ARM based MacBook Pro discussion [merged]

[awesomedeluxe]

Your comments are well thought out. It's made me envision a thinner 16 inch MacBook Pro. The base of the machine could have a few millimeters shaved off by removing the tapering. It will probably look more like the iPad Pro. Hopefully with the reduction in tapering this will allow for more space in the display for a better camera & Face ID? All that + mini LEDs, the battery life should be great!

Thanks! I would love to see an iPad Pro style design for the new machines, where the display can be raised an inch or two off the keyboard. I'm a little surprised we haven't seen more machines like this.

Apple could just do what Microsoft is doing with their Surface Book line. The screen would hold the AX chip part while the base holds the dGPU. They can still make it almost as thin as the current MBP 13/16.

Whether they make it detachable or not depends on what they do the with OS and the hinge. Hopefully they will make a hybrid OS that is suitable for touch and desktop. Slowly updating iPadOS might not be the best move with Microsoft coming out with Win10X and Google with Fuschia (If Google ever releases it)

GPU might finally be a priority for Apple. With their push for gaming (Apple Arcade) and their new commitment towards creative professionals, we might see more GPU improvements. The eventual arrival in AR (glasses) might progressively move them towards VR, that needs a good GPU to perform well.

I think we'll find out more at WWDC this year. Interesting to see what Apple will do.

They definitely could, and a lot of people would be happy to see that happen. I actually discussed it a bit in the Waiting for the 2020 MBP thread, although I envisioned it as running macOS and iPadOS separately. A hybrid OS would solve the need for duplicate hardware.

But it leaves you wondering if Apple would be willing to let such a machine exist and have access to apps outside the walled garden. Hopefully they don't care too much about potentially cannibalizing iPad app sales if they're already selling those machines for 4-5k.

I really expect Apple to design ARM CPUs specific for laptops and desktops and doesn't re-use phone CPUs.

You mean like a totally different architecture? I'm not certain what the upside would be. There's plenty you can do to scale up the iPhone CPU. My reading of the current "Project Kalamata" rumors is Apple is using the same cores, just more of them. I'm sure they get a lot of mileage just going from two to eight high performance cores - but they can also make other tweaks like adding more/faster cache or nudging the clockspeeds up a little without having to radically redesign anything.

 

[sublunar]

I can't see Apple partnering with AMD for a GPU in an ARM Mac, certainly not for a 12" MacBook like model which could end up being fanless. Makes it pointless if Apple is looking to entirely control that side of the supply chain.

Drop an A14X type CPU into it - maybe the same as in a 2021 iPad Pro - call it an M1 CPU for marketing differentiation because you've got extra cores in it for example. I think 8Gb of RAM would be baseline for this model to be able to do multitasking sufficiently, especially if Pro-motion screens are likely to be considered for the ability to down clock the display refresh to save battery rather than go all the way up to 120Hz for anything other than video playback.

For specific tasks the ARM architecture is great - for example video decoding and encoding - which is why something like FCPX or Logic might go well.

Intel might be better for other jobs but we would have to see. Releasing a boutique product would enable Apple to judge the market and give developers a real product to code towards before releasing something more mainstream and usable like a 15" laptop variant or - indeed - a 23" iMac.

I think Apple would be looking for developers to recompile their apps for ARM and deliver solely via a specific App Store at least initially rather than seeking emulation for Intel on the ARM CPUs.

 

[DanMan619]

I can't see Apple partnering with AMD for a GPU in an ARM Mac, certainly not for a 12" MacBook like model which could end up being fanless. Makes it pointless if Apple is looking to entirely control that side of the supply chain.

Drop an A14X type CPU into it - maybe the same as in a 2021 iPad Pro - call it an M1 CPU for marketing differentiation because you've got extra cores in it for example. I think 8Gb of RAM would be baseline for this model to be able to do multitasking sufficiently, especially if Pro-motion screens are likely to be considered for the ability to down clock the display refresh to save battery rather than go all the way up to 120Hz for anything other than video playback.

For specific tasks the ARM architecture is great - for example video decoding and encoding - which is why something like FCPX or Logic might go well.

