History Repeating? (68K --> PPC --> Intel --> ARM)

[im_to_hyper]

Scenario: Intel (IBM in 2004) hasn't been able to bring the thermals down in order to ensure a powerful, thin and light laptop that sips battery. The Core series (G5 in 2004) is simply too hot even at 10nm (90nm in 2004). There have been promises of a cooler-powered CPU with the next 7nm Core architecture (POWER5/G6 in 2004) but delays have made Apple grow impatient and source other CPUs.

------

As someone who recently crossed that 30 year threshold in life, I've seen (and been through) quite a few Mac transitions. This whole ARM rumor thing is starting to feel suspiciously like 2003-2005 when Intel rumors were swirling around here on MR.

The speculation is so similar. The "should I buy" posts are so similar. The "wondering how Apple will support old apps" curiosity is so similar.

As a long-time Mac user I've become accustomed to Apple's architecture shifts whenever necessary in order for the hardware to meet their expectations.

-------

Each time Apple changes architecture; a paradigm shift in computing seems to happen and new doors open. What will happen this time?

 

[repairedCheese]

Well, unless they plan on having universal binaries for a good long time, this means the new Mac Pro is already doomed to running Windows or Linux in about 5 years. That's the thing, too. Computers, new, high end computers, last a lot longer than they use to. So even if x86 isn't good enough for Apple, it's going to be good enough for a lot of people for a long time.

What's more, last time they moved from an architecture no one supported to one everyone did. This time? They're moving from that same architecture that still gets that same support. Don't get me wrong, we may see Windows for Arm on Mac, but I have doubts about bootcamp, among other things.

Is it possible they will have Arm and x86 simultaneously? Sure! But I genuinely doubt that. Given how Apple ended 32-bit support on all their hardware, why would they run two different families of cpu at the same time if they don't have to?

 

[RogerWilco6502]

Do we know why Apple switched from 68k to PPC? I haven't done much research into that transition, not as much as I have on the PPC to Intel one at least.

 

[maflynn]

Do we know why Apple switched from 68k to PPC?

If memory serves me, performance, also with the PPC, Apple would have to manufacturers (Moto and IBM), instead of just Moto

 

[RogerWilco6502]

If memory serves me, performance, also with the PPC, Apple would have to manufacturers (Moto and IBM), instead of just Moto

So similar reasoning to the other architecture changes?

 

[eyoungren]

Do we know why Apple switched from 68k to PPC? I haven't done much research into that transition, not as much as I have on the PPC to Intel one at least.

So similar reasoning to the other architecture changes?

Apple and software designers needed room to allow the OS and software to do things it could not on the old chips. So yeah, same old story.

Around here we have long passed around the term 'universal' when referring to apps and OS that can run either on PowerPC or Intel. I can remember, when the word being used was 'fat'. If you had a fat app, it was one that could install on the old chips or PowerPC.

 

[RogerWilco6502]

Apple and software designers needed room to allow the OS and software to do things it could not on the old chips. So yeah, same old story.

Around here we have long passed around the term 'universal' when referring to apps and OS that can run either on PowerPC or Intel. I can remember, when the word being used was 'fat'. If you had a fat app, it was one that could install on the old chips or PowerPC.

Ok, cool. Thanks!

Yeah, that aligns with what I remember from my research: FAT Binaries are 68k/PPC and Universal Binaries are PPC/Intel.

 

[filmbuff]

They basically already make Arm computers - the ipad. I think it’s inevitable they’re going to start to use their own chips in a laptop form, probably soon. I think they will run concurrently with Intel for quite a while. At least a few years?

 

[rampancy]

I think it's more than just that. Apple has weathered multiple successive points where a large part of their ability to succeed, or at least deliver on their intended goals, were tied inextricably to another company: there were the PPC development and scaling issues with Motorola (and later IBM), and their traditional dependence on Adobe and Microsoft to provide architecture-optimized versions of their premiere apps to drive Mac sales. I even remember reading once about how the initial dearth of PPC-optimized software was due to a critical delay in the release of Metrowerks' PPC-optimized version of CodeWarrior.

Now Apple is likely going to be making its own wireless modem chips for iPhones and iPads, removing their dependence on Qualcomm.

 

[rampancy]

Do we know why Apple switched from 68k to PPC? I haven't done much research into that transition, not as much as I have on the PPC to Intel one at least.

