Kuo: Apple might release two ARM-based MacBooks this year

nanoguy

Posts: 1,355   +27
Staff member
Rumor mill: While it wouldn't be a complete surprise, Apple's first ARM-based Macs would reportedly keep the same industrial design of current Intel-based models. The first machine to receive Apple's all-new computer SoC would be the 13-inch MacBook Pro, followed by the MacBook Air possibly this year. The rest of the Mac lineup would have to wait until 2021 and 2022.

Apple has made it official that it will transition its entire Mac lineup from Intel processors to its own custom silicon based on the ARM64 architecture. The company hinted that the new CPUs will debut in a Mac later this year, which naturally invites to speculation as to what that may be.

According to analyst Ming Chi-Kuo, that device will most likely be a 13-inch MacBook Pro with a very similar design, if not identical, to the 2020 model. That means it's going to feature the Magic Keyboard with scissor switches, but won't have the slimmer bezels of the 16-inch MacBook Pro.

Kuo believes Apple will also launch a new ARM-powered MacBook Air later this year, with the possibility of that release being pushed to early 2021.

A 16-inch MacBook Pro and the long-rumored 14-inch MacBook Pro with Apple silicon wouldn't arrive until the second or third quarter of 2021. These newer models may also come with mini-LED displays.

Notably absent from Kuo's report are iMacs, but since the Cupertino company is planning to complete the CPU architecture transition in two years, they should also land sometime in 2021 or 2022 at the latest.

The new chips could give Apple the ability to release new Macs on the same cadence as new iPhones, as well as reduce their cost and potentially sell them at a lower price (possible, but unlikely). The company shipped around 15 million Macs in 2019, and Kuo believes the new ARM-based models will increase that to anywhere between 18 to 20 million in 2021.

Permalink to story.

 
I look forward to seeing what the thermal difference is on these models versus the standard x86 based models.
 
Last edited:
I look forward to seeing what the thermal difference is on these models versus the standard x86 based models.
It depends over anandtech there is a comparison of the A13 and he stated the thermals and voltage is a bit out of control pulling more watts than the chip should with the thermals throttling performance.
All indications point to Apple pulling out of the power user desktop and laptop market as they don't make a large portion of profit and they haven't been able to gain traction of market share in the marketplace.
I imagine on their end as well, Apples strength has always been ease of use and the optimizations made to the os and applications due to limited hardware configurations. So it will be 10x easier to optimize performance for ARM and streamline any overhead from supporting separate instruction sets. The problem is ARM just doesn't have the scalability to compete with x86 and doesn't have the advantage of supporting CISC and RISC like simple instructions so it will lose certain advantages in regards to productivity and more complex software.
 
I don't think anyone except the "tech-as-fashion" crowd will keep buying Apple's junk computers for much longer. The platform has very few exclusive programs left that make it worth investing in. PCs are cheaper, more powerful, more reliable and have many times more software options. The Apple ecosystem is OK I guess but are people really excited about being able to run iOS apps on their Macbooks?
 
What make Macs popular are ease of use, doesnt get viruses (its a grey area as they can get them n malware but windows does get them more frequently) and longivity. Apple tends to last longer than most devices on the market. Thats not to say that cant or dont have issues, they can n do. Most just dont hear about them.
This is what people hear or get told so they buy a mac and most are happy with the decision. Macs arent for everybody, in some cases Windows will be the only option due to the type of work or software needed. Some stick to Windows as that is what they know. Similar how many stick with Intel despite AMD being just as good and cheaper.

While not everyones like Apple, they arent going anywhere. People still do and will buy them for years to come. You have to make the decision if Apple is for you or not.
 
Tell that to the people who complain about Windows 10 all the time and see where that gets you. You'll be laughed out of the room.
They same could be done to the people laughing as peolle would laugh at them for not knowing how a computer works or simply didnt bother learning Windows 10.

Also dont forget the ppl who complain are mostly in the minority not the majority.
 
They same could be done to the people laughing as peolle would laugh at them for not knowing how a computer works or simply didnt bother learning Windows 10.

Also dont forget the ppl who complain are mostly in the minority not the majority.
Trust me, I know this. I, myself, run Windows 10 with no issues at all. Then again, I'm not tweaking it to hell and back on the advice of some random person on the Internet.

Amazing what happens if you let the OS run as it was meant to run, you have no issues.

I'm just bringing up a point where people have gone away from the PC because they somehow broke Windows 10.
 
Apple has an opportunity here to leave Intel and AMD in the dust, at performance level, with the clear advantage of an easier optimization having full control of hardware and software.
But the question is: are they going to do that, of Mr “Greedy “ Cook will do the barely minimum to maximize profits?
 
