At the end it is what it is: A mobile computer. It has a strong GPU that can run external monitors but it's meant to be used without. It would probably better to just use an eGPU when external monitors are connected.
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
Apple's product offerings drive us to using laptops with External Displays. If they had a desktop in-between the Mini and Pro, things would be easier. Yes, an eGPU would be one way to go but spending an extra $1K wouldn't be my preference. Meanwhile my 2014 MacBook Pro runs an external 4K on integrated just fine.
100%. I’d even take a Mini with i9 and 128GB+ RAM options at this point to use in combo with eGPU, docks and external arrays. MBP16,1 is a fantastic machine and can do high intensity “typical” desktop work, but there are tradeoffs. It is just not a desktop.
The iMac is in between the MacMini and the MacPro ... but this is mostly a compromise.
I have a 2010 27 inch iMac. The video card went many years ago. It awaits recycling. I don't want to toss a machine because a component fails. I also like to bring my older monitors with me when I upgrade. I do not like the concept of AIO.
I got myself a base model 16", and connected it via a Call Digit TS3+ to a 1440p monitor. And indeed the fans are going crazy just by watching youtube.
I'm waiting for a Radeon VII to try it with epgu, and see how it goes.
I'm waiting for a Radeon VII to try it with epgu, and see how it goes.
Issues with the VII and sleep due to firmware on card. Likely won’t be fixed. See the Sonnet eGPU documents for more info:
Page 6 specifically has the VII notes.
I got my third 16" MacBook Pro yesterday. This time, I think I got a keeper 
The first one was a base model Core i9 early December. It had quite noticeable ghosting/smearing which does not disturb me and the sound pop problem when playing videos. And it's fan noise when connected to my external 4K monitor either directly or via my Caldigit TS3+ dock drove me crazy, it made lots of noise with low CPU load even in clamshell mode.
Early January I tried another base model Core i9. The display was noticeable better, it had no sound pop problems, but the fan noise was still not acceptable to me. When the machine gets taxed I am ok with that but not when the machine is doing hardly anything at all! So I sent it back and decided to try the base model core i7.
What can I say, this one seems to be a keeper I have less power than before (6 cores instead of 8 cores) and the Radeon 5300M instead of 5500M but the machine is noticeably more silent.
I am doing software development in Java with it. But currently I am chilling because it is weekend and I did some simple testing. When I use it on my lap it is very very quiet, the fans mostly spin at 1700rpm which I cannot hear.
I connected it to my 4K 43" display via a USB-C to HDMI cable and tried clamshell mode and tried letting the laptop display turned on. As to be expected, non clamshell mode drove the energy consumption up.
But I do not need that, my 4K display is a huge 43" mammoth, so no need for the additional laptop display.
I could watch 4K 60fps on YouTube via Chrome for a long time without hearing anything from the fan, very nice. Since Mac OS has no hardware support for VP9 encoded videos via Chrome the CPU gets taxed by playing these videos. CPU load goes up to 30% but the fans stayed very silent and went from 1700 to about 2400 rpm while watching videos, very nice The core i9 MacBooks went crazy while I tested that.
I surfed around using Safari, wrote some documents, downloaded some games via Steam and Epic Launcher, tried Cinebench, ... When the machine gets under load (e.g. while gaming) the fans ramp up but I am ok with that, that's what they are there for.
The machine is generally more silent than the i9s, and the fans seem to be more silent under full load than the i9.
Tomorrow I will tax the machine with software development using IntelliJ Idea and later this week will also install Parallels 15 for doing some Visual Studio development under Windows 10. I hope the machine will perform well, too I will then see how the fan noise performs when connected to the 4K monitor via my Caldigit TS3+ dock.
At the moment I cannot tell if the CPU downgrade, the GPU downgrade or both were the reason for the better thermal performance, but ok.
I used a MacBook Pro 2016 15" before and the upgrade from 4 cores to 6 cores should be noticeable for software development.