Intel might be better for other jobs but we would have to see. Releasing a boutique product would enable Apple to judge the market and give developers a real product to code towards before releasing something more mainstream and usable like a 15" laptop variant or - indeed - a 23" iMac.

I think Apple would be looking for developers to recompile their apps for ARM and deliver solely via a specific App Store at least initially rather than seeking emulation for Intel on the ARM CPUs.

I don't think Apple would mind partnering with AMD to do GPUs for ARM Macs. That is, long as AMD stays on schedule putting out GPUs to Apple's standards/wants/needs for a given product. That seems to be the main reason Apple is ditching Intel for ARM in the first place, because Apple's Mac products in recent years have been hurt by Intel's inability to make any meaningful progress. I don't think Apple is currently interested in developing an in-house GPU. I think they're fine giving up control long as the 3rd party vendor can keep up with what Apple wants. If it was solely a control thing, i think Apple would have started developing an in-house GPU a while ago.

 

[sublunar]

I don't think Apple would mind partnering with AMD to do GPUs for ARM Macs. That is, long as AMD stays on schedule putting out GPUs to Apple's standards/wants/needs for a given product. That seems to be the main reason Apple is ditching Intel for ARM in the first place, because Apple's Mac products in recent years have been hurt by Intel's inability to make any meaningful progress. I don't think Apple is currently interested in developing an in-house GPU. I think they're fine giving up control long as the 3rd party vendor can keep up with what Apple wants. If it was solely a control thing, i think Apple would have started developing an in-house GPU a while ago.

I draw your attention to the newly struck deal with imagination technologies which would allow Apple to explore its own gpu tech. They don’t need to then concern themselves with AMD for graphics in their power sipping products.

 

[DanMan619]

I draw your attention to the newly struck deal with imagination technologies which would allow Apple to explore its own gpu tech. They don’t need to then concern themselves with AMD for graphics in their power sipping products.

"(“Imagination”) announces that it has replaced the multi-year, multi-use license agreement with Apple, first announced on February 6, 2014, with a new multi-year license agreement under which Apple has access to a wider range of Imagination’s intellectual property in exchange for license fees."

I don't know... This seems like Apple has had a relationship with this company for 6 years already and is simply re-upping to get access to more IP and will be paying more in licensing fees. This doesn't seem like any major sign of Apple doing an in-house GPU (in-house meaning there's not a third party company involved for the design or using other IP let alone licensing fees to make a new chip).

 

[vyruzreaper]

Random question - would you have preferred the recent iPadOS on iPad -or- an Ipad that when docked on the magic keyboard turns into MacOS with the same data/docs/user info. MacOS running ARM

 

[UltimateSyn]

Random question - would you have preferred the recent iPadOS on iPad -or- an Ipad that when docked on the magic keyboard turns into MacOS with the same data/docs/user info. MacOS running ARM

This is my dream - the same files with UI wrappers around them that change contextually. I had thought about it previously in the context of an iPhone being docked into a monitor / trackpad / keyboard and displaying iPadOS, but a docked iPad displaying macOS would be just as cool.

 

[AFPBoy]

Random question - would you have preferred the recent iPadOS on iPad -or- an Ipad that when docked on the magic keyboard turns into MacOS with the same data/docs/user info. MacOS running ARM

I think Apple is going to skip the iPad 2-in-1 convertible type of setup, and go straight to using their (highest-end) iPhone as the CPU and trackpad for your computer, and possibly having another chip to drive graphics on your laptop. The current top story on the MacRumors homepage about a port-less iPhone (Leaker: Apple to Stick With Lightning Over USB-C for 'iPhone 12' Before Going Port-Less Next Year).

Maybe this is why Apple ditched 3D touch for haptic touch? To keep it consistent with the current trackpad?

