Become a MacRumors Supporter for $50/year with no ads, ability to filter front page stories, and private forums.

Eric_WVGG

macrumors 6502
Oct 25, 2016
389
747
gentrification fallout zone
In fact, Apple doesn't even need to add an additional display controller to the base chip. They could just allow the controller for the built-in display to instead be used to drive an external monitor.

Based on the specs of the M2 Mini (where they do essentally that, since there is no internal display), that would enable the Airs and base 14" MBP to support 6k@60 + 5k@60.

This could be implemented in a variety of ways, but my preference would be to to configure the laptop to automatically disable the internal display when two externals are attached.
dang dude NAILED IT
 

leman

macrumors Core
Oct 14, 2008
19,516
19,664
Though actually, to my prior point. The MacBook Air may indeed be wired differently. It may not be that this is possible on the 14" MacBook Pro's display wiring. I hope it is; It's a bit funky for the Air to have better external display support than the Pro, but the wiring may still be different to support this even if the chip is the same


I am sure you are correct about this.

What I find most interesting is that it proves that the need is real and that Apple is aware of it. M4 will likely include dual monitor support.
 

theorist9

macrumors 68040
May 28, 2015
3,880
3,059
I don't think internal wiring is a problem. Perhaps on M1, but M2 is definitely flexible enough. In the M2 Mac Mini you can drive two monitors through Thunderbolt ports, and that implies both display controllers have access to the routing matrix which allows sending display refresh data through either of the two Thunderbolt/USB-C ports. (We know this routing must exist because in laptops, you don't end up with one special USB/TB port that is the only one which can drive an external display.)
I believe the original question wasn't whether the laptops could be easily wired to support this, but whether the wiring was already in place such that this change could be made to the M2 Air's and M2/M3 MBP's via a software update. I suspect the answer to this is, alas, no. But it would be cool if it were otherwise.
 
Last edited:
  • Like
Reactions: Tagbert

ArkSingularity

macrumors 6502a
Original poster
Mar 5, 2022
928
1,130
Frankly, I'm thrilled to see that Apple made this change (will help people who use MacBooks for office setups quite a bit, provided they're okay with clamshell mode).

I am, however, quite surprised that Apple actually did this. I did not see this one coming.
 

theorist9

macrumors 68040
May 28, 2015
3,880
3,059
I hope the clamshell mode requirement doesn't create problems. I've read there can be issues with clamshell.

I'd personally like to see the user have the option to use this without clamshell, where the internal display would automatically deactivate when connecting a 2nd external, and reactivate when disconnecting it. To give the user additional control, they should be allowed to switch which display is active in Display Preferences
 
Last edited:

ArkSingularity

macrumors 6502a
Original poster
Mar 5, 2022
928
1,130
I hope the clamshell mode requirement doesn't create problems. I've read there can be issues with clamshell.

I'd personally like to see the user have the option to use this without clamshell, where the internal display would automatically deactivate when connecting a 2nd external, and reactivate when disconnecting it. To give the user additional control, they should be allowed to switch which display is active in Display Preferences
I wouldn't be surprised if someone creates a program to enable this sort of functionality (if the APIs allow it, at least).
 

mr_roboto

macrumors 6502a
Sep 30, 2020
856
1,866
I run my 16" M1 Max in clamshell a lot and there's no problems with it. I put much heavier loads on it when it's in clamshell, too.

I don't think there's been a real clamshell problem for a long time. The airflow path is basically the same in clamshell mode, and Apple doesn't depend on heat going up through the keyboard and palmrest at all. In fact, they put in small amounts of thermal insulation to discourage heat from leaving the machine through those paths since high temps on any surface likely to be in contact with skin are not good.
 
  • Like
Reactions: Chuckeee

mr_roboto

macrumors 6502a
Sep 30, 2020
856
1,866
Hector Martin tooted about this, saying it's possible this could be backported to M2 systems too.


(Technical details: the M1 has one primary display controller hardwired to eDP and one external display controller connected to a crossbar that leads to the Type C DP PHYs and the Thunderbolt packetizer. The M2 moves the primary controller to the crossbar too, so it can be repurposed to output via the Type C ports too. M3 is presumably identical. All baseline chips still only have 2 display controllers total, and this is cost driven and not easy to increase without a major die size increase since the display controllers are huge. The change to the crossbar, on the other hand, is essentially "free" and makes perfect sense to do.)

eDP = embedded DisplayPort. M1/M2/M3 all have an eDP port dedicated to driving internal displays, and on the Mac Mini, it's connected to a DP-to-HDMI translator chip to provide the HDMI output.