For gaming I use a dedicated gaming PC. But I tested some games with the 5300M like Civilization VI, Tomb Raider, Subnautica, ... and the difference from the Radeon 5300M to the Radeon 455 in the MacBook Pro 2016 is huge Very nice to being able to play some games directly on the Mac without having to switch to my PC gaming machine!
That is not what I bought the machine for, but it is nice to have.
Otherwise, the phenomenon which some people call "ghosting" or "smearing" is happening on this machine, too, but not very noticeable for me. It only caught my attention because it was mentioned in this forum, it does not bother me.
Keyboard is fine. Much better than the butterfly keyboard on my MacBook Pro 2016" (which was replaced by the 2017 revision keyboard). It is definitely more silent and very precise. I hope that this keyboard does not fail. I am using a MacBook 12" 2017 in addition. I used it mainly in the time between end of November when I sold my MacBook Pro 2016" and now and I must say that I really liked the precision of this butterfly keyboard, it is not identical to the butterfly keyboards on the Macbook Pros!
Sound is great, but that was no surprise. The build quality of the device is very good. The display is gorgeous, the trackpad is perfect, ...
All in all I really like the machine. I will continue using it the next days and will keep you informed in case something changes.
One thing to note: the machine is more silent that the Core i9 versions I tested but it still is overall louder than my MacBook Pro 2016 that I have used the last three years before. This could be interesting for people wanting to switch from such a machine and only use it with moderate CPU load to heavy CPU load.
People who use the machine full tilt most of the time like video editing won't have a problem with the fan noise I assume. But in software development the CPU is more used in little bursts while editing/refactoring code, building code and then running tests/code, etc. So this is a different scenario which demands CPU power but not 100% of the time if I could make this clear.
The first one was a base model Core i9 early December. It had quite noticeable ghosting/smearing which does not disturb me and the sound pop problem when playing videos. And it's fan noise when connected to my external 4K monitor either directly or via my Caldigit TS3+ dock drove me crazy, it made lots of noise with low CPU load even in clamshell mode.
Early January I tried another base model Core i9. The display was noticeable better, it had no sound pop problems, but the fan noise was still not acceptable to me. When the machine gets taxed I am ok with that but not when the machine is doing hardly anything at all! So I sent it back and decided to try the base model core i7.
What can I say, this one seems to be a keeper
I have less power than before (6 cores instead of 8 cores) and the Radeon 5300M instead of 5500M but the machine is noticeably more silent.I am doing software development in Java with it. But currently I am chilling because it is weekend and I did some simple testing. When I use it on my lap it is very very quiet, the fans mostly spin at 1700rpm which I cannot hear.
I connected it to my 4K 43" display via a USB-C to HDMI cable and tried clamshell mode and tried letting the laptop display turned on. As to be expected, non clamshell mode drove the energy consumption up.
But I do not need that, my 4K display is a huge 43" mammoth, so no need for the additional laptop display.
I could watch 4K 60fps on YouTube via Chrome for a long time without hearing anything from the fan, very nice. Since Mac OS has no hardware support for VP9 encoded videos via Chrome the CPU gets taxed by playing these videos. CPU load goes up to 30% but the fans stayed very silent and went from 1700 to about 2400 rpm while watching videos, very nice
The core i9 MacBooks went crazy while I tested that.I surfed around using Safari, wrote some documents, downloaded some games via Steam and Epic Launcher, tried Cinebench, ... When the machine gets under load (e.g. while gaming) the fans ramp up but I am ok with that, that's what they are there for.
The machine is generally more silent than the i9s, and the fans seem to be more silent under full load than the i9.
Tomorrow I will tax the machine with software development using IntelliJ Idea and later this week will also install Parallels 15 for doing some Visual Studio development under Windows 10. I hope the machine will perform well, too
I will then see how the fan noise performs when connected to the 4K monitor via my Caldigit TS3+ dock.At the moment I cannot tell if the CPU downgrade, the GPU downgrade or both were the reason for the better thermal performance, but ok.
I used a MacBook Pro 2016 15" before and the upgrade from 4 cores to 6 cores should be noticeable for software development.