It depends over anandtech there is a comparison of the A13 and he stated the thermals and voltage is a bit out of control pulling more watts than the chip should with the thermals throttling performance.
All indications point to Apple pulling out of the power user desktop and laptop market as they don't make a large portion of profit and they haven't been able to gain traction of market share in the marketplace.
I imagine on their end as well, Apples strength has always been ease of use and the optimizations made to the os and applications due to limited hardware configurations. So it will be 10x easier to optimize performance for ARM and streamline any overhead from supporting separate instruction sets. The problem is ARM just doesn't have the scalability to compete with x86 and doesn't have the advantage of supporting CISC and RISC like simple instructions so it will lose certain advantages in regards to productivity and more complex software.
Hold on. Thermals and power consumption out of control were referred to A12 SoC, which was more efficient.
But we are still speaking about a 5/6W SoC giving performance comparable to 28W notebook CPU. And it still use passive cooling.
Apple has the potential for a very fast CPU here, if they don’t follow greedy Cook indications...

try to imagine a 45W 8 cores “A20”. It would be a monster.
 
Hold on. Thermals and power consumption out of control were referred to A12 SoC, which was more efficient.

More efficient because it's designed for lower power, low clock scenarios. Surprisingly Intel's Atom cores and AMD's Cat cores are more efficient than YouNameIt Lake or Ryzen. Surprisingly Atom and Cat cores lose efficiency when clock speed is even near YouNameIt Lake or Ryzen. Because they are designed to be low power low clock speeds parts.

But we are still speaking about a 5/6W SoC giving performance comparable to 28W notebook CPU. And it still use passive cooling.
Apple has the potential for a very fast CPU here, if they don’t follow greedy Cook indications...

try to imagine a 45W 8 cores “A20”. It would be a monster.

28W notebook CPU is miles faster than Apple's design. Moreover Apple has 2 "fast cores" and 4 "slow cores" whereas on PC there are always just "fast" cores. That makes comparison pointless. Intel seems to change this with Alder Lake though.

There is no potential for very fast CPU. Apple's current SOC's with high performance would be total disasters. So far Apple has shown ZERO high performance CPU designs, so it's nothing but imagination at this time.
 
More efficient because it's designed for lower power, low clock scenarios. Surprisingly Intel's Atom cores and AMD's Cat cores are more efficient than YouNameIt Lake or Ryzen. Surprisingly Atom and Cat cores lose efficiency when clock speed is even near YouNameIt Lake or Ryzen. Because they are designed to be low power low clock speeds parts.

A12 was designed for the same scenario, so NO your comment has no sense.

28W notebook CPU is miles faster than Apple's design. Moreover Apple has 2 "fast cores" and 4 "slow cores" whereas on PC there are always just "fast" cores. That makes comparison pointless. Intel seems to change this with Alder Lake though.
lol, miles faster...
I love people making things up.
Apple A12X, from 2018, was quite close to i7-8750H used on MacBook Pro mid 2018, and faster than most of the i5 used in other notebooks.

There is no potential for very fast CPU. Apple's current SOC's with high performance would be total disasters. So far Apple has shown ZERO high performance CPU designs, so it's nothing but imagination at this time.
So far Apple A12X and A12Z (basically the same CPU) are still faster than most of the mobile CPU on the market (except 10th Gen Intel CPU H-series and Ryzen 4000 Mobile APUs).
And we are speaking about a 2 years old SoC.
Apple is bigger than AMD and Intel TOGETHER and they have a very close relationship with TSMC.
The point is not if they are capable. The point is they want to... and this is not granted under Tim "Greedy" Cook guidance.
 
A12 was designed for the same scenario, so NO your comment has no sense.

28W Intel notebook CPU is NOT designed for 5-6W scenario.

lol, miles faster...
I love people making things up.
Apple A12X, from 2018, was quite close to i7-8750H used on MacBook Pro mid 2018, and faster than most of the i5 used in other notebooks.

I expect you don't have anything better to show than Geekbench. So much about that.

i7-8750 is from 2015 so basically 2 year old Apple struggles against 5 year old Intel.

So far Apple A12X and A12Z (basically the same CPU) are still faster than most of the mobile CPU on the market (except 10th Gen Intel CPU H-series and Ryzen 4000 Mobile APUs).
And we are speaking about a 2 years old SoC.
Apple is bigger than AMD and Intel TOGETHER and they have a very close relationship with TSMC.
The point is not if they are capable. The point is they want to... and this is not granted under Tim "Greedy" Cook guidance.

Faster? Where?? Geekbench???

So much about it.
 
28W Intel notebook CPU is NOT designed for 5-6W scenario.

so read again the comments above and think twice before posting.
The discussion was about thermals and power consumption bigger on A13 COMPARED TO A12, not to Intel CPUs.


I expect you don't have anything better to show than Geekbench. So much about that.

Anandtech has a very good analysis about A12 and A13, if you are able to understand it. And they defined it "desktop level performance".
It is a good article to read.