So, in M1, the internal display refresh engine is routed only to the eDP port, but in M2 and M3, it can also connect to the crossbar which serves display data to Type C ports.
 

senttoschool

macrumors 68030
Nov 2, 2017
2,626
5,482
I am sure you are correct about this.

What I find most interesting is that it proves that the need is real and that Apple is aware of it. M4 will likely include dual monitor support.
I assume you meant that M4 will support 3 total screens (including built in display). If so, I'm not convinced. I think there is a small vocal minority that wants 3 total screen support for the Macbook Air.

I think M3's approach is good enough.

I, for one, would not want to pay for the extra transistors (as many as 2 P cores) for something I will never use.
 

ArkSingularity

macrumors 6502a
Original poster
Mar 5, 2022
928
1,130
yeah, Intel models — and even the old G4s — had a reputation for overheating in clamshell mode.

But honestly when have you felt an ARM Mac even get warm, and even if it did I'm pretty sure they're smart enough to throttle.
Yea, they have several sensors throughout the main board to measure temperature at various parts of the chassis. It's not going to get dangerously hot, the worst thing that could happen is a tiny performance loss from a marginal amount of thermal throttling.

(Frankly, that's probably an easy enough problem to solve in clamshell anyway just by having a fan blowing in the vicinity).
 

Chancha

macrumors 68020
Mar 19, 2014
2,307
2,134
Yea, they have several sensors throughout the main board to measure temperature at various parts of the chassis. It's not going to get dangerously hot, the worst thing that could happen is a tiny performance loss from a marginal amount of thermal throttling.

(Frankly, that's probably an easy enough problem to solve in clamshell anyway just by having a fan blowing in the vicinity).
But I have to say, in the long run, the internal display panel and the battery cells are better off not getting that extra heat, if longevity is to be considered.

I see this M3 clamshell mode being a compromise solution for those who need it, instead of an overall ideal one for everybody. The fact that you lose access to the internal KB and trackpad and even the TouchID makes this even more obvious.
 

theorist9

macrumors 68040
May 28, 2015
3,880
3,059
Hector Martin tooted about this, saying it's possible this could be backported to M2 systems too.




eDP = embedded DisplayPort. M1/M2/M3 all have an eDP port dedicated to driving internal displays, and on the Mac Mini, it's connected to a DP-to-HDMI translator chip to provide the HDMI output.

So, in M1, the internal display refresh engine is routed only to the eDP port, but in M2 and M3, it can also connect to the crossbar which serves display data to Type C ports.
Interesting...so he thinks the physical configuration to do this might already be present in the M2 Airs, and thus they could be given this capability purely through a firmware+software update. How likely do you think this is (that the physical capability is already there, not that Apple will actually do it)?
 
  • Like
Reactions: Chuckeee

theorist9

macrumors 68040
May 28, 2015
3,880
3,059
I wonder if they'll also do this for the M# Pro MBP -- allow it to drive three externals by routing the internal display controller to one of them. Three externals is the sweet spot for many users who need ample display space, since it gives a central monitor you can look at head-on, and one on either side. IMO it's ergonomically superior to having two externals where you need to look either left or right.

My guess is they were motivated to enable two externals for the Air by enterprise customers whose standard setup is 2 externals, and who buy Airs for their employees in bulk. There's probably less pressure to support 3 externals with the M# Pro MBP, since 3 externals is less common.
 
Last edited:
  • Like
Reactions: Chuckeee

jdb8167

macrumors 601
Nov 17, 2008
4,859
4,599
I hope the clamshell mode requirement doesn't create problems. I've read there can be issues with clamshell.

I'd personally like to see the user have the option to use this without clamshell, where the internal display would automatically deactivate when connecting a 2nd external, and reactivate when disconnecting it. To give the user additional control, they should be allowed to switch which display is active in Display Preferences
I’ve run both my M1 and M2 MacBook Airs in clamshell for years. Absolutely no problems. The MBAs have very effective thermal throttling.

The only issues will be the perennial third-party external display issues.
 

dmccloud

macrumors 68040
Sep 7, 2009
3,138
1,899
Anchorage, AK
Though actually, to my prior point. The MacBook Air may indeed be wired differently. It may not be that this is possible on the 14" MacBook Pro's display wiring. I hope it is; It's a bit funky for the Air to have better external display support than the Pro, but the wiring may still be different to support this even if the chip is the same

The wiring is likely not the same because the MBA display itself is different. The MBP displays support Mini LED/local dimming, ProMotion, and 1600 nits peak brightness. The MBA displays lack all of those features and apparently only has a 500 nit peak brightness (according to Apple's tech specs for the new M3 Airs.
 

dmccloud

macrumors 68040
Sep 7, 2009
3,138
1,899
Anchorage, AK
But I have to say, in the long run, the internal display panel and the battery cells are better off not getting that extra heat, if longevity is to be considered.

