Pegatron Tipped to Manufacture Upcoming 'ARM-Based MacBook'

[barmann]

Yeah, and it's nothing to do with the age of the products....

The entire mac lineup right now is depressing...
[doublepost=1527591206][/doublepost]


Moving to ARM will give apple far more freedom to put out the hardware THEY want to make rather than being tied to what intel produce for the mass market.

This isn't aimed at people who want to run windows virtual machines. This is aimed at the 99% of users who want to run mac apps, and maybe some 365 office apps which you can even run in a browser these days.

Priced appropriately with decent screen, trackpad and keyboard it will sell like hot cakes.

- Apple already make the hardware THEY want to make - that's the problem .

- Mac apps are not a thing .

- There is not a 99% of users for anything , apart from the 99% of users hitting a cieling hard if the hardware doesn't allow for their growing demands .

- It's virtually impossible to sell a laptop that doesn't offer the same support for 3rd party apps as the rest of the Mac line . The very thought is ridiculous .

 

[Smeaton1724]

You're not looking deep enough.

Apple have not been getting what they want from intel for a long time now.

This is a way to float the idea and get the processors into their notebooks. But mark my words, this isn't the end of it.

Intel has been getting sub 10% performance improvements for almost a decade now. Apple has been getting 1.5x or more per year. Apple's AxX CPUs are already set to overtake intel on the low end this year - it won't be long before they're poised to put higher end variants in things like the Mac Pro, but they need the software support to be there.

Which is why they're going to stick a CPU of that architecture in a volume seller to start with.

The ARM chip in the iPad Pro 10.5 and 12.9 overtook the Macbook 12'' and the base model Macbook Pro 13'' last year! Go look at the geekbench scores for Single and Multi core.

 

[asiga]

It's certainly a bad moment for purchasing new Macs: CPU vendors will be offering redesigned CPUs in the next years with hardware fixes for Meltdown/Spectre, and you'll want a CPU with hardware fixes when they are released. If you add to that the uncertainty regarding the butterfly keyboard, as well as a slower than expected migration to USB-C (plenty of older USB ports still in brand-new accessories in the market) which makes you guess you'll depend on dongles for quite a few years to come... everything suggests that whatever Mac you purchase today, you might regret the purchase a few months from now when all these uncertainties are conveniently addressed.

 

[throAU]

- Apple already make the hardware THEY want to make - that's the problem .

- Mac apps are not a thing .

- There is not a 99% of users for anything , apart from the 99% of users hitting a cieling hard if the hardware doesn't allow for their growing demands .

- It's virtually impossible to sell a laptop that doesn't offer the same support for 3rd party apps as the rest of the Mac line . The very thought is ridiculous .

So, do you remember the intel transition?

Were you perhaps one of the PPC guys who always claimed PPC was where it was at, and that Apple would never use a dirty intel processor?

Apple has done this before multiple times. 68k to PPC. PPC to intel. This isn't their first rodeo - when the performance and supply chain side makes sense, they jump.

Given their improvements and control of their own destiny with AX processors, the time is ripe.


Bookmark this thread and come back to it in 18 months....


edit:
this thing will run native mac software (yes, including third party applications), guaranteed. no different to the way things were using rosetta back for the PPC transition.

 

[bluecoast]

I think this makes lots of sense and defer to the great comments above as to why.

As to how this could happen:

• In March 2019, the new (ARM) MacBook (not pro) is launched
• It’s surprising fast and uses an Apple GPU - and Face ID & two usb 3.0 ports & a lightening port.
• Memory and storage are soldered into the motherboard
• Like an iPad it comes in WiFi only or WiFi & 4G
• Its keyboard is more like the iPad Pro keyboard (and can be relatively easily removed & replaced by a service professional) as the body of the computer is built more like an iPad ie it ‘splits’ open.
• Like the latest iPhones, it’s waterproof
• It runs macOS. Xcode enables compilation to MacOS ARM.
• This version of macOS is forked from Intel macOS until the next major release version. Why?:
• Apple reveals that it’s been working with vendors such as Microsoft to get their apps onto ARM, using a pre-release version of the Marzipan APIs.
• Twitter also revea their iOS app has been ported to the Mac using this method.
• It also reveals details of the developer version of Marzipan which will more easily enable iOS devs to get their iOS apps onto the Mac - with more details coming to WWDC
• This ARM fork of macOS also has the revamped App Store.
• It’s revealed also a special point release of macOS featuring marzipan will be released about 6 weeks after WWDC (ARM only)
• In fall (October/November) 19, the next major point release of macOS is released featuring the App Store and marzipan apps

 

[throAU]

The ARM chip in the iPad Pro 10.5 and 12.9 overtook the Macbook 12'' and the base model Macbook Pro 13'' last year! Go look at the geekbench scores for Single and Multi core.