For gaming I use a dedicated gaming PC. But I tested some games with the 5300M like Civilization VI, Tomb Raider, Subnautica, ... and the difference from the Radeon 5300M to the Radeon 455 in the MacBook Pro 2016 is huge
Very nice to being able to play some games directly on the Mac without having to switch to my PC gaming machine!That is not what I bought the machine for, but it is nice to have.
Otherwise, the phenomenon which some people call "ghosting" or "smearing" is happening on this machine, too, but not very noticeable for me. It only caught my attention because it was mentioned in this forum, it does not bother me.
Keyboard is fine. Much better than the butterfly keyboard on my MacBook Pro 2016" (which was replaced by the 2017 revision keyboard). It is definitely more silent and very precise. I hope that this keyboard does not fail. I am using a MacBook 12" 2017 in addition. I used it mainly in the time between end of November when I sold my MacBook Pro 2016" and now and I must say that I really liked the precision of this butterfly keyboard, it is not identical to the butterfly keyboards on the Macbook Pros!
Sound is great, but that was no surprise. The build quality of the device is very good. The display is gorgeous, the trackpad is perfect, ...
All in all I really like the machine. I will continue using it the next days and will keep you informed in case something changes.
One thing to note: the machine is more silent that the Core i9 versions I tested but it still is overall louder than my MacBook Pro 2016 that I have used the last three years before. This could be interesting for people wanting to switch from such a machine and only use it with moderate CPU load to heavy CPU load.
People who use the machine full tilt most of the time like video editing won't have a problem with the fan noise I assume. But in software development the CPU is more used in little bursts while editing/refactoring code, building code and then running tests/code, etc. So this is a different scenario which demands CPU power but not 100% of the time if I could make this clear.
Last edited:
I got my third 16" MacBook Pro yesterday. This time, I think I got a keeper
The first one was a base model Core i9 early December. It had quite noticeable ghosting/smearing which does not disturb me and the sound pop problem when playing videos. And it's fan noise when connected to my external 4K monitor either directly or via my Caldigit TS3+ dock drove me crazy, it made lots of noise with low CPU load even in clamshell mode.
Early January I tried another base model Core i9. The display was noticeable better, it had no sound pop problems, but the fan noise was still not acceptable to me. When the machine gets taxed I am ok with that but not when the machine is doing hardly anything at all! So I sent it back and decided to try the base model core i7.
What can I say, this one seems to be a keeper
I am doing software development in Java with it. But currently I am chilling because it is weekend and I did some simple testing. When I use it on my lap it is very very quiet, the fans mostly spin at 1700rpm which I cannot hear.
I connected it to my 4K 43" display via a USB-C to HDMI cable and tried clamshell mode and tried letting the laptop display turned on. As to be expected, non clamshell mode drove the energy consumption up.
But I do not need that, my 4K display is a huge 43" mammoth, so no need for the additional laptop display.
I could watch 4K 60fps on YouTube via Chrome for a long time without hearing anything from the fan, very nice. Since Mac OS has no hardware support for VP9 encoded videos via Chrome the CPU gets taxed by playing these videos. CPU load goes up to 30% but the fans stayed very silent and went from 1700 to about 2400 rpm while watching videos, very nice
I surfed around using Safari, wrote some documents, downloaded some games via Steam and Epic Launcher, tried Cinebench, ... When the machine gets under load (e.g. while gaming) the fans ramp up but I am ok with that, that's what they are there for.
The machine is generally more silent than the i9s, and the fans seem to be more silent under full load than the i9.
Tomorrow I will tax the machine with software development using IntelliJ Idea and later this week will also install Parallels 15 for doing some Visual Studio development under Windows 10. I hope the machine will perform well, too
At the moment I cannot tell if the CPU downgrade, the GPU downgrade or both were the reason for the better thermal performance, but ok.
I used a MacBook Pro 2016 15" before and the upgrade from 4 cores to 6 cores should be noticeable for software development.
For gaming I use a dedicated gaming PC. But I tested some games with the 5300M like Civilization VI, Tomb Raider, Subnautica, ... and the difference from the Radeon 5300M to the Radeon 455 in the MacBook Pro 2016 is huge
That is not what I bought the machine for, but it is nice to have.