 

[PeterJP]

Random question - would you have preferred the recent iPadOS on iPad -or- an Ipad that when docked on the magic keyboard turns into MacOS with the same data/docs/user info. MacOS running ARM

It sounds nice in theory, but I think it would be confusing in practice. Also, some apps just aren't ready for it. I just decided to start cataloguing all my photos in Photos. Just create a new library, put it on OneDrive and I can access it from my iPad, no? Well, the Mac Photos app uses a character in a filename that is unsupported by OneDrive. But even if it would sync via another cloud system, e.g. iCloud, it still wouldn't work. The iPad Photos app can only work with the system library, not with user created ones.

So when I am working on my photos but I have to leave to catch a train, I undock and I can't continue working on the train.

 

[subjonas]

I agree that in theory it sounds good but in practice not sure. It may be possible that frequently moving between a touch UI and mouse UI on the same device may mess with the brain a bit.
And someone correct me if I’m wrong, but also I believe the difference between macOS and ipadOS is more than just surface level UI; they differ at a deeper system level. Particularly, ipadOS puts priority on battery efficiency and sacrifices things like background processes, while macOS sacrifices battery efficiency for better multitasking. I have doubts that one set of hardware can be optimized for both sets of priorities.

Random question - would you have preferred the recent iPadOS on iPad -or- an Ipad that when docked on the magic keyboard turns into MacOS with the same data/docs/user info. MacOS running ARM

 

[DanMan619]

I agree that in theory it sounds good but in practice not sure. It may be possible that frequently moving between a touch UI and mouse UI on the same device may mess with the brain a bit.
And someone correct me if I’m wrong, but also I believe the difference between macOS and ipadOS is more than just surface level UI; they differ at a deeper system level. Particularly, ipadOS puts priority on battery efficiency and sacrifices things like background processes, while macOS sacrifices battery efficiency for better multitasking. I have doubts that one set of hardware can be optimized for both sets of priorities.

I think you're right. Those under the hood differences wouldn't work on iPad. Otherwise iPad already would run MacOS i'd think. To do this, i think ultimately this theoretical iPad Pro would pretty much still be running iPadOS, and the "MacOS" switch that happens when you dock it to the magic keyboard would be purely UI/visual. Which i think they could realistically pull off if they wanted to. iPadOS already steals a lot of the core "desktop-ish" things from MacOS like a file system and what not. So they'd just have to implement the smart UI switching to make it look like MacOS when docked, but under the hood it'd still be functioning like iPadOS. That, or they could put out a new magic keyboard with a T3 chip or whatever in it that beefs up performance to run in the multitasking over battery efficiency way you describe when docked?

 

[subjonas]

I think you're right. Those under the hood differences wouldn't work on iPad. Otherwise iPad already would run MacOS i'd think. To do this, i think ultimately this theoretical iPad Pro would pretty much still be running iPadOS, and the "MacOS" switch that happens when you dock it to the magic keyboard would be purely UI/visual. Which i think they could realistically pull off if they wanted to. iPadOS already steals a lot of the core "desktop-ish" things from MacOS like a file system and what not. So they'd just have to implement the smart UI switching to make it look like MacOS when docked, but under the hood it'd still be functioning like iPadOS. That, or they could put out a new magic keyboard with a T3 chip or whatever in it that beefs up performance to run in the multitasking over battery efficiency way you describe when docked?

Those seem theoretically possible, but both ideas also seem like a lot of work for Apple and developers. And the latter has some trade offs like a heavier and more expensive keyboard accessory. So it’s just a matter of would Apple consider it worth it, which I’m not sure.

 

[awesomedeluxe]

For a fusion device, you really have to determine if you want the tablet part to run an updated macOS with touch controls or to just switch between macOS and iPadOS.

The former is in all ways neater than the latter. But in addition to making substantial changes to macOS, you are also cannibalizing the iPad Pro and its ecosystem by providing a more powerful alternative that extends beyond the walled garden.

The latter requires a lot of redundant hardware, driving up the price. At a minimum, both the tablet and macOS need their own storage. You are also going to be limited to a handful of marquee apps that truly work seamlessly between the platforms and a wide ocean of ones that do not.

