M3 Die Shot suggests display engines take up nearly the same area as P-cores

[ArkSingularity]

The key point is not whether or not you are connected to power when using externals, but whether you have easy access to it if needed. And you do. For instance, whenever I'm giving a lecture using a projector, I can always easily plug in my laptop. That's why the efficiency of the external controller(s) is not as critical as that of the internal.

That's true, but we also don't necessarily know how much smaller the display controller could would be if they had designed it to be more compact, nor do we really know how much power is saved by the design they have now. We're sort of relying on hypotheticals here, it could very well be that Apple had good reasons for doing it the way they did it.

A good compromise, in my opinion, would have been for them to allow one display controller to drive two non-6K displays in software, but at least we have displaylink that can do that as a third-party solution (not as good as native, but at least it's possible).

Speaking of access to charging, there were actually scenarios where I needed easy access to charging and didn't have it. Most of the time it's available, but I also am glad that the computer doesn't run itself down unreasonably fast when it's not (I couldn't say the same of the 2012 I was using in college). Nowadays I actually do hook up my laptop to an external monitor without it being hooked up to the charger fairly often.

 

[theorist9]

Feature mix seems exactly right to me. I highly doubt people buying base M-series chips are in need of external monitors at all, let alone more than one. Battery life on the other hand is exactly what they need, because its not at a desk plugged in to monitors (and power).

I'll repeat what I said above:

I recall one poster who said his boss was thinking of converting to Macs when AS was released, but their budget only allowed for Airs, which would have worked great, except all their workstations use dual monitors. And dual monitors is pretty common in many offices these days.

 

[pgolik]

I mean, it's always nice to be able to skip bringing the charger if you're bringing a laptop around a lot. Constantly having to charge was definitely an added inconvenience for me in college.

Definitely. Not every lecture room I work in has a conveniently placed free socket for a charger. And you don’t want to trip on a power cable snaking on the floor in the middle of your lecture.

 

[salamanderjuice]

3840×2160 = 8,294,400 px (100%) 4K UHD
4480×2520 = 11,289,600 px (136.1%) 4.5K iMac
5120×2880 = 14,745,600 px (177.7%) 5K ASD
6016×3384 = 20,358,144 px (245.4%) 6K XDR

Do those Intel display engines support a 6K display at 60 Hz?

Yes, max res of the i7 1370p for example is 7680x4320 @ 60 Hz. AKA 8K.

 

[ArkSingularity]

Definitely. Not every lecture room I work in has a conveniently placed free socket for a charger. And you don’t want to trip on a power cable snaking on the floor in the middle of your lecture.

This, or someone else is already using the outlet or the cord isn't long enough. I encountered this pretty often.

 

[theorist9]

That's true, but we also don't necessarily know how much smaller the display controller could would be if they had designed it to be more compact...

Yeah, and that's why it's unfortunate the dies are not annotated to show the size of those internal controllers.

Speaking of access to charging, there were actually scenarios where I needed easy access to charging and didn't have it.

Were these cases when you were driving an external monitor? I'm sure that can happen sometimes, but I would think it would be uncommon.

 

[ArkSingularity]

Were these cases when you were driving an external monitor? I'm sure that can happen sometimes, but I would think it would be uncommon.

As a matter of not being able to get to charging at all when connected to a display, maybe only twice. As a matter of it simply being inconvenient because I didn't have a spare outlet or a spare USB-C port, I've had that happen more often.

I'm pretty mobile with my workflows, I move things around a lot. Probably more so than the average person.

 

[AlastorKatriona]

I'll repeat what I said above:

I recall one poster who said his boss was thinking of converting to Macs when AS was released, but their budget only allowed for Airs, which would have worked great, except all their workstations use dual monitors. And dual monitors is pretty common in many offices these days.

These are not office computers. They're not replacements for the Windows 95 boxes plugged into the VGA monitors that populate office spaces. That's a poor, poor, poor argument and comparison.

 

[salamanderjuice]

These are not office computers. They're not replacements for the Windows 95 boxes plugged into the VGA monitors that populate office spaces. That's a poor, poor, poor argument and comparison.

No? It's really common for businesses to deploy laptops and docks so employees can take them to meetings or home if need be. Something like a M1 MBA should be able to chew through basic office tasks.

 

[pgolik]

Were these cases when you were driving an external monitor? I'm sure that can happen sometimes, but I would think it would be uncommon.

