Got a tip for us?
Let us know
Become a MacRumors Supporter for $50/year with no ads, ability to filter front page stories, and private forums.
TSMC to Remain Sole iPhone Chip Supplier, Could Provide ARM-Based Chips for Mac in 2020, Apple Car Chips in 2023
- Thread starter MacRumors
- Start date
- Sort by reaction score
Lol. I remember when you had to run Primetime and cadence on Solaris, and people were saying “i would never buy a Linux x86 computer...”
Also, even back when i designed microprocessors for AMD and Sun and Exponential, my work computer (sparc, linux, whatever) was very different from my home computer. I had no interest in doing a place and route at home (other than by Vnc). So I wasn’t making purchasing decisions for work, and they weren’t making purchasing decisions for me at home.
I would think Apple have been tinkering with MacOS on ARM for at least as long as iOS has existed, given the amount of shared code. In that way, and with iOS being such a superlative platform for developers Apple are probably way better placed to move to ARM now than they were to move to intel from power pc, particularly given they are also going to have 100% control over the chip’s design.
‘Massive base of software’ is a bit of a stretch for MacOS - it’s actually a relatively software poor platform. Making it more interoperable with iOS will, again, mean iOS developers will be able to step across onto the Mac platform much more easily.
I suspect in 2020, the mac-mini will be using the same chips they've been using for the past four years.
They wouldn't. And besides, the logical selection is AMD with the Zen 2 7nm already done at TSMC, and being released this Spring. The architecture is superior to Xeon in every conceivable means, it works with Thunderbolt and it's 50% less expensive.
With TSMC being the supplier for both Apple and AMD there are no more excuses for them to keep using Intel.
HobeSoundDarryl
macrumors G5
How can we think this? Think it through: some "Mac" software will run on some Macs but not others? Consumers will need to get software savvy enough to ask is this for ARM-based Macs or only Intel Macs?
And let's not swallow this "just throw a switch in a compiler" nonsense so easily. It is never as easy as that for all software much beyond maybe "Hello World" apps. Note how many apps on the massively-more popular iPhone could not even make the leap from 32 bit to 64 bit. If throwing a switch to go from Intel to ARM is so easy, certainly throwing a switch to go from 32 bit to 64 bit would be just as easy... and yet, instead, those apps just didn't make the transition at all... even with plenty of notice... and even with far more potential users than equivalent apps would ever have on Macs. Let that foretell how it goes for lots of Mac apps should (much less popular) Macs (relatively) hop from Intel to ARM.
Actually, for most software it is as easy as flipping a switch.
The switch is going to happen. Likely only for non-Pro Models for a few years. And developers will start to get their Software running on it. By the time that Apple is ready with the Pro machines, Developers will be ready too.
I'm also not worried about Bootcamp. Windows on ARM is one of the biggest priorities of Microsoft at the moment. It already emulates x86 programs at a reasonable speed. x64 should follow soon.
But noone will be forced to buy an ARM Mac for a very long time. They will likely coexist for quite some time.
I'm also not worried about Bootcamp. Windows on ARM is one of the biggest priorities of Microsoft at the moment. It already emulates x86 programs at a reasonable speed. x64 should follow soon.
But noone will be forced to buy an ARM Mac for a very long time. They will likely coexist for quite some time.
HobeSoundDarryl
macrumors G5
I'll believe it when I see it. Why didn't all the 32-bit popular apps that didn't survive the 64-bit mandate just flip one little switch and recompile?
And why did we need Rosetta back in the PPC to Intel transition... and need it for a couple of years while software was redone to make it work with a different Mac CPU? I still have a bootable Snow Leopard drive to be able to use some software that STILL hasn't been "flip 1 switch to" work with macOS.
I'll try to be as optimistic as anyone here, but I've been through this... and code myself. I'll concede that some software might be so easily recompiled but I'll not believe it (until I see it) that lots of popular Mac software will be able to flip to ARM overnight if it's just flipping a switch in a compiler.
Why the assumption that this would be an A series chip? It could be a different design called Mac series CPU / GPU with similar instruction set to A series yes.
Because they didn't think the sales would be worth even the minimal effort? Just because I update something to 64-bit doesn't mean anyone wants to buy it. Not to mention that 64-bit wasn't the only change - typically there are other things that need to be updated with new OS versions, so any time you touch the source code you need to address these other things too.
HobeSoundDarryl
macrumors G5
Tell that to people who use their Macs to work with Clients. Windows rules the business world and Macs are (relatively) rare. I have no great love for Windows myself but, IMO, the "killer app" for Macs is the ability for one working laptop to be a 2-birds-with-one-stone computer. Client needs me in Windows mode- my Mac can do that. Client OK with Mac mode- my Mac can do that.
If future Mac loses Bootcamp, the business computer that goes is probably a Windows laptop. Why? Because odds are, it will be easier to interface with the real world than taking a Mac and hoping it will be OK.
Again, no love here for Windows- just being real in terms of using computers and computer software for work. I WISH that Macs fit in as well as many of us here like to imagine.
Yea, that's fair and where my mind immediately goes. I just think protecting their investment in one the best fabricators out there may be a smart move.
[doublepost=1539803509][/doublepost]
Good point, maybe they do it that way to keep compatibility. It would be impressive.
The number of people using bootcamp is very small. If you need a windows machine in business, chances are you purchase a windows machine. You can get one for a couple hundred bucks. You can also use parallels, vmware, etc. Losing bootcamp will be fine. Nobody will care.
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.
You know more than me in this area. I hadn't even thought of it that way.
HobeSoundDarryl
macrumors G5
But with much-less-popular Macs, software will be able to hop from Intel to ARM with...
So which is it? Is it "there are other things that need to be updated" and "any time you touch the source code..." or is it "as easy as flipping a switch"?
That's because the A12X doesn't need those things. Obviously a chip built for a laptop or desktop would be different.
It would have the same instruction set. ARM.
All of them? Anyhow my main point is that it doesn’t have to be a CPU that’s used in mobile phones. Desktops and laptops can both support plenty more power and heat than a phone.
HobeSoundDarryl
macrumors G5
I use bootcamp in business. I don't want to but clients are not always macOS friendly/capable/setup or have software or networks that require something that only runs on Windows.
I definitely do NOT want to take 2 laptops if one can do most/all of the job. So again, in this future Mac world where there's no bootcamp, the working man computer choice gets much more complicated than the "no brainer" (2-birds-with-one-stone) choice now. Except in odd situations, the working computer NEEDS Windows capabilities but only (maybe) WANTS Mac capabilities.
Taking a stake in them might be a possibility, though I'm not certain it would necessarily give Apple any more clout than it already has as a large volume, big name client? For outright ownership though, it would probably cut right into Apple's margins - and considering how much pressure they're under from Wall St. I would say nonstarter.
If it works anything like the AX chips, likely they'd repackage existing core designs, probably tweaked and clocked higher to take advantage of higher power draw and thermal headroom. That's basically what the X chips in iPads are (plus a GPU bolted on that's man enough to run the bigger, higher res screens). So for an "A12M" variant you'd get, say, 4-6 of the performance "Vortex" cores from the vanilla chip and maybe also 4 of the lower power "tempest" cores to handle background tasks. Of course if they really wanted to push into a 15-20W TDP, for example, it might become worth it to completely redesign the cores around that, I'm not honestly sure how well it would scale beyond iPad level (~7.5W).