Apple ARM Based MacBooks and iMacs to come in 2021

erek

[H]F Junkie
Joined
Dec 19, 2005
Messages
10,785
Low-end, right? Entry Level offerings?

"According to analyst Ming-Chi Kuo, Apple has started an "aggressive processor replacement strategy", which should give some results by the end of 2020, around Q4, or the beginning of 2021 when the first quarter arrives. According to Kuo, the approach of doing in-house design will result in not only tighter control of the system, but rather a financial benefit, as the custom processor will be 40% to 60% cheaper compared to current Intel CPU prices."

https://www.techpowerup.com/265165/apple-arm-based-macbooks-and-imacs-to-come-in-2021
 
Another massive blow for Intel, and another contract soon to be severed.
It will be interesting to see the real-world benchmarks of Apple's ARM CPUs compared to x86-64 CPUs on the market.
 
I won't pretend to know when these are coming, but they'll be interesting to see for sure.

I'm kind of excited(?) to see some real world testing in the desktop space.
 
  • Like
Reactions: erek
like this
I'm kind of excited(?) to see some real world testing in the desktop space.
Considering the hardware that passes as usable for desktops and that heavy work can be farmed off to specialized hardware, and ARM on the desktop seems like an eventuality.

It'll just take someone like Apple with their vertical integration to do it well first.
 
Why?

Given that iPads now support KB/Mice/Touch/stylus, and runs on ARM, it seems kind of pointless to switch KB/Mouse only Macs to ARM.
 
After seeing what happened to Windows RT, would Apple really want to go down that same road. ;)
 
Considering the hardware that passes as usable for desktops and that heavy work can be farmed off to specialized hardware, and ARM on the desktop seems like an eventuality.

It'll just take someone like Apple with their vertical integration to do it well first.
We're at the point where Apple is using ASICs and FPGAs (not just the Afterburner card, Apple have been using FPGAs in their laptops since 2011 at least - though to be fair, I'd guess a lot of other vendors were/are doing the same for glue logic or I/O handling purposes) to offload work onto. The T2 chip is even handling video decode, instead of the Intel SoC.

IMO, both Apple's and Microsoft's approaches are very interesting, in terms of approaching ARM with respect to professional workloads.

EDIT: out of the embedded markets, Nvidia is also moving their software development stack to ARM. Can already develop (and compile) CUDA code natively on my Jetson Nano, though I suspect Nvidia is aiming more for the Thunder X3/X4 and Ampere Altra crowd.

EDIT2: Though actually having a Linux-based ARM development station (in the form of the Jetson Nano) is nice. Wish it was in a laptop form factor.
 
"According to analyst"

Might as well consult a Magic 8-Ball while you're at it.
You don't read Apple news. Ming-Chi Kuo has a long history and track record. He has plenty of people in the supply chains of Apple that give him information at all levels. He's probably the only reliable source of information on what Apple is doing.
 
Apples silicon will be a game changer. i7-i9 performance with passive cooling and 1/4 the power consumption!

After seeing what happened to Windows RT, would Apple really want to go down that same road. ;)

Apple aren’t idiots like Microsoft, this has been in the works for years and most of the apps will be ready to go already for the platform.
 
Apples silicon will be a game changer. i7-i9 performance with passive cooling and 1/4 the power consumption!



Apple aren’t idiots like Microsoft, this has been in the works for years and most of the apps will be ready to go already for the platform.

You must be too you to Remeber the G5 Apple times...yes, Crapple is that stupid.
 
You must be too you to Remeber the G5 Apple times...yes, Crapple is that stupid.
Yup. One of the highest valuated companies that has market dominance in multiple consumer markets is completely stupid. They can’t do anything right and anything to the contrary is just lies.
 
Seems like it would move them further into the niche category.

Yup. One of the highest valuated companies that has market dominance in multiple consumer markets is completely stupid. They can’t do anything right and anything to the contrary is just lies.

One of their "market dominances" isn't the notebook or desktop computer space...
 
One of their "market dominances" isn't the notebook or desktop computer space...
Getting rid of an expensive (and failing) Intel contract, in place of using in-house CPUs and designs - how is that not a win for Apple?
It will eventually be a win for everyone else, as Apple will lead the way to mainstream (non-mobile) ARM technologies and support, which will thus eventually become a true competitor (and hopefully killer) of x86-64.
 
Getting rid of an expensive (and failing) Intel contract, in place of using in-house CPUs and designs - how is that not a win for Apple?
It will eventually be a win for everyone else, as Apple will lead the way to mainstream (non-mobile) ARM technologies and support, which will thus eventually become a true competitor (and hopefully killer) of x86-64.