I'm increasingly of the mind that the second approach is untenable. It's a $5,000 device which will constantly break the user experience. As for the first option, my opinion is more mixed. It would still be enormously profitable at $4k, and would be an incredibly attractive product for consumers. It would also help Apple rapidly rebuild their ecosystem after the move to ARM. also also, it would be a great way to justify dropping that bloody touchbar without having to concede what a dumb idea it was. This might make the downsides worth it.

 

[sublunar]

"(“Imagination”) announces that it has replaced the multi-year, multi-use license agreement with Apple, first announced on February 6, 2014, with a new multi-year license agreement under which Apple has access to a wider range of Imagination’s intellectual property in exchange for license fees."

I don't know... This seems like Apple has had a relationship with this company for 6 years already and is simply re-upping to get access to more IP and will be paying more in licensing fees. This doesn't seem like any major sign of Apple doing an in-house GPU (in-house meaning there's not a third party company involved for the design or using other IP let alone licensing fees to make a new chip).

There was a period where it looked like Imagination technologies was going to be unceremoniously dumped. Something made Apple do a u-turn even after they hired a bunch of staff from the company. Maybe it was in the list of patents that the company owns.

 

[simonmet]

Apple has been moving incredibly slowly on the Mac for years. It’s why we see things like 4-5 years between meaningful Mac Mini updates; longer for the Mac Pro. Outdated iMac designs, etc. The laptops fair better but still progress is slow. Intel is partly to blame but by no means the main culprit.

The ARM switch is a tricky one. They want to keep a unique market for the iPads, and they can’t make something like the Air more powerful than a Pro. Many Pros still rely on Boot Camp, so the 16” would be the last holdout, but also when Apple changes architecture they have a tendency to transition entirely and quite quickly.

Basically, I don’t see radical redesigns to the physical design of the laptops as much as I think it’s overdue for a refresh. I’m not sure what they’ll do, but the screen is the thing that needs the most improvement.

 

[awesomedeluxe]

The ARM switch is a tricky one. They want to keep a unique market for the iPads

This is a good point, but the counterpoint is that they could probably sell a Surface-Book style 13" for $3,000 and have a huge margin. It may kill the iPad, but if it sells more than the iPad and makes more per unit sold, this may not matter.

and they can’t make something like the Air more powerful than a Pro.

Literally true. I tried to scale up the A14X to the thermal capacity of the current 16" and found it fell way short. What I think Apple would have to do instead is make the Pro more like the Air. The A14 chipset will never provide the kind of performance you would get from a 45W x86 CPU, but I think it could provide 75% of that performance for 50% of the power.

Many Pros still rely on Boot Camp, so the 16” would be the last holdout, but also when Apple changes architecture they have a tendency to transition entirely and quite quickly.

This is a casualty I can see Apple being willing to accept. They are in a much stronger position now.

Basically, I don’t see radical redesigns to the physical design of the laptops as much as I think it’s overdue for a refresh. I’m not sure what they’ll do, but the screen is the thing that needs the most improvement.

2021 could go either way. If Apple transitions to ARM a redesign is necessary. But even if they do not we are seeing big changes from AMD and Intel. If they stay on x86, their processors will be on TSMC 5 or Intel 7 by 2022 at the latest, and I think that would demand a redesign as well.

 

[sublunar]

Apple has been moving incredibly slowly on the Mac for years. It’s why we see things like 4-5 years between meaningful Mac Mini updates; longer for the Mac Pro. Outdated iMac designs, etc. The laptops fair better but still progress is slow. Intel is partly to blame but by no means the main culprit.

The ARM switch is a tricky one. They want to keep a unique market for the iPads, and they can’t make something like the Air more powerful than a Pro. Many Pros still rely on Boot Camp, so the 16” would be the last holdout, but also when Apple changes architecture they have a tendency to transition entirely and quite quickly.

Basically, I don’t see radical redesigns to the physical design of the laptops as much as I think it’s overdue for a refresh. I’m not sure what they’ll do, but the screen is the thing that needs the most improvement.

Making the iPad Air/Mini not as powerful as the (current 2020) Pro should be a simple case of retaining the 3Gb RAM/A12 SoC while putting it into the larger form factors that have been rumoured. Apple already did this with the iPad - moving it from the 6th gen 9.7" model to 10.2" 7th gen seemingly without altering the SoC.