I see this M3 clamshell mode being a compromise solution for those who need it, instead of an overall ideal one for everybody. The fact that you lose access to the internal KB and trackpad and even the TouchID makes this even more obvious.

With Apple Silicon, the additional heat generated in clamshell mode would be negligible at most, and definitely not enough to noticeably or significantly impact the longevity of the device. The issues people saw running Intel-based machines in clamshell mode with respect to heat generation have largely been eliminated with Apple Silicon.
 

Chancha

macrumors 68020
Mar 19, 2014
2,307
2,134
With Apple Silicon, the additional heat generated in clamshell mode would be negligible at most, and definitely not enough to noticeably or significantly impact the longevity of the device. The issues people saw running Intel-based machines in clamshell mode with respect to heat generation have largely been eliminated with Apple Silicon.
When Apple Silicon was new, I did extensive tests on a base M1 MBA, and measured the surface temp of the keyboard aluminum frame to reach 45 degrees C before throttling kicks in (the CPU P-Cores are around 98 degrees at this moment). I have since moved on real workflows to Apple Silicon Macs with fans but the point stands.

I wouldn't say I am qualified to say sustained 45 degrees is negligible for clamshell mode or not, but, using laptops for a while now I can see this sort of heat generated and concentrated in a small area against the display panel is likely to cause discoloration inside, if not doing something to the coating along with what oily stains the user leaves on the surface. And among the dozens of MacBooks in our studio (both Intel and AS), the ones where the user use clamshell modes have their battery health dropping much faster, but it is of course unclear if heat is the issue or is it the fact the power is constantly plugged in.

My take away is that, Apple designed their Intel MacBooks not that much different than the AS MacBooks in terms of thermals, it just happens that the Intel chips heats up so quickly and then sustain at high temps way too often for the issues to be much more common among users. What Apple Silicon brought to the table is a much lower starting point in temps so we got much more headroom to work with, so much so the MacBook Airs can become fanless and most users don't notice a difference. But with clamshell mode especially sustained usage, well, time will tell.
 

Tagbert

macrumors 603
Jun 22, 2011
6,254
7,280
Seattle
When Apple Silicon was new, I did extensive tests on a base M1 MBA, and measured the surface temp of the keyboard aluminum frame to reach 45 degrees C before throttling kicks in (the CPU P-Cores are around 98 degrees at this moment). I have since moved on real workflows to Apple Silicon Macs with fans but the point stands.

I wouldn't say I am qualified to say sustained 45 degrees is negligible for clamshell mode or not, but, using laptops for a while now I can see this sort of heat generated and concentrated in a small area against the display panel is likely to cause discoloration inside, if not doing something to the coating along with what oily stains the user leaves on the surface. And among the dozens of MacBooks in our studio (both Intel and AS), the ones where the user use clamshell modes have their battery health dropping much faster, but it is of course unclear if heat is the issue or is it the fact the power is constantly plugged in.

My take away is that, Apple designed their Intel MacBooks not that much different than the AS MacBooks in terms of thermals, it just happens that the Intel chips heats up so quickly and then sustain at high temps way too often for the issues to be much more common among users. What Apple Silicon brought to the table is a much lower starting point in temps so we got much more headroom to work with, so much so the MacBook Airs can become fanless and most users don't notice a difference. But with clamshell mode especially sustained usage, well, time will tell.
You may want to turn on Optimized Charging to help keep them from charging to 100% all the time.

edit: I use Aldente to set my maximum to 80% because I prefer the manual control over the automated control.
 
Last edited:

mr_roboto

macrumors 6502a
Sep 30, 2020
856
1,866
Interesting...so he thinks the physical configuration to do this might already be present in the M2 Airs, and thus they could be given this capability purely through a firmware+software update. How likely do you think this is (that the physical capability is already there, not that Apple will actually do it)?
It's 100% likely.

As Hector explained, the M1 Mini can drive two displays, but one must be connected to the HDMI port and the other to one of the type C ports. This restriction exists because the HDMI port is driven by a DisplayPort-to-HDMI conversion chip on the Mac Mini's motherboard, and the DP side of that chip is connected to the M1 SoC's eDP display output. That M1 eDP port is what drives the internal display on all other M1 products, and in M1, it's the only output path the first of the two display refresh controllers can use.

The other display out on the M1 Mini is either type-C port. This means there's a crossbar switch to connect the second display refresh controller to either type-C.