Yup. Though that was perhaps inflated somewhat by some of the benchmarks in geekbench being "Friendly" to the A10X's strengths.

My "set to overtake" was a conservative statement. I own a 10.5 and that CPU is extremely fast for the power it draws. It could drive a low end macbook already quite happily. Give it more headroom via cooling and additional power (and more cores via bigger budget in terms of die size as well) and it would smoke anything intel have inside of 28 watts. At half the power draw.

 

[cicalinarrot]

This is big.
This could mean Apple has some plans about their computers and, at this point, any direction is better than not moving.

 

[BBCWatcher]

If Apple were able to obtain a single chip that incorporates a couple competent X86-64 cores plus a more than competent ARM section, that'd be a great foundation for a MacPad. It could happen.

 

[762999]

You're not looking deep enough.

Intel has been getting sub 10% performance improvements for almost a decade now. Apple has been getting 1.5x or more per year. Apple's AxX CPUs are already set to overtake intel on the low end this year - it won't be long before they're poised to put higher end variants in things like the Mac Pro, but they need the software support to be there.

you're assuming too much! the cpu is not important here. Apple already have a fast arm CPU, do they use it in all their devices? no! they don't mind selling you the old **** for years. So right now, they will release some computers with it, but even if new cpus come out, they will milk the cow by using the same cpu for years.

 

[bigtomato]

I don't think this is a good idea since we would be dependant on apple only for everything. Innovation happens everywhere and only by utilizing other technologies can a product be better. This will turn into more of a monopoly and you will get underpowered computers because of apples desire to make more profits. We're even limited to AMD graphics right now when there are titan v's available from Nvidia.

 

[Dave245]

Ok i'm a little confused, will this be a Surface type of device? haven't Appel said they don't like the idea of such devices.

 

[ColdShadow]

I have a speced out 2016 12" MacBook (Fanless) and the performance of Intel M processor is just terrible.
I really hope Apple goes full ARM for 12" MacBooks.the Apple processors in iPads have much better performance and speed than the awful, slow and crippled M series of lazy ass intel.
can't wait.

 

[Stingray454]

I just have a very hard time believing that they would go full ARM. In case you didn't notice, Windows already tried x86 emulation on ARM and the performance is just rubbish. Sure Apple could probably spit out some better ARM chips (last test I saw was on a Snapdragon 835 I think), but if they don't add specific emulation hardware I have a very hard time thinking it will work out.

Sure, it could be a last resort for vendors that doesn't port their software quick enough, but completely removing the possibility to use both hackintoshes or to dual-boot your mac will not help Apples case. I personally dual boot, and that's probably the only reason I still have a macbook pro.

I don't think it's ready yet, maybe by 2019-2020. IF something like this is happening, don't expect Apple to mention it this year.

 

[sblemmy]

If we go back to Steve Jobs' analogy that tablets are cars and PCs are trucks, an ARM-only Mac doesn't seem to make sense. There would be too many compromises in creating what is essentially a crossover device.

 

[throAU]

I think this makes lots of sense and defer to the great comments above as to why.

As to how this could happen:

• In March 2019, the new (ARM) MacBook (not pro) is launched
• It’s surprising fast and uses an Apple GPU - and Face ID & two usb 3.0 ports & a lightening port.
• Memory and storage are soldered into the motherboard
• Like an iPad it comes in WiFi only or WiFi & 4G
• Its keyboard is more like the iPad Pro keyboard (and can be relatively easily removed & replaced by a service professional) as the body of the computer is built more like an iPad ie it ‘splits’ open.
• Like the latest iPhones, it’s waterproof
• It runs macOS. Xcode enables compilation to MacOS ARM.
• This version of macOS is forked from Intel macOS until the next major release version. Why?:
• Apple reveals that it’s been working with vendors such as Microsoft to get their apps onto ARM, using a pre-release version of the Marzipan APIs.
• Twitter also revea their iOS app has been ported to the Mac using this method.
• It also reveals details of the developer version of Marzipan which will more easily enable iOS devs to get their iOS apps onto the Mac - with more details coming to WWDC
• This ARM fork of macOS also has the revamped App Store.
• It’s revealed also a special point release of macOS featuring marzipan will be released about 6 weeks after WWDC (ARM only)
• In fall (October/November) 19, the next major point release of macOS is released featuring the App Store and marzipan apps

I'd agree with most of that.