Otherwise, the phenomenon which some people call "ghosting" or "smearing" is happening on this machine, too, but not very noticeable for me. It only caught my attention because it was mentioned in this forum, it does not bother me.
Keyboard is fine. Much better than the butterfly keyboard on my MacBook Pro 2016" (which was replaced by the 2017 revision keyboard). It is definitely more silent and very precise. I hope that this keyboard does not fail. I am using a MacBook 12" 2017 in addition. I used it mainly in the time between end of November when I sold my MacBook Pro 2016" and now and I must say that I really liked the precision of this butterfly keyboard, it is not identical to the butterfly keyboards on the Macbook Pros!
Sound is great, but that was no surprise. The build quality of the device is very good. The display is gorgeous, the trackpad is perfect, ...
All in all I really like the machine. I will continue using it the next days and will keep you informed in case something changes.
One thing to note: the machine is more silent that the Core i9 versions I tested but it still is overall louder than my MacBook Pro 2016 that I have used the last three years before. This could be interesting for people wanting to switch from such a machine and only use it with moderate CPU load to heavy CPU load.
People who use the machine full tilt most of the time like video editing won't have a problem with the fan noise I assume. But in software development the CPU is more used in little bursts while editing/refactoring code, building code and then running tests/code, etc. So this is a different scenario which demands CPU power but not 100% of the time if I could make this clear.
Nice, this is the post I've been waiting for. Can you post a screenshot of iStatsMenu as others have done using the direct connection and dock (with external monitor running)? This will show the GPU load.
Type of connection? Directly with a cable (and what cable) to the laptop or through some hub/dock/adapter?
Directly by cable was fine. I put in a dock after I posted that and still no problem.
Of course I can. But not today, I am tired now, it was a long day :-(
I will post iStatsMenu screenshots for direct connection to the 4K monitor via a USB-C to HDMI cable and for the connection via my CalDigit TS3+ dock which I have not tested yet.
[automerge]1579546433[/automerge]
Which base model do you mean? The core i9 or the core i7?
I just solved this issue. Bought a 2015 Pro 15 machine without dGPU. Fingers crossed somewhere along the way Apple does the same treatment to a 16" refresh. RIP dGPU, long live eGPU (if you need it anyway).
Same. I have the LG 27UL850-W 27 inch monitor connected by USB-C to my 8 core i9 16" base model and run with no problems (yes including 4k YouTube videos). The monitor even powers my laptop for the most part. I didn't perceive any heating issues until I read this thread. And sure if I stare at iStat long enough the computer does indeed get warmer, versus not having a monitor connected. But to call it hot and call out a design fault? Nope. And if the fans spin up under continuous hard load, well, what did people think the fans were there for? I suppose I could obsess, or just get on with work. I don't think it's reasonable to expect that running under full load conditions with extra monitors is going to the the same as low load conditions with only the built in monitor.
I don't think people don't know how electronics works. But looking at this specific use case when you connect an external display and the dGPU is consuming ~20W no matter if it's idle or under load is suspicious.
No, it isn’t. I believe 17w is the minimum that dGPU electronics draw. When using dGPU it’s not only the graphic processor that is drawing current, you have to account for dedicated memory too!
As should be obvious by now I am a strong opositor to having dGPUs on a laptop This thread just proves that I am not alone and Apple should deliver a 15/16 machine without dGPU as well as a 13/14 with dPGU. I believe there is market for such systems.
You're wrong because when using MacBook in clamshell mode the dGPU takes only 5W.
Radeon High side takes 4-5w with integrated panel on only. But check on iStat Menus the total energy on the CPU use section and you will be amazed... Its not 5W.
@pneves1975 is right. In clamshell mode the AMD dGPU uses around 18W when attached to a external monitor.
Unfortunately... nobody of you is either wrong or right. There is some serious messed up things going on with the dGPU drivers, and can be attributed to a lot of variants in displays, adapter, configurations. I did some testings, and here are the results. Please read until the end if you're really interested.