In my experience teaching at university - yes, it happens often enough, especially in older buildings. There may be an extension cord available, but that’s awkward. It really feels great to be able to present without having to rig some Rube Goldberg contraption every time. For me, it was annoying enough that before getting the M3P, I took my iPad with a dongle instead of the old Mac (even though Keynote on iOS sucks). With that I could get by with a small powerbank (the battery life of the M1 iPad pro driving an external isn’t that great, 5 hours give or take).

 

[theorist9]

These are not office computers. They're not replacements for the Windows 95 boxes plugged into the VGA monitors that populate office spaces. That's a poor, poor, poor argument and comparison.

Actually, it was an excellent argument. You simply didn't understand it. I'm talking about business use as it is now, where dual office externals are common, and where busineses would like to issue employees laptops instead of desktops so they can move easily between office and home. Yet for some bizarre reason you're referencing 1995, when people used single VGAs, and didn't commonly work from home.

The fact that the best you could do was to say "poor, poor, poor", rather than actually give a substantive response, shows how vapid your counterargument is. Seriously, do you somehow think the strength of your argument is in proportion to how many times you use that adjective?

 

[name99]

Here's another one, do they compress that framebuffer cache? And will we ever find out ;-)

They appear to:

US10049427B1 - Image data high throughput predictive compression systems and methods - Google Patents

Systems and methods for improving operation of an electronic device that includes a processing pipeline. The processing pipeline includes a prediction block that determines prediction residuals based on difference between uncompressed image data corresponding with a pixel tile and prediction of...

patents.google.com

The display controllers do a ton of stuff that probably differs across products.
On watches, for example, they store a number of the future frames (eg tick by tick update of the screen) so that the entire SoC can sleep and the display controller rolls out those successive frames at the right times.
On phone and laptop they remap colors depending on ambient light (brightness and white point). If Apple control the screen (so phone and laptop) they also send a lossless compressed image to the screen so that the cost of "long distance" (ie off chip) data transfer is reduced, and the result is decompressed as close to the screen as possible.
Also on Apple controlled screens they track brightness across the screen (probably not at pixel level, but at some small tile level) and keep statistics, so that they know approx how the screen is aging, and compensate for that aging. There's similar instantaneous color correction for ambient temperature of the screen.

 

[name99]

The thing that I'm left wondering about from this video is the mention of "black silicon" or the "seemingly unused" portions of silicon on the chip. That seems expensive in a rising-wafer-cost world.

Does anyone have an idea there? In my layman's thinking, I have an idea that if a company ever started stacking dies (vs using chiplets), they'd need some kind of design for an elevator shaft between floors - not for moving people between floors, but bits.

Some huge blocks that I have not seen marked out are media (ISP, encode, decode).
Which makes me, not exactly skeptical, but willing to be flexible on the claims of the areas that are supposed to be display controller.
Other large logic blocks that aren't being considered are obvious things like SEP, not very obvious things like the system wide power and communications control, and almost unknown things like the smart DMA (do things like compress, encrypt and otherwise perform SmartNIC functions on network packets).

As a more general point, where area cannot be used for "real" logic (eg in the boundaries between two irregularly shaped block) that area is frequently fabbed to act as a capacitor, to help with the overall SoC capacitance (ie to hold small amounts of excess charge to handle very short term bursts of excess current).

 

[name99]

The key point is not whether or not you are connected to power when using externals, but whether you have easy access to it if needed. And you nearly always do. For instance, whenever I'm giving a lecture using a projector, I can always easily plug in my laptop. That's why the efficiency of the external controller(s) is not as critical as that of the internal.

Yeah, but it's a question of convenience.
When you know that power is not an issue, fiddling with a power connector is one more hassle that's nice to ignore.

(And of course in Apple's optimal world, you'd probably have an Apple TV connected to the projector so that you could display to it without having to plug in even one cable...)

 

[name99]

I'll repeat what I said above:

I recall one poster who said his boss was thinking of converting to Macs when AS was released, but their budget only allowed for Airs, which would have worked great, except all their workstations use dual monitors. And dual monitors is pretty common in many offices these days.

And I'll repeat what I've said a dozen times which is that I call BS.
Because if that boss were in the slightest bit competent then the dual screen situation could easily have been handled via either wireless, or via USB monitors.

