TSMC to Remain Sole iPhone Chip Supplier, Could Provide ARM-Based Chips for Mac in 2020, Apple Car Chips in 2023

[techwhiz]

This is possible. I always thought that the MacBook would get it first. The problem is that if the MacBook gets ARM, it's already going to be about as fast as the 13" quad core Intel MacBook (we'll see what the A12X benchmarks show). The other problem is the question of why it has taken Apple this freaking long to build the new Mac Pro if they weren't working on putting ARM in it? Surely since they've rethought it top to bottom to be modular they will want to build their own proprietary components. ARM seems like a good way to lock that up while also providing speed that the competition can't match—especially since they can fine tune macOS to take full advantage of this hardware.

There is more than just a synthetic benchmark to performance.
The A12x does not support multichannel DDR, does not have multiple lanes of PCIe along with the sizes and and types of caches of Intel processors.

It might beat a low end MacBook, but not a MacBook Pro.
I for one would never buy an ARM based computer, because my work life runs on RedHat/CentOS designing chips for a living. When Synopsys, Cadence and others port their tools to an ARM processor then I'll think about switching.

 

[Baymowe335]

Unless I am mistaken, It seems that Apples reliance on TSMC is only growing. Should Apple buy the company? Or maybe put together a consortium like they did with Toshiba?

No. TSMC is a massive hardware manufacturer with huge capital requirements. Apple doesn’t want to be in that business.

Not only that, TSMC is worth over $200B. Apple can buy almost any company, but this is even too big for Apple with premium.

 

[obiwan]

It will completely break compatibility. You would need to emulate the X86 -> Slow.
Right now, Boot Camp works because Macs are X86.

I thought that might be the case. If future Macs break this compatibility, I’d hazard a guess that this would be a sure fire way to kill off the Mac. Either that, or suddenly gain significant market share from the PC market (which will never happen!)

 

[ValO]

This rumor keeps popping up.
Apple is not making transitioning from Intel to ARM.
A mobile processor is a far cry from a desktop processor. The reason for ARM in cell phones is that there were two camps ARM and MIPS. ARM won. The first smartphones were actually MIPS (Windows Mobile).

For desktop processors see the new Xeon and Ryzen Thread Ripper for reference.
Symetric processing, multilevel caches, multi-socket support, multithreaded, PCIe root complex with 48 to 64 lanes of Gen4 (by the time they would come out).

The eco system for ARM existed in the mobile space. It does not exist in the desktop space.
Yeah, I know there are ARM Windows devices, and Chrome OS laptops.

I for one would not buy another MacBook Pro or MacPro if they had ARM.
I use VMWare to run CentOS/RedHat x86 and I don't have a substitute for running that for chip design and simulation tools. I'd switch to a Windows machine.

Emulating x86 on ARM would be too slow.
I currently own a MacBook Air, two MacBook Pros, and a MacPro tower that I will not replace with that coffee can thing.
If they release another tower with current Xeon or Ryzen chips, they can have my money. They can have my money as long as they stay intel compatible. Moving away from Intel compatibility, means I move away from Apple.

Yeah, I was waiting for a reaction like that. It was rather late this time.

Seeing what apple can do with an a series soc and imagining what they could do with an arm based cpu without some thermal, power and space restrictions of a phone , I think you would be surprised.

Next step: release some kind of arm based MacOS(for simplicity, skip the finder/filemanager for now) with the next iPad Pro , together with a new Smart Keyboard cover with trackpad.
This could help developers with preparing for the future.

 

[jjhny]

The desktop paradigm as we know it is dying

Yes, for consuming content mostly. But creating is another story. I use a 38" ultrawide monitor and it just barely gives me enough room for all the windows my apps need to have open to do any work. Also use a 24" Wacom screen (weighs 58 lbs with its stand).

Many associates run two to three 30" plus sized screens. Run drive arrays. Use PCI cards for various purposes, including array controllers, etc.

If you are consuming content only, sure. I use a iPad Pro for email and web surfing, reading books, PDF manuals, checking the weather and watching youtube videos. Fine for being a consumer.

 

[Glideslope]

Maybe it’s time to buy TMSC. On the other hand, we have no idea what agreement they have in place. They very well may be an exclusive partner well into the future, but it makes sense to stay flexible on Apple’s part. Either way, good to see. I expect big things from both companies in the coming years. Can’t wait till they drop Intel, fix macOS, and come up with some new computing concept ideas. The desktop paradigm as we know it is dying, let’s see what’s next aside from a strictly mobile experience.

It would be interesting to see how the Taiwanese Govt. would view such an acquisition if it were ever to be floated. I could see opposition.