They could in theory put these iPads onto the iPhone SE 2020 SoC (A13 with 3Gb RAM). Understandably this could lead to a brief period where in certain circumstances the iPad Air and Mini could outperform the iPad Pro. That choice will be up to them though.

Obviously the iPad Pro in 2021 is expected to move onto the A14x next year so any overlap won't be for long.

With respect to Apple tackling the issue of the ARM switch, I don't think it's a simple case of upping the TDP wattage to the equivalent of Intel's to see what sort of performance is feasible.

Apple will be far more likely to go after a 20 hour battery life and seek benchmarks that will aid certain work flows. This is where you might expect video exporting to be an eye opener if you look at various reviews mentioning LumaFusion.

The worry for me would be in connectivity options. Selling a 12" iBook (Macbook replacement) is fine because that model only ever came with a single USB-C port - thus equivalent to the MacBook that came before it.

If Apple added a MagSafe smart connecter as rumoured for a future iPhone or iPad - for charging only it could be argued that such a combination could replace the MacBook Air 13" too which runs with a 10w Intel CPU.

I think GPU performance from an A14x should allow Apple to consider replacing all the 13" MacBook Pros but there's nothing quite like Thunderbolt publicly available for ARM systems - unless Apple have been forging ahead with a motherboard that had Titan Ridge controllers on it.

If Apple can somehow produce a laptop motherboard with Titan Ridge controllers on it I don't see any reason why the 16" MacBook Pro could not be challenged even if it was initially with a large screened iBook 'Air' model.

The challenge arrives with the desktop models and the expectations on hand for those.

 

[littlepud]

If Apple can somehow produce a laptop motherboard with Titan Ridge controllers on it I don't see any reason why the 16" MacBook Pro could not be challenged even if it was initially with a large screened iBook 'Air' model.

That shouldn’t be difficult. Titan Ridge probably only needs a PCIE interface to the CPU. It’s only a question of device drivers.

That being said, I don’t think Titan Ridge would even be necessary since TB3 is fully integrated into the USB4 spec. Apple could very well just implement USB4 directly into their SOC.

 

[high heaven]

I'm not sure if third party developers are willing to participate in the ARM based Mac system. Even now, Mac itself is not a popular platform and not many people are using it. iOS apps? I don't need them. If I cant even use x86 based apps, then it's totally useless. Emulating is not a great option too.

 

[Unregistered 4U]

I don't think Apple is currently interested in developing an in-house GPU.

They already are. And it outperforms other “embedded” GPU’s already. The big difference between their solution and Nvida/AMD is that those come is separate chip packages while Apple’s GPU is built into the A series chip. Will they separate their graphic chip into a separate package in a laptop/desktop A series chip? I don’t know... having the GPU that close to the processor is generally good and if they hit the performance they’re looking for (Fortnite at 120?), then it doesn’t really matter where it is as long as it does what it’s supposed to.

Maybe it was in the list of patents that the company owns.

It was rumored that Apple couldn’t get the deal they wanted from the company, but new leadership came in and suddenly, they’re more willing to cut a deal with Apple.

and they can’t make something like the Air more powerful than a Pro.

They absolutely could, though. If they came out with an Air tomorrow that could output to FCP twice as fast as a Pro (without the Afterburner card), the Pro still has MANY advantages over the Air that would make performance only part of the picture. There’s a lot you give up with an Air (Intel compatibility, expansion, upgradability).

I'm not sure if third party developers are willing to participate in the ARM based Mac system.

They may not be, but it’s wholly their decision to lose out on sales to new Mac users. As far as customers go, they will find either find other apps to do what they want on their new ARM Mac OR will just not buy a new system.

 

[DanMan619]

I'm not sure if third party developers are willing to participate in the ARM based Mac system. Even now, Mac itself is not a popular platform and not many people are using it. iOS apps? I don't need them. If I cant even use x86 based apps, then it's totally useless. Emulating is not a great option too.