The thing is that some people are missing though is that there is no need for apple to get software developers onboard yet. For some time applications uploaded to the app store have been in an intermediate byte-code format that apple can recompile themselves (without your source) to be native code for future platforms. This is one of the features of LLVM.

Additionally, most of the heavy work in a typical mac or IOS application (audio/video processing, etc.) is handled by the Apple frameworks. Which Apple can recompile as native (in the supplied/installed version of macOS) and your old application calling them will get native speed for much/most of the processing heavy things it is doing.

For the stuff that isn't being done by the frameworks - apple will rely on a new version of rosetta.


All this has been done before with the intel transition and it worked just fine.

The original intel machines had processors much stronger than the PPCs they replaced, but the software ran "about the same" via translation.

As the old machines were phased out and application vendors recompiled their code performance improved.

It will be even easier for apple this time around due to my notes regarding LLVM and the app store above.

App store apps will get recompiled by apple as native. Non-app store apps (if supported, I suspect so as there's a lot of them) will be no worse off than PPC native applications were running on intel in 2004 (or was it 2005? 2006? memory hazy).
[doublepost=1527594607][/doublepost]

I just have a very hard time believing that they would go full ARM. In case you didn't notice, Windows already tried x86 emulation on ARM and the performance is just rubbish. .

Microsoft are muppets when it comes to this. Don't think that just because Microsoft fail at something, it can't be done.

The windows software landscape is not the same as the apple software landscape. Apple's frameworks and the applications that use them are a lot easier for apple to port. Apple binaries for example have been capable of being multi-architecture since back in the NEXTSTEP days.

Microsoft? Nope...

Microsoft's win32 platform is a dogs breakfast, the applications that run on windows generally reinvent libraries or include rafts of third party code microsoft do not control to get things done. The frameworks apple provide are far cleaner and more likely to be used by a developer.

 

[Neodym]

Perhaps that rumored "MacBook" is more a docking station on steroids, coming with display and keyboard, but needing an iPhone/iPad connecting to it.

Connection between that dock and an iOS device could be done (optionally) wirelessly, so they wouldn't have to adjust a dock connector to different device sizes now and in the future (remember that old patent with an iDevice having to be inserted in some sort of iMacesque docking station?). Plus they could reduce cable clutter / keep the Lightning port free for charging.

No more syncing your documents between Mac & iDevice, it's all on one device.

That would also put a different light on the rumor of soon being able to run iOS & Mac apps in parallel on one device.

Edith reminded me of the odd triple-point connector on the iPad Pro's. So far it seems that's only for a keyboard. But perhaps it was designed with a full-fledged dock in mind? iPad acting as screen & computer at the same time, the keyboard offers additional ports (USB, Monitor etc.) not present on "only-keyboards". Sure - nothing new there, but Apple is rarely known for completely new stuff, but rather for implementing existing approaches in a better and more seamless way.

 

[Lesser Evets]

Apple will definitely transfer their desktops to A-chip architecture sooner-or-later.

Sooner is very-very unlikely.

They might drop an "experimental" laptop on the consumers in the near future to begin the bridging. Perhaps that is the MacBook Air 2018/19, but the switch over for most or all Macs won't be for another 2+ years. The way Apple moves these days, it probably won't be until 2021 at the earliest, though it could be done anytime with a custom CPU and proper software.

Intel is now extremely unreliable. Their chips are inefficient for Apple's needs. Intel will only become less efficient for Apple in the next years unless they innovate to cater to Apple... and then their architecture is available to the competitors! That is a lousy business model for a company that wishes to consume their industry and increase growth and revenue with innovative product. By producing their own chip designs, Apple will keep most of the competition at bay with a year or two lead in release dates.

Low-end notebooks and iPads should begin to mesh ASAP at Apple. Then the ARM designs should spread upward into desktops over the next decade.

 

[throAU]

you're assuming too much! the cpu is not important here. Apple already have a fast arm CPU, do they use it in all their devices? no! they don't mind selling you the old **** for years. So right now, they will release some computers with it, but even if new cpus come out, they will milk the cow by using the same cpu for years.