While TSMC clearly has the 7nm fab down at this point it could be an issue as a sole provider. The geology around Fab 15 is not the most stable. My understanding is all of Apple’s Silicon comes from that one Fab. Another big earthquake could be crippling.

Pluse the trend is to outsource chip fabs today. Such as AMD and IBM have done.

 

[theluggage]

Will current apps on the Mac (intel) run on ARM-based processors??? OR will developers have to rewrite everything?

Well it will vary on a case-by-case basis, but for a modern App written in high-level-language and relying on OS frameworks for hardware access (Metal etc.) it should be just a case of ticking "ARM" in the compiler settings and re-compiling. OK, so not everything will be that simple, but the need for major re-writes should be rare.

MS already have Office on ARM (not the Mac version - but one would assume that lots of the code is shared so any major hurdles should have been solved). Adobe have just announced "full" Photoshop for iPad - and hence ARM (and I'm sure Apple will have given them the heads-up if MacOS on ARM is in the pipeline). Lots of other Mac developers are now offering iOS versions of their products (and, where possible, will have avoided having two sets of code).

Also, MacOS is dropping 32-bit support after Mojave, so an awful lot of old, no-longer-supported applications and plug-ins aren't going to run on 2020 Macs anyway.

 

[whooleytoo]

Losing virtualisation would be a deal-breaker for me. The only reason I'm able to use Macs at work is because I have a multiple VM setup, running Mac & Windows side-by-side. Lose that, and I have little justification for having an office Mac.

(And at home, my iMac is almost permanently boot-camped into Windows for Elite Dangerous. If upcoming Macs can't be booted into Windows, there's no reason to own a Mac at home either. I can buy a faster gaming PC, cheaper, which is easier to upgrade).

I don't see the end-user benefits of this.

 

[Joniz]

Oh, I didn’t know Apple made Macs anymore.

Well, I guess that tells us when the Mac Mini will be updated. Hoped it would be this year, but…oh, well.

 

[Internet Enzyme]

This is possible. I always thought that the MacBook would get it first. The problem is that if the MacBook gets ARM, it's already going to be about as fast as the 13" quad core Intel MacBook (we'll see what the A12X benchmarks show). The other problem is the question of why it has taken Apple this freaking long to build the new Mac Pro if they weren't working on putting ARM in it? Surely since they've rethought it top to bottom to be modular they will want to build their own proprietary components. ARM seems like a good way to lock that up while also providing speed that the competition can't match—especially since they can fine tune macOS to take full advantage of this hardware.

I think this is one of the issues apple has internally had with the mac pro. Initially i believe the mac pro was ignored due to the reality that it is a niche product, but then when the mac team finally got around to looking at a refresh they realized that it simply wasnt going to happen with that design. All the while, they have to reconcile the timing issue of its impending 2019 release with the more cosmic and grand goal of porting the mac to arm. If they had regularly updated the mac pro at least once or twice since 2013, then they'd have a case to delay the 2019 mac pro until arm is ready. But at this point they pretty much have to update it and they must figure that pros wont want arm initially anyways.

 

[ph001bi]

Will current apps on the Mac (intel) run on ARM-based processors??? OR will developers have to rewrite everything?

Officially, they redeveloping their GUI builder to help App transition from ARM to Intel, unofficially they are probably working for it to work both ways, so MacOS app will just need a recompile.
They could also start mandating that all Apps be compiled using their in House solution starting June 2019, we will then have the first ARM powered Macbook released in May-June 2020 followed by Arm Powered Mac mini October 2020, followed by Macbook Pro in June 21, followed by Imac in October 2021 and MacPro in June 2022.

 

[cardfan]

A move to an ARM chip means me probably switching to windows and basically the death of macs. Not being able to run windows is a dealbreaker. These "full" versions of Office or photoshop aren't really that. Windows Office does not equal and will never equal Office on iOS.

I understand Apple does great things with their mobile chips but hardware is hardly their problem in the Macs area.

 

[kemal]

Pro customers will have nothing to worry about. Software developers will make their latest versions ARM compatible. And the new Apple hardware will overheat just like always.

 

[DNichter]

Yes, for consuming content mostly. But creating is another story. I use a 38" ultrawide monitor and it just barely gives me enough room for all the windows my apps need to have open to do any work. Also use a 24" Wacom screen (weighs 58 lbs with its stand).

Many associates run two to three 30" plus sized screens. Run drive arrays. Use PCI cards for various purposes, including array controllers, etc.

If you are consuming content only, sure. I use a iPad Pro for email and web surfing, reading books, PDF manuals, checking the weather and watching youtube videos. Fine for being a consumer.

