.
A lot of people are claiming here they experience excessive heat and fan noise as soon as they connect external monitor.

Mine is nether hot nor noisy (unless I'm using it for heavy tasks and then it is hot & noisy with or without external monitor). So my hardware is not doing the same as some others reported here. Just stating the fact.
I let you interpret if mindset has anything to do with it.

p.s.
my glass was always half full
[automerge]1589020972[/automerge]

Thanks very much for running those tests. Confirms my suspicion that CPU performance is degraded by this issue. True that many people won't notice because the machine is not grinding to a halt. This result is way better than what I see on my 2019 15" where it has 15W additional power draw and can grind to a halt with medium usage.

I should have clarified earlier, for the luxmark tests I was interested in the luxmark test result number to determine if GPU performance is degraded by the additional power draw. Intel power gadget CPU data doesn't help with that.

Without re-running those two tests, did you happen to note the luxmark test result numbers?

Sorry - no. Did not look at test results at all - just launched it in stress test mode and captured intel gadget screen.
The only observation I had was about Radeon High Side in stats - it was nearly identical in both cases. I understand this may not be enough to conclude.

 

[TJ82]

Okay so after posting originally that my 1080p monitor was fine after getting the 16inch this week, I've now just came back into the room to hear the fans blazing. There's only one website up, no other apps running, and the temperature above the Touch Bar is scorching.

I thought I'd dodged a bullet, maybe because this screen is low res and on USB-C.

 

[Abaganov]

I’m still confuse to if this issue happens to 100% of 16” users or to select few? Surely if this would happen to all 16” owners there would be much more talk about this?

 

[MrGimper]

I’m still confuse to if this issue happens to 100% of 16” users or to select few? Surely if this would happen to all 16” owners there would be much more talk about this?

Unless you hook up an external monitor, it's not an issue. And TBH apart from inconvenience of a bit of noise and heat, I don't think it seems to be a problem.

 

[cptcaveman]

I’m still confuse to if this issue happens to 100% of 16” users or to select few? Surely if this would happen to all 16” owners there would be much more talk about this?

Maybe many people use don't use external monitors at all or use clamshell.

I personally don't believe that my MBP 16 returned today was defective or any different from everyone else's here. It's a thermal issue related to the dGPU power draw as raised so many times in this thread, related to driver or hardware I don't know, but very likely by design. I think the new thermal architecture, new default fan curve and bigger and presumably louder impeller fans also exacerbate the issue.

The different reports are likely a result of people having different expectations, levels of hearing, ambient environments, and differences in their general tolerance for noise.

I am now back on my 2019 MBP 15 2.4 I9, Radeon Pro 560X 4 GB with my external LG HDR monitor attached and can't hear a whisper. Radeon Highside is 11W. This morning on the returned machine it was double that with fans blaring in the exact same scenario.

I am so glad I went back and will try again later in the year with a revised model.

 

[PeterJP]

It's not a red herring, but it's not a problem by itself. I can use an external screen with Radeon High Side at 20W all day without the fans coming up.

That is, until I start an online meeting eg in MS Teams. Teams by itself works fine without fan noise, too. But with an external screen connected, the fans come up loud.

The same for any sort of load. The 16" can take a lot. When connected to an external screen, it can take a lot less. It's not entirely problematic but it's definitely there.

 

[decipherkl]

IMO most if not all 16" MBP 2019 exhibits this characteristic.

1) Those who don't use an external monitor won't be coming to this forum to complain

2) Some who do use an external monitor but don't notice loud fan noises probably won't either

3) Some who do use an external monitor and don't notice loud fan noises but are more observant probably will complain. They noticed the fault in their machine but don't want to be conned (by now most in this group probably won't agree with Apple's marketing promise)

4) Those who do use external monitor(s), observes high temperatures and hear loud fan noises are probably the most dissatisfied and they believe they have a lemon


I am in group 3 above.

 

[mrmachine79]

I am now back on my 2019 MBP 15 2.4 I9, Radeon Pro 560X 4 GB with my external LG HDR monitor attached and can't hear a whisper. Radeon Highside is 11W. This morning on the returned machine it was double that with fans blaring in the exact same scenario.