For the last few days I played around with two different 32" monitors, to try to find the "coolest" option for my MacBook. I'm almost always in clamshell mode while docked, and I want the real estate of a 2560x1440 monitor on a 32" panel. The two monitors are both LG:
32UD99 (32" 3840x2160 - 60hz - USB-C)
32GK850G (32" 2560x1440 - 165hz)
I didn't really have a problem with the 4K monitor: Great panel, gorgeous sharpness, nice viewing angle etc. Thing is, I usually play a lot of media for work or entertainment, and I noticed that pushing around 2560x1440 in HiDpi (which renders at 5120x2880 and scales down to panel resolution) in an already hot enclosure made the fans spin up quite a bit. I did some testing using 2560x1440 on the same panel and especially when in full screen the difference in watt consumption was noticeable, but of course all the text was blurry because it wasn't using native resolution. I decided to order the 32GK850G, which is a little inferior in quality due to the VA panel, but at least it's a 32" 2560x1440, and decently rated.
The monitor comes in, and I'm excited to get back at using a single cable to dock. I'm using a standard USB-C dock with power pass-through, HDMI connection to the monitor, and jack for audio. Not so fast: the same system and dGPU that was easily pushing 2560x1440 using only 5W on the 4K monitor, now it's using 18W on a native resolution monitor JUST TO STAY IDLE on the desktop. Of course the minimum usage made the fans spin up like crazy.
Tinkering around all night with both monitors, and one USB-C to DisplayPort cable order later, I finally get the hang of it: the dGPU for some strange reason hates the HDMI connection (which can only go up to 60hz on that monitor) and uses 18W just to idle. But, when connected to with a proper DisplayPort cable instead of the USB-C dongle, it can properly interface itself with the monitor and give me the coveted 5W. The strangest part of all of this? If I select 2560x1440 60hz in the system panel, it still jumps up to 18W, but If I go 85, 100, 120hz with SwitchResX it goes back to 5W. That doesn't make any sense.
So here are some results at 2560x1440 85hz. There are six major steps in this graphs:
1. Clamshell, idle on desktop.
2. Clamshell, streaming from Safari on Prime Video (The Office S01E01) in full screen.
3. Clamshell, playing a 1080p, 12GB BluRay rip on IINA in full screen.
4. Clamshell, back to idle on desktop.
5. Clamshell, switching to "Default for display" resolution in the system menu.
6. Undocking and using internal monitor.
Ignoring the temperatures of the CPU which can vary by a lot of factors, one thing is comparable for all of our systems: watt usage. And you can see in the system total, that for some configurations, you can be burning 27W for staying idle on a 2560x1440 monitor while the internal monitor (which being HiDpi is A LOT more pixels) can be pushed, with all of the system, for just 3W. I also noticed that using different types of interface to the monitor can give more choices both in the system display setting and in SwitchResX.
By the way I still find strange, as per my last post, that something is firing up on the side you plug the monitor in, or even if you just plug in a USB device like a hard drive. While the 15" didn't fire up the Thunderbolt controller if there wasn't any Thunderbolt connection to establish, it seems that the 16" does even for USB. If the system undocked is 3,70W, and the dGPU in idle is 4,90W, there seems to be another 5W lost by powering the Thunderbolt chip, even if the monitor is just using DisplayPort through USB-C.
So, after all of this testing, and to go back at the replies above my advice to everyone is: play around with different interface connections to your monitor, different cables and different resolution/refresh rates. The dGPU at idle should not burn more than 5W even at 4K. The lowest system total I managed to get in clamshell, 2560x1400 was those 13,40W.
A lot of people make it seem like this is not a big deal, but it is: The system can push work extremely efficiently while on internal monitor, as low as 3W. This means a lot less heat generated and a healthier system, and more importantly, a healthier battery. While I find now "acceptable" idling at 13W, it still takes your CPU to around 50-52°C and the batteries at 34-36°C, which is uncomfortably hot for their health. Also, 13W on a 100W rated system, is like burning 13% of your thermal capacity for doing nothing at all.