I'm not so sure they'd be that quick to jump out. Overall personal computer market share, Apple isn't that huge, but they are still a big enough chunk that it's worth developing for. Apple's user base isn't small, it's just the other PC brands just have that much more. If they were really that small, third party developers already wouldn't bother developing software for Mac. They still have to maintain x86 Mac versions of their software and if that wasn't financially viable, they wouldn't be doing it. The added revenue from that user base is clearly worth the while. I think it'll really come down to how hard porting the softwares is. If Apple has some way to make that process easier and thus less expensive/time consuming for the developers, i don't think it'll be much of an issue for developers to want to participate in the ARM Mac system. Plus, if these ARM machines end up having really good performance, that might spur some growth in Apple's market share if it's on par or somehow better than what AMD/Intel is offering (that's a big if though and may take a few iterations to get there).

 

[awesomedeluxe]

I didn't realize AMD's roadmap had leaked last week.

Two takeaways:

• Apple could beat both Intel and AMD to DDR5. Even the rumored Tiger Lake H does not use DDR5, and we now know AMD will not adopt the faster memory until 2022. This gives Apple an opening next year. Apple has used desktop-class RAM in its notebooks before, and it will certainly have the thermal room to do so after moving to ARM.
• Apple's internal graphics will stack up - at least on the 14". AMD will be sticking with Vega for its APUs again next year, which means the iGPU to beat will be the 96EU Tiger Lake. I think Apple is up to the task.

I am still worried about the 16", but there is definitely an opportunity for Apple to get ahead of its competitors in the 14" space and below.

 

[high heaven]

I'm not so sure they'd be that quick to jump out. Overall personal computer market share, Apple isn't that huge, but they are still a big enough chunk that it's worth developing for. Apple's user base isn't small, it's just the other PC brands just have that much more. If they were really that small, third party developers already wouldn't bother developing software for Mac. They still have to maintain x86 Mac versions of their software and if that wasn't financially viable, they wouldn't be doing it. The added revenue from that user base is clearly worth the while. I think it'll really come down to how hard porting the softwares is. If Apple has some way to make that process easier and thus less expensive/time consuming for the developers, i don't think it'll be much of an issue for developers to want to participate in the ARM Mac system. Plus, if these ARM machines end up having really good performance, that might spur some growth in Apple's market share if it's on par or somehow better than what AMD/Intel is offering (that's a big if though and may take a few iterations to get there).

Mac still exists because of its OS, software, and system. Final Cut Pro X is a great example. Also, in order to develop iOS apps, you still need Mac. However, it would be very difficult to attract developers to ARM-based Mac if it's not that worth it.

 

[DanMan619]

Mac still exists because of its OS, software, and system. Final Cut Pro X is a great example. Also, in order to develop iOS apps, you still need Mac. However, it would be very difficult to attract developers to ARM-based Mac if it's not that worth it.

They also still exist because of the ecosystem effect as a whole, that's their greatest strength. I really don't think they'll have that hard a time attracting developers considering how ingrained in the ecosystem people are. And with Apple using things like Catalyst/Swift to make iOS apps be able to run native on Mac and iPadOS and vice versa or have full versions of apps be able to run on iPadOS (like Photoshop recently, meaning Adobe already has a small start on porting to ARM).

If Apple is doing what i think they're doing and trying to have MacOS/iOS/iPadOS apps be effectively hardware agnostic (or at least very easily be made natively for all Apple hardware via Catalyst) once all their devices are on ARM. That greatly expands the user base while also cutting down on versions of apps (or effort to make those versions) developers need to make. I think that would be enticing to them. If my wild theory is right, it seems like a smart long game to me, but people will struggle in the transition short term as they did in the PowerPC to Intel transition.

Also, Microsoft did the SQ1 ARM chip for the Surface Pro X, so there's some familiarity there as well. Third party developers didn't take to porting for that because it was one individual product. If it's a whole ecosystem of ARM products, then i think they won't be as hesitant.

 

[firehawk12]

If ARM MBPs are in the near future, is it still worth considering an Intel MBP 13" 2020? Would that be like buying a PowerPC Mac right before that transition?