If you want to connect two monitors AND it's very important to you that the screens are HDR 120Hz 5K or the other fanciness that can only be accessed by the "official" display ports then you're not doing MacBook Air work...

 

[name99]

Yeah, and that's why it's unfortunate the dies are not annotated to show the size of those internal controllers.


Were these cases when you were driving an external monitor? I'm sure that can happen sometimes, but I would think it would be uncommon.

A simple example is my friend where I set up for her a second monitor in her home office.

She has a laptop power supply in her bedroom, but there was no need to bother with another power supply in the home office, just set up the monitor and plug the HDMI cable into her M1 Macbook Pro when she needs to use the large screen.

Amory Lovins, in the context of building efficiency, talks about how when you cross certain cliffs you unlock various "second order" savings because stuff you used to need you no longer need.
A dedicated laptop power supply in the home office is an (admittedly minor) example of this idea. A slightly more serious one might be if you had to have a box fan or something in that same home office, and now once again you don't need that.

 

[ArkSingularity]

She has a laptop power supply in her bedroom, but there was no need to bother with another power supply in the home office, just set up the monitor and plug the HDMI cable into her M1 Macbook Pro when she needs to use the large screen.

This is the exact scenario I find myself in often.

 

[Chancha]

Look, I myself manage a small music / media related business and I make purchasing decisions also. Sometimes I do wish the minis and Air we deployed could have more external display capability, like they used to when on Intel.

However;
this doesn’t mean I cannot see the validity of Apple‘s decisions here. I would have come to the exact same conclusion as they did: close to no one use more than 2 display on base model Macs. It is so insufficiently small that they “deserve” to be upsold to the Mx Pro models. Putting more display buffers on die of the base M chip, conversely, actually takes away silicon budget for the other 99% targeted users.

 

[JPack]

Feature mix seems exactly right to me. I highly doubt people buying base M-series chips are in need of external monitors at all, let alone more than one. Battery life on the other hand is exactly what they need, because its not at a desk plugged in to monitors (and power).

If nobody is connecting an external display, then Apple wouldn't need to emphasize support for 6K monitor.

Not sure where you've been living for the past three years. Everyone is doing remote work or hybrid work. If you're at a desk, why wouldn't you have a monitor?

 

[theorist9]

And I'll repeat what I've said a dozen times which is that I call BS.
Because if that boss were in the slightest bit competent then the dual screen situation could easily have been handled via either wireless, or via USB monitors.

If you want to connect two monitors AND it's very important to you that the screens are HDR 120Hz 5K or the other fanciness that can only be accessed by the "official" display ports then you're not doing MacBook Air work...

I remember you. You're the guy I caught lecturing (and insulting) someone about physics principles when you have no understanding of physics yourself:

https://forums.macrumors.com/threads/m2-max-geekbench-scores-surface-again-online-ahead-of-release-in-2023.2372999/page-4?post=31794239#post-31794239

https://forums.macrumors.com/threads/m2-max-geekbench-scores-surface-again-online-ahead-of-release-in-2023.2372999/page-6?post=31797562#post-31797562

And now you're attempting to characterize my post as "BS". So no, I'm not inclined to take anything you say on technical matters seriously, given your history of criticizing people who understand more than you do.

Yes, there are aftermarket kludges you can use to run 2 externals on an Air (see https://www.macworld.com/article/67...ternal-displays-to-apple-silicon-m1-macs.html ), but a boss is hardly "incompetent" if he doesn't want to rely on the continued function of such kludges—which are not officially supported by Apple—to keep his business running. It's striking, but at the same time hardly surprising, that you don't understand that.

 

[ArkSingularity]

To be fair, I can see both sides of the debate. Most people who are connected to an external display could (or will) connect to power in most cases. Theorist9 has a valid point. I still personally find it very convenient that the energy consumption is low even when connected to a display (I actually do use my laptop like this fairly frequently), but I would hardly say it's a dealbreaker if it were to consume more.

I don't think that Apple would have put two external display controllers into the base model anyway though (even if they were to design smaller display controllers for them). That, however, is just my speculation.

 

[theorist9]

Theorist9 has a valid point.

Only very rarely .

To be fair, I can see both sides of the debate. Most people who are connected to an external display could (or will) connect to power in most cases....I still personally find it very convenient that the energy consumption is low even when connected to a display (I actually do use my laptop like this fairly frequently), but I would hardly say it's a dealbreaker if it were to consume more.