Not yet. But when they do they'll be able to properly re-unify iOS and macOS.

No, not in terms of UI, but in terms of having common back-end frameworks to get the actual work done inside the applications. Right now they're still similar but iOS has had a lot of the cruft in macOS removed and re-implemented more securely.

 

[pubwvj]

Apple is really missing the boat with their mantra of not crossing iOS and MacOS and not having touch screens on MacOS. It is very possible to do both on the same hardware with all the best of each. The Google PixelBook looks most excellent and combines a touch screen tablet in a keyboarded thin notebook, it's only problem being it doesn't run MacOS and iOS software.

 

[Marx55]

Did Apple learn from the PowerPC fiasco? We need x86 compatibility with the 97% of the world, and that means Intel chips inside Macs! Otherwise, it is a deal breaker and switch to Windows. A shame for all.

 

[throAU]

If Apple were able to obtain a single chip that incorporates a couple competent X86-64 cores plus a more than competent ARM section, that'd be a great foundation for a MacPad. It could happen.

The problem with that idea is that it isn't a clean break.

Apple try to do clean breaks. Providing x64 in hardware is dead hardware once the transition is made. It is wasted die space, wasted R&D to do the design, etc. They just won't do it. It would blow the power consumption, cost advantages, performance per watt advantages out to make the shift pointless. And it would just increase the transition time due to developers being lazy and not porting their code...

No... since woz, Apple have been about doing everything they can in software and being hardware independent as much as possible. hence things like openCL (instead of CUDA) being pushed, rosetta for the previous transition, etc.

There will be a performance penalty for not including x64 hardware, sure. But it will go away as more of the OS and third party applications are ported.
[doublepost=1527595561][/doublepost]

Did Apple learn from the PowerPC fiasco? We need x86 compatibility with the 97% of the world, and that means Intel chips inside Macs! Otherwise, it is a deal breaker and switch to Windows. A shame for all.

This is less important now. Applications are shifting to the cloud / mobile / java / javascript / etc.

Running x86/x64 code is less relevant now than ever.

 

[asiga]

• This ARM fork of macOS... also has the revamped App Store.

No need to write an ARM fork. MacOS inherits multi-architecture support from the days of NeXTSTEP. The kernel, the dynamic loader, the executable format, the app-bundles... everything has a multi-architecture design. Apple didn't do an "Intel fork" when they moved to Intel: they just had Intel support in the system. Years later, they gradually removed PowerPC support from the kernel, dynamic loader, etc... Today they don't even need to add ARM support: it's already implemented (they implemented it for iOS, which shares most of that low-level code with MacOS).

 

[Neodym]

Did Apple learn from the PowerPC fiasco? We need x86 compatibility with the 97% of the world, and that means Intel chips inside Macs! Otherwise, it is a deal breaker and switch to Windows. A shame for all.

For the majority of standard applications, virtualization is more than sufficient on current hardware. The small remainder would need more powerful hardware than offered by Apple anyway. This is not 2006 anymore, when you absolutely needed Intel hardware to run x86 software properly.

 

[762999]

Not yet. But when they do they'll be able to properly re-unify iOS and macOS.

No, not in terms of UI, but in terms of having common back-end frameworks to get the actual work done inside the applications. Right now they're still similar but iOS has had a lot of the cruft in macOS removed and re-implemented more securely.

your point is totally unrelated to mine.

My point IS:

even if they fully switch to ARM cpus and completely free themselves of INTEL cpus, they will still won't update their computer line each year.

 

[throAU]

your point is totally unrelated to mine.

My point IS:

even if they fully switch to ARM cpus and completely free themselves of INTEL cpus, they will still won't update their computer line each year.

Your thoughts on this are based on what exactly?

They update the ipad and iphone every year.

Being in control of their own CPUs will give them the ability to put out new CPUs for their computers just as frequently. They haven't updated the mac lineup in ages because, to put it bluntly, intel haven't built anything they want.

A 5% performance jump (which is all intel have been putting out per generation since sandy bridge) is not worth deprecating the existing mac lineup and putting out new models. PC vendors can do what they like - they do this because they don't have a captive market. This is why they make 5-10% margins and include heaps of third party crapware with their machines to try and re-coup some money.

Apple isn't interested in 5% margins, and their customers aren't interested in crapware pre-installed on their machines by apple in order to try and make up some of the lost margin.