It saves them money. It does nothing to assist their market share if anything it will hurt it.
 
Seems like it would move them further into the niche category.

Yep. Think of all the people that need some Windows applications on their Mac, through Parallels, Boot Camp, Wine, etc... That goes away.
 
If Apple ditches Intel, it will need to find a replacement for Thunderbolt 3. Every MacBook in Apple's current arsenal relies on a Thunderbolt connection, but the Intel-developed hardware interface is available only on laptops with Intel CPUs.

Apple will reportedly sidestep this problem by using USB4, a new interface that merges the Thunderbolt and USB protocols. USB4 will enable 40 Gbps transfer speeds and support up to two 4K displays or one 5K display

https://www.laptopmag.com/amp/news/macbooks-with-custom-apple-cpus-usb4-arriving-next-year-report

That leaves the question of GPU IP licensing. Considering the R&D & licensing costs, won't it be cheaper for apple to go for AMD's upcoming APUs such as Van Gogh with RDNA iGPUs !?
 
  • Like
Reactions: dgz
like this
After seeing what happened to Windows RT, would Apple really want to go down that same road. ;)

Microsoft to developers: look at these new apis! They are so much better! You should update your code! But we are leaving all the old ones in place for legacy compatibility so you can still count on everyone having internet explorer 11 and file com dialog boxes from windows 3.1. So there’s no real reason to do it.

devs: pass.

Microsoft to developers after winrt was announced: you need to rewrite everything for the App Store so the handful of people who buy these machines can run them!

devs: pass.

Apple to devs: we’re removing Rosetta support for PowerPC architecture. Update now or no one will use your software anymore.

devs: okay!

(and the ones that didn’t update... no one even remembers their names)
 
Yes because they were completely stupid switching from IBM to Intel.

Tinted glasses much?
Tehy went from mocking Intel...to "Intel is the best".
Crapple and their fanbois (reality distortion field).

Remember Mac servers? :wtf:
 
EDIT2: Though actually having a Linux-based ARM development station (in the form of the Jetson Nano) is nice. Wish it was in a laptop form factor.

Ah, that saved me the effort of asking what you've been using your Jetson Nano for. Figured if you'd prefer it in a laptop form, you're probably not into building miniature nano kill bots with facial recognition. Good to know
 
Tinted glasses much?
Tehy went from mocking Intel...to "Intel is the best".
Crapple and their fanbois (reality distortion field).

Remember Mac servers? :wtf:
Are we talking about the present and future of Apple and what they are going forward with or are we talking about the past? Because one is a huge waste of time.

If you want to do that, we could bring up, hey remember Windows Vista? That thing that is supposed to be directly under Microsoft competence (you know, OS' that one big thing they do) that no one wanted? Or Windows 8? PC users and their fanbois (reality distortion field).
If you want to go toe to to about mistakes, or issues in software we could go all day. In fact if all we want to do is talk about failures that companies have done we can go ahead and post garbage from twenty plus years ago. How any of that is remotely relevant I struggle to see, but for some reason all the conversation on this news post is incredibly bias. Hrmmm, wonder why.
 
Last edited:
It saves them money. It does nothing to assist their market share if anything it will hurt it.
If this were 2010, I would say you are probably right.
Not in 2020, though.

It is also going to serve a greater function by leading the way with ARM CPUs in workstations, servers, and non-mobile and non-embedded platforms.
In other words, it is going to lead the way for other developers to start working on non-x86-64 tech, which has been direly needed for years now.

AMD is the only corporation keeping x86-64 alive and competitive, and eventually, the process nodes won't be able to shrink any further, and then what?
Oh, right, exactly what is happening to Intel with its 14nm+++++ designs - it may eventually be 5nm+++++ or 2nm+++++, but that will be it, end of the road.

ARM has a much more efficient design, and while we have the hardware optimization available, we don't exactly see much of the software optimization side of things outside of the mobile space.
Apple releasing ARM-based desktops and laptops will lead the way for other developers and corporations to start optimizing and capitalizing on these changes.

Sticking with x86-64, and especially a contract with Intel and its failing offerings, would be an incredibly bad business decision.
x86-64 is nearing the end of its proverbial road, and while AMD, and perhaps Intel, might be able to drag things out with it for another decade, the writing on the wall for x86-64 is obvious in that it is a dead-end technology.

Unless of course, you are ok with market and innovation stagnation as what Intel did with x86-64 over the last decade, and certainly within the last five years.
I'm not all-for Apple by any means, as they are out for themselves, but I am all-for the roads of innovation in other technology markets that their decision will inadvertently open.

Long story short, innovation is good.
 