In the M2 Mini, you no longer have to use the HDMI port. Instead, if you wish, you can connect two displays to the two type C ports. That proves the M2 supports disconnecting the first controller from the eDP output path and instead connecting it to the type-C crossbar switch, and that's all you need (at the hardware level) to support this feature.
 

theorist9

macrumors 68040
May 28, 2015
3,880
3,059
It's 100% likely.
...That proves the M2 supports disconnecting the first controller from the eDP output path and instead connecting it to the type-C crossbar switch, and that's all you need (at the hardware level) to support this feature.
Sorry, I'm not following. As you know, the question at hand is whether the M2 Airs and 13" M2 MBP can be given this capability solely with a firmware+software update, i.e., without any changes at the hardware level. What you're describing are instead the changes Apple would have needed to make at the hardware level for this to work. The fact that those changes are minimal doesn't prove they've actually been made to the M2 (non-Pro/Max) laptops.

I don't know whether they made this hardware adjustment to the M2 laptops or not. But if they did, why bother doing it, only to not offer the associated capability it provides?
 
Last edited:

mr_roboto

macrumors 6502a
Sep 30, 2020
856
1,866
Sorry, I'm not following. As you know, the question at hand is whether the M2 Airs and 13" M2 MBP can be given this capability solely with a firmware+software update, i.e., without any changes at the hardware level. What you're describing are instead the changes Apple would have needed to make at the hardware level for this to work. The fact that those changes are minimal doesn't prove they've actually been made to the M2 (non-Pro/Max) laptops.

I don't know whether they made this hardware adjustment to the M2 laptops or not. But if they did, why bother doing it, only to not offer the associated capability it provides?
I'm both describing the changes and stating that the M2 Mac Mini's display output capabilities prove they're present. Specifically, the display controller which formerly (in M1) could only drive the eDP output port can now be rerouted to drive one of the type-C connectors instead. Since all the required circuits are internal to the M2 SoC (this is all on-silicon plumbing), they exist in laptops too.

Why did they do it but not take advantage of it in this way immediately? (note, they did take advantage of it on the M2 Mini.) Who knows, but I did raise the idea (right before they announced M3 notebooks would get this ability, heh) that the user experience would be a bit bumpy on laptops in certain scenarios, and maybe Apple just didn't want to go there. Perhaps that got overridden by the practical consideration that so many corporate buyers really love twin external monitors on employee desks, and forcing them onto Mx Pro or Max was limiting sales.
 
  • Like
  • Love
Reactions: Sikh and theorist9

theorist9

macrumors 68040
May 28, 2015
3,880
3,059
I'm both describing the changes and stating that the M2 Mac Mini's display output capabilities prove they're present. Specifically, the display controller which formerly (in M1) could only drive the eDP output port can now be rerouted to drive one of the type-C connectors instead. Since all the required circuits are internal to the M2 SoC (this is all on-silicon plumbing), they exist in laptops too.

Why did they do it but not take advantage of it in this way immediately? (note, they did take advantage of it on the M2 Mini.) Who knows, but I did raise the idea (right before they announced M3 notebooks would get this ability, heh) that the user experience would be a bit bumpy on laptops in certain scenarios, and maybe Apple just didn't want to go there. Perhaps that got overridden by the practical consideration that so many corporate buyers really love twin external monitors on employee desks, and forcing them onto Mx Pro or Max was limiting sales.
Ah, I get it now--the part I didn't understand earlier is that this is internal to the SoC, and no additional rewiring is required to get it to work in the M2 laptops (M2L's). Thus whatever hardware changes they had to make to the M2 SoC to allow it to drive two externals in the M2 Mini is also sufficient for the M2L's

This also provides a possible explanation for why they didn't give the M2L's this capabilty: If this required laptop-specific rewiring, then it wouldn't have made sense to do this to the M2L's and then not give them that capabilty. However, since this is something they had to do anyways for the M2 Mini, I can see them then deciding not to extend that capability to the M2L's, for business reasons, to maintain product differentiation vs. the M2 Pro laptops.

Now Apple realizes the business case favors giving the M3L's (or at least the M3 Airs) that capability (as both of us have guessed, this likely comes from enterprise customers). My guess is they will also update the firmware/software on the M3 MBP, since otherwise the latter would be left less capable than the M3 Airs in this regard.

But they might not provide it to the M2L's, to maintain product differentiation vs. the M3L's. I suppose it depends on how much pressure they get from enterprise clients who want to save money by buying the M2 Air instead of the M3 Air--and how much Apple wants to push those toward the M3.
 
Last edited:
Register on MacRumors! This sidebar will go away, and you'll see fewer ads.