I don't think that Apple would have put two external display controllers into the base model anyway though (even if they were to design smaller display controllers for them). That, however, is just my speculation.

Plus this whole discussion about efficiency may be moot. After all, as you've said, we're just speculating that the internal controller may be more efficient, for the reasons we've been discussing. But the difference might not be significant, in which case the whole question of battery vs. plugged for external displays becomes unimportant (at least so far as the controller is concerned).

I also wonder if the additional GPU power to run an external is more significant than any added demand from the controller.

Ultimately, I think the efficiency discussion was perhaps a distraction (though an interesting one) from the essence of—from what really underlies—what's being argued, which is this: To the extent Apple didn't supply an additional display controller on the Base M chip, was it for inherent product optimization reasons, or for business reasons?

I.e., did they do it because they thought leaving it out would lead to better Base-M products (Base MBP, Air, Base Mini, iMac), or because they wanted to maintain market segmentation between the Base and Pro lines? Or was it both?

[Even though the discussion has focused on the Air, it's really about all the Apple products that use the Base-M chip.]

I don't know the answer. But if we knew an additional 4k controller would be physically large, that would suggest product optimization. If we knew it's physically small, that would suggest business segmentation.

 

[ArkSingularity]

Ultimately, I think the efficiency discussion was perhaps a distraction (though an interesting one) from the essence of—from what really underlies—what's being argued, which is this: To the extent Apple didn't supply an additional display controller on the Base M chip, was it for inherent product optimization reasons, or for business reasons?

I.e., did they do it because they thought leaving it out would lead to better Base-M products (Base MBP, Air, Base Mini, iMac), or because they wanted to maintain market segmentation between the Base and Pro lines? Or was it both?

Yea, and that one is really hard to say. If I had to take a wild guess, I'd say it was probably a little bit of both.

 

[mr_roboto]

My first real "holy **** I'm in the future" moment with Apple Silicon was when I started futzing with the resolution scaling, and then plugged my laptop into an external display.

None of the bouncing around or flickering like on Intel and even PowerPC Macs, just seamless changes. Dang.

If any of that has to do with Apple's oversized display controllers — and again, this is a thing we've been living with for decades — that is absolutely a valid trade-off against running four displays on a flippin' $999 Macbook Air.

It's not the display controllers, it's software. On an x86 PC or Mac, there's a lot of layers of GPU driver, OS, and firmware code involved in video resolution switching, and the layers weren't necessarily written by the same organizations with the same attention to detail. The result is what you see: video mode switches are often slow and produce visible weirdness.

With Apple Silicon, Apple writes the entire driver and firmware stack. They put in some effort to make it all work a lot more smoothly and efficiently, which is why video mode switches are much faster and usually glitchless.

I also wonder if the additional GPU power to run an external is more significant than any added demand from the controller.

The proper way to think about this is that GPUs use power to draw into buffers residing somewhere in RAM. They aren't too closely coupled to how many displays happen to be around. While an external display is showing a static image that's not being recomputed at all each refresh interval, it needs 0W of additional GPU power.

This is true (or can be) even for PC-style discrete GPUs, but you might not think it since I've long heard of dumb driver/firmware stuff producing the illusion of something else. For example, you might run into driver/firmware stacks that force the GPU to clock higher as long as there's more than 1 display attached, even while there's no actual demand for more GPU throughput.

 

[Confused-User]

Ultimately, I think the efficiency discussion was perhaps a distraction (though an interesting one) from the essence of—from what really underlies—what's being argued, which is this: To the extent Apple didn't supply an additional display controller on the Base M chip, was it for inherent product optimization reasons, or for business reasons?

I.e., did they do it because they thought leaving it out would lead to better Base-M products (Base MBP, Air, Base Mini, iMac), or because they wanted to maintain market segmentation between the Base and Pro lines? Or was it both?

In the real world, for any given feature, the answer will always be "both", because the two are not separable.

They set a budget for chip size (cost, really, but it's mostly the same thing, until the advent of chiplets) and then decide what to spend that budget on. Sometimes someone will go to bat with management to argue that a certain feature is worth increasing the budget for. Sometimes they might even get their way.

In your terms, the budget is a "business reason". Deciding what to spend it on is "product optimization".