Correct, for creation it's definitely a different story, but it will become less and less as more and more desktop-like features come to iPad. A desktop in it's traditional form will likely only be used by large corporations and niche creators, but it will become less and less as time goes on.

 

[fokmik]

Question: Why would Apple spend years redesigning the Mac Pro to release them on Intel chips in 2019 only to move everything to ARM in 2020?

Either they're moving to ARM in 2019 or a few years later. If Apple can do for desktop chips what they did for mobile chips, then this is going to be amazing.

this will be only in macs like mac mini,12" macbook, macbook air or entry level imac...these are not for mac pro or macbook pro or 27" imac or whatever that is using quad core/six core i7 or i9

 

[Appleaker]

Based on my information, the switch to ARM for Mac will start in 2021, but 2020 is possible for an introduction or perhaps the introductory Mac, if it’s target is developers, it may be a Mac mini but the first consumer focused product will no doubt be a MacBook.

 

[fokmik]

Pro customers will have nothing to worry about. Software developers will make their latest versions ARM compatible. And the new Apple hardware will overheat just like always.

Can ARM tech can be in near future compatible with x86 apps?
I mean my sun from time to time loves to play league of legends on my mac...but this game is only for macOS or windows for x86 cpu

 

[wigby]

I think they’ll keep the Pro line using Intel initially. I wouldn’t expect Apple designed desktop level chips until 3 years from now, personally. They’ll likely start with a MacBook maybe the Mac Mini to test the waters.

I think Apple will definitely add A series chips to all high end Macs alongside Intel chips. The A series chips in iMac Pro already handle a lot of security and I/O tasks which allow the Intel CPUs and GPUs to work more efficiently.

 

[Falhófnir]

Maybe it’s time to buy TMSC. On the other hand, we have no idea what agreement they have in place. They very well may be an exclusive partner well into the future, but it makes sense to stay flexible on Apple’s part. Either way, good to see. I expect big things from both companies in the coming years. Can’t wait till they drop Intel, fix macOS, and come up with some new computing concept ideas. The desktop paradigm as we know it is dying, let’s see what’s next aside from a strictly mobile experience.

Apple are a strictly design company; generally the manufacturing side of things is deadweight which requires a huge amount of investment (think updating machinery to handle <7nm chips) so I can’t see Apple moving towards in-house fabrication.

 

[now i see it]

Soldering ARM chips into the Mac will do zero to help move the Mac forward. It could very well be the tipping point of the beginning of the end. I doubt it will happen.

 

[fde101]

ARM was originally used on desktop computers, back before the whole "mobile" thing existed.

It just turned out to be a good architecture for low-power applications so it has been used extensively in that field now, but I wouldn't be too quick to rule it out as a possible contender for the desktop world.

Also, people seem to be ignoring the possibility that Apple might opt to make its own x64-architecture chips for their computers rather than more advanced ARM-based ones...

or even something else completely.

 

[guzhogi]

If Apple does go to ARM, I'm curious to see if they'll make a stick PC? Call it a Mac Nano or something?

I realize that an iPhone/iPad CPU is way different than the CPU in something like a Mac Pro, but I wonder if Apple could make that jump? I just wonder what Apple could do with an A# chip can do with the power & thermal limits of a Mac Pro would be like.

 

[SeminalSage]

It's excellent that A-series chips are rivaling desktop CPUs, but I wish Apple's assistant would catch up…

View attachment 795924

I'm not good enough with an editor to post a proper reply, but:

Microsoft: "Brake before we collide with the car ahead of us!"
"Update 42% complete. Please do not turn off or reset your car."

Google: "Brake before we collide with the car ahead of us!"
"You can skip this ad on brake shops in 4...3...2..."

Bixby: "Brake before...why are the brakes on fire!"

 

[bjet767]

The old ARM vs Intel chip debate goes on...

Before we see any new Apple product it has probably been in design and pre-production process for as long as three years. Additionally the software base is currently all compiled for the Intel chip instruction set. Changing the CPU to ARM would require an immediately available emulation chip or software that would not out date the current massive base of software. Of course Apple could have both CPUs installed in their Macs, but that would be cost prohibitive and really just a silly notion.

I would expect Apple would introduce the ARM chip based PCs at one of the Developer's conferences and make Xcode completely seamless, as far as coding goes.

 

[cmaier]

This would be a really interesting development for macOS if it happens. I’m not sure if this would break compatibility with running Windows (via boot camp), which let’s face it, is one the reasons the Mac didn’t ultimately slide into oblivion.

It’s much more important now for Mac to be “compatible” with iOS than it is for it to run boot camp.

The days of Windows mattering are done.