Yes Geekbench is an useful tool to compare different architectures running different operative systems.
It is not perfect, but it gives us a rough idea.
The fact that you don't like it, doesn't change its validity (it is used worldwide by tech experts).
 
Anandtech has a very good analysis about A12 and A13, if you are able to understand it. And they defined it "desktop level performance".
It is a good article to read.

They defined it wrongly. Ever heard about clickbaiting?

Fact that i9-9900K was both faster and more energy efficient than Ryzen 9 3900X tells everything about that test.

Yes Geekbench is an useful tool to compare different architectures running different operative systems.
It is not perfect, but it gives us a rough idea.
The fact that you don't like it, doesn't change its validity (it is used worldwide by tech experts).

It does not give any idea. It has absolutely no validity. It's used because there are virtually no alternatives. Being only alternative does not make it good in any way.

Just simple look tells that changing OS easily gives 10% improvement on Geekbench with x86 CPU. It's supposed to be cross platform CPU test so OS shouldn't matter at all. No need to prove any more about it's invalidity-
 
They defined it wrongly. Ever heard about clickbaiting?

Fact that i9-9900K was both faster and more energy efficient than Ryzen 9 3900X tells everything about that test.



It does not give any idea. It has absolutely no validity. It's used because there are virtually no alternatives. Being only alternative does not make it good in any way.

Just simple look tells that changing OS easily gives 10% improvement on Geekbench with x86 CPU. It's supposed to be cross platform CPU test so OS shouldn't matter at all. No need to prove any more about it's invalidity-
lol, so Geekbench and Anandtech are both wrong because ... you said so. :joy:
And it is not even just Anandtech since the analysis was reported by a relevant number of tech websites like ArsTechnica, Extremetech and dozens of others.
 
lol, so Geekbench and Anandtech are both wrong because ... you said so. :joy:
And it is not even just Anandtech since the analysis was reported by a relevant number of tech websites like ArsTechnica, Extremetech and dozens of others.

Of course I'm right. Like I said, ever heard about clickbaiting? Geekbench has proven to be unreliable and it's heavily optimized for mobile CPU's. I have enough experience not believe everything I read. Expecially on cases where results are clearly flawed.

Anandtech's analysis was done with SpecCPU 2006. Old and obsolete software, as results clearly show.

Analysis was reported some sites, and so what? Ever seen websites reporting totally imaginary rumours as fact? I see daily. And so "sites reporting about something" doesn't mean anything.
 
Of course I'm right. Like I said, ever heard about clickbaiting? Geekbench has proven to be unreliable and it's heavily optimized for mobile CPU's. I have enough experience not believe everything I read. Expecially on cases where results are clearly flawed.

Anandtech's analysis was done with SpecCPU 2006. Old and obsolete software, as results clearly show.

Analysis was reported some sites, and so what? Ever seen websites reporting totally imaginary rumours as fact? I see daily. And so "sites reporting about something" doesn't mean anything.
Sure. One of the best website on the planet is wrong, Mr Noone here is right.
If you said so...
 
Hold on. Thermals and power consumption out of control were referred to A12 SoC, which was more efficient.
But we are still speaking about a 5/6W SoC giving performance comparable to 28W notebook CPU. And it still use passive cooling.
Apple has the potential for a very fast CPU here, if they don’t follow greedy Cook indications...

try to imagine a 45W 8 cores “A20”. It would be a monster.
It doesn't scale like that, it's a pipe dream, the problem is when you start applying high wattage ARM loses what makes it powerful, there are technical challenges behind the current materials and process, you can only make things efficient for their specialization.
 
It doesn't scale like that, it's a pipe dream, the problem is when you start applying high wattage ARM loses what makes it powerful, there are technical challenges behind the current materials and process, you can only make things efficient for their specialization.
You can't know for sure.
Apple A series SoC are quite different from Qualcomm Sanpdragon SoC, and they are built since the beginning with a "desktop like" approach.
 
You can't know for sure.
Apple A series SoC are quite different from Qualcomm Sanpdragon SoC, and they are built since the beginning with a "desktop like" approach.
Per watt ARM is superior, for overall power they will never achieve parity and scaling, people have been dreaming and still are waiting for far too long. Fujitsu has to use a staggering amount of chips in the supercomputer to achieve above level of performance and that was on a better process node than desktop or server processors.

It's a case of the future, but desktop processors and companies aren't standing still AMD broke Intel's marketing strategy of selling the same slightly refined product year after year R&D will improve year after year, that stifling of the market allowed ARM to catch up, I doubt the future will bode well however.
 
You are speaking of ARM as if it was a fixed design. It is not. Apple ARM implementation is different from other’s.
You can’t know yet how scalable it is (the only indications we have are the tablet X models, somewhat bigger than smartphone counterpart, and it worked well).
 
Back