Last edited:
Remember Mac servers? :wtf:
Yep, the main flaw with those servers wasn't that they were inherently bad themselves, but that OS X had a stable but slow job scheduler, and servers which had thousands of smaller threads running struggled to perform on OS X.
A similar PowerPC or x86/x86-64 server with identical hardware running Linux would run circles around the XServe servers because of this, and around 2009, Apple abandoned it.

This was a smart decision, since it was an inherent flaw in their existing OS, which aside from the enterprise market did not need to be fixed, and their newly emerging smartphone and tablet market was starting to take off.
Money pit vs profits - I think Apple made the right choice.
 
I guess there goes the Hackintosh community. I do like macOS but I don't like Apples hardware old tech at top dollar prices.
 
I guess there goes the Hackintosh community

Nah they'll start hacking cheaper ARM devices and move on.

I'm betting this will only be lower-end stuff, like iPads with keyboards built-in.
 
You must be too you to Remeber the G5 Apple times...yes, Crapple is that stupid.
I remember, and worked extensively on their IBM and Motorola PowerPC-based systems.
If you don't remember, the IBM 970MP was a dead-end design, and the overhead was far too expensive for IBM to continue to develop for, especially with the minimal amount of sales the G5 Quads made, with them being a flagship workstation.

Apple moving to Intel's x86, and later x86-64, CPUs at the time was a great decision, especially with the Intel Core CPUs, and later Core 2 CPUS, regaining performance and market share compared to AMD's then-offerings.
So, what exactly is your point about Apple being "stupid", again?
 
iPads already have ARM CPUs in them.

But they don't have keyboards built-in....and iMacs are hardly workstations; I had one as a workstation for a few years and it was a struggle. These will be big iPads on a stand.
 
If Apple ditches Intel, it will need to find a replacement for Thunderbolt 3. Every MacBook in Apple's current arsenal relies on a Thunderbolt connection, but the Intel-developed hardware interface is available only on laptops with Intel CPUs.
Not quite.
https://en.wikipedia.org/wiki/Thunderbolt_(interface)#Royalty_situation
...and now vendors are freely allowed to produce Thunderbolt controller silicon...

It's like nearly everyone in this thread is stuck in 2010.
Technology changes, markets change, get with the times...
 
But they don't have keyboards built-in....and iMacs are hardly workstations; I had one as a workstation for a few years and it was a struggle.
iMacs are classified as workstations, and have Intel x86-64 CPUs while natively running MacOS - they are not a mobile or embedded platform.
They might be lower-end workstations relative to what is on the market now, but regardless of your personal opinion of their performance, they do remain classified as workstations.

Also, we have had the ability to connect Bluetooth keyboards to iPads since their inception over a decade ago.
By this logic, Microsoft Surfaces do not have keyboards built-in - does that make them any less of a mobile workstation?

These will be big iPads on a stand.
smh
 
Not quite.
https://en.wikipedia.org/wiki/Thunderbolt_(interface)#Royalty_situation
...and now vendors are freely allowed to produce Thunderbolt controller silicon...
That and Thunderbolt 3 has been rebranded as USB4... the connector and cable situation is still romeo foxed, but at least the data standard is there. This isn't going to be a sticking point for Apple.

iMacs are classified as workstations
I'll have to say that I'm still quite skeptical here. I get the laptop angle; running MacOS X on ARM with the desktop experience shouldn't be too difficult once Apple has the appropriate CPU parts available, but in general, these are likely to succeed mostly in terms of providing said desktop experience as opposed to a tablet experience -- not as being higher performance computers.

To me, the iMac seems to exist for those instances where either the computer is not expected to move, or where it is expected to provide higher performance than would be available from a Macbook. In the first case the iMac could be cheaper, so I get that angle for switching to ARM, but in the second, I expect ARM to fall far short.

Maybe the first case is enough for Apple to produce them? It would make sense even for businesses if the heavy lifting, whatever that may be, is largely constrained to local or remote cloud resources, with the iMac simply being a responsive 'terminal' so to speak.
 

It will be a tablet mainboard hooked up to a bus for expansion hardware like a hard drive and input devices running a windows-based iOS UI. (Not Microsoft Windows, just windows like every other desktop/laptop OS UI uses.)

What will be interesting, hardware-wise, will be how Apple transitions to running apps on their actual laptop/desktop machines. Because they're going to have to have some shared capabilities. People will be upset if software they use on their iPad-based MacBooks and iMacs doesn't work on the higher-end machines.

I think the plan, based on what Apple's done so far, will be to use ARM co-processors in their x86 machines as a device within a device, and from there move on to merging iOS and MacOS.
 
Back
Top