What resolution do you run the LG monitor at? I have the same 2019 15" and I find performance is seriously degraded with external monitor at looks like 2560x1440 resolution (15W Radeon) and trying to do a video conference (uses GPU) while doing light coding work. I have similar heat and noise issues as reported here and am replacing this machine with a 16" thinking although it is (equally as) noisy hopefully the performance degradation is not as severe due to better cooling.

Could you by chance run the same tests that I asked am2am above to run on his 16", on your 15"?

 

[cptcaveman]

It's a LG 27UL600-W with a native resolution of 3840 x 2160 connected via HDMI over a Caldigit Thunderbolt 3 Mini. I am running at Default for display.

 

[esphil]

It's not a red herring, but it's not a problem by itself. I can use an external screen with Radeon High Side at 20W all day without the fans coming up.

That is, until I start an online meeting eg in MS Teams. Teams by itself works fine without fan noise, too. But with an external screen connected, the fans come up loud.

The same for any sort of load. The 16" can take a lot. When connected to an external screen, it can take a lot less. It's not entirely problematic but it's definitely there.

This happens to my 2015 mbp connected to a external screen open, its not a specific sceneric around the 16" only. At the end of the day the fans are still pretty quiet even when they come on with the 16" to me. I think Apple users have gotten too spoiled thinking when they hear fans on a Mac come on. My pc desktop fans are much much louder just running idle.

 

[MrGimper]

This happens to my 2015 mbp connected to a external screen open, its not a specific sceneric around the 16" only. At the end of the day the fans are still pretty quiet even when they come on with the 16" to me. I think Apple users have gotten too spoiled thinking when they hear fans on a Mac come on. My pc desktop fans are much much louder just running idle.

The fans on my 16" at max speed (both fans) = 50dB, which isn't at all bad.

 

[alexweej]

I can connect external display and with light work (several tabs in safari, YouTube 1080, emails) my fans are inaudible idling at minimum.

As per my post here summarising as best I can everyone's evidence, a few people claim to have an 18 W draw while their fans stay at a minimum 1800rpm. Nobody has explicitly claimed they can leave it like this for 10 minutes and the fans still stay actually at minimum (are they creeping up to 2800-3000?). My friend who has a 'working' model however does spin up to 3300/3500 with one external monitor plugged in - this is quiet...ish, but nowhere near as quiet as my 2016 MBP "Touch Bar". It's marginally better than my new 16", but my flat is regularly 25 degrees C in the summer so this difference could just be chalked up him having a lower ambient temperature.

 

[onyx_eagle7]

I have a 2016 MacBook Pro 15” that I use in clamshell mode and works well with the LG UltraFine 5k. I’m looking to upgrade my MBP but want to use it in screen open mode. From what I see in this threads, this could be an issue with the 16” MBP. I got one earlier (before I had my UltraFine display) and then returned it due to GPU panics on sleep. With these issues, I’m considering the following options -

1) upgrade to this machine and hope it works or Apple sorts it out

2) get the $1999 version of the 13” MBP with integrated graphics (probably won’t have the kernel panics either?)

3) just wait for the next revision.

Thoughts?

 

[skalfyfan]

I would assume the new Macbook Pro 13" (high- end versions) also has a similar new thermal architecture. Would be curious to know whether this new model also gets significantly hotter when hooked up to external or does it remain cool, calm, and idle? Is the 16" external monitor issues solely down to getting forced to use the dGPU, and it's excessive power draw?

 

[MrGimper]

I have a 2016 MacBook Pro 15” that I use in clamshell mode and works well with the LG UltraFine 5k. I’m looking to upgrade my MBP but want to use it in screen open mode. From what I see in this threads, this could be an issue with the 16” MBP. I got one earlier (before I had my UltraFine display) and then returned it due to GPU panics on sleep. With these issues, I’m considering the following options -

1) upgrade to this machine and hope it works or Apple sorts it out

2) get the $1999 version of the 13” MBP with integrated graphics (probably won’t have the kernel panics either?)

3) just wait for the next revision.

Thoughts?

Depends what you define as an issue? It gets hotter, but stays withing spec, and louder as the fans spin more, but within spec.

 

[onyx_eagle7]

Depends what you define as an issue? It gets hotter, but stays withing spec, and louder as the fans spin more, but within spec.

How does it perform? Does it lag? Say you’re on a Teams (or any other VP9 based) video call and are using an external monitor, with a few browser windows and MS office apps open.