Performance and frequency scaling on the 68040 had stalled; even though it was supposedly superior clock-for-clock with the 80486, Intel had gotten the 80486 to scale much higher, much faster (Apple's Star Trek port of System 7 to the 80486 was supposedly much faster than on 68040). And the Pentium was coming too.

Apple had done research on an internally-designed RISC-based quad-core/SIMD CPU that was called "Scorpius", part of a wider project called Aquarius. That project later fell through but it showed that movement to an RISC-based CPU was the way of the future. Soon after, IBM approached Apple about developing a desktop version of its POWER CPU architecture, and Apple brought on Motorola to add manufacturing capacity -- that led to the AIM alliance and PowerPC.

 

[RogerWilco6502]

Performance and frequency scaling on the 68040 had stalled; even though it was supposedly superior clock-for-clock with the 80486, Intel had gotten the 80486 to scale much higher, much faster (Apple's Star Trek port of System 7 to the 80486 was supposedly much faster than on 68040). And the Pentium was coming too.

Apple had done research on an internally-designed RISC-based quad-core/SIMD CPU that was called "Scorpius", part of a wider project called Aquarius. That project later fell through but it showed that movement to an RISC-based CPU was the way of the future. Soon after, IBM approached Apple about developing a desktop version of its POWER CPU architecture, and Apple brought on Motorola to add manufacturing capacity -- that led to the AIM alliance and PowerPC.

Interesting. Thank you!

 

[sawpits]

I would imagine that the new A14 chip will be the first into an Apple computer.

 

[Macbookprodude]

Well, unless they plan on having universal binaries for a good long time, this means the new Mac Pro is already doomed to running Windows or Linux in about 5 years. That's the thing, too. Computers, new, high end computers, last a lot longer than they use to. So even if x86 isn't good enough for Apple, it's going to be good enough for a lot of people for a long time.

What's more, last time they moved from an architecture no one supported to one everyone did. This time? They're moving from that same architecture that still gets that same support. Don't get me wrong, we may see Windows for Arm on Mac, but I have doubts about bootcamp, among other things.

Is it possible they will have Arm and x86 simultaneously? Sure! But I genuinely doubt that. Given how Apple ended 32-bit support on all their hardware, why would they run two different families of cpu at the same time if they don't have to?

I am sure Apple will abandon Intel like it did to PowerPC since Intel has failed Apple. Watch, they will do the same as they did to us.. Tim Cook will say just like Steve "And we have many new intel products in the pipeline" - then after a year, it will be dead as a platform on the Mac just as Steve got rid of PowerPC, though he could have kept it going till 2010 at least.

 

[DearthnVader]

Scenario: Intel (IBM in 2004) hasn't been able to bring the thermals down in order to ensure a powerful, thin and light laptop that sips battery. The Core series (G5 in 2004) is simply too hot even at 10nm (90nm in 2004). There have been promises of a cooler-powered CPU with the next 7nm Core architecture (POWER5/G6 in 2004) but delays have made Apple grow impatient and source other CPUs.

------

As someone who recently crossed that 30 year threshold in life, I've seen (and been through) quite a few Mac transitions. This whole ARM rumor thing is starting to feel suspiciously like 2003-2005 when Intel rumors were swirling around here on MR.

The speculation is so similar. The "should I buy" posts are so similar. The "wondering how Apple will support old apps" curiosity is so similar.

As a long-time Mac user I've become accustomed to Apple's architecture shifts whenever necessary in order for the hardware to meet their expectations.

-------

Each time Apple changes architecture; a paradigm shift in computing seems to happen and new doors open. What will happen this time?

With each of Apple's previous transitions 68k->PPC PPC->x86 it required that the new CPU could run the old code at acceptable speeds while developers ported their code.

Arm can not run x86 code at anywhere near "acceptable" speeds.

Tho, it would seem Apple is going to go that way because they can fab their own Arm cpu's.

I don't think Apple will dump x86, likely the Air the Mini and the MacBook will get Arm, and maybe the low end iMac, while the Pro line will retain x86.

Apple will do it just to make more profit, not because Arm is better for desktop macOS, and for that reason I think it's a mistake.

 

[Amethyst1]

Apple's Star Trek port of System 7 to the 80486 was supposedly much faster than on 68040)

Oh, the irony.

 

[DearthnVader]

This whole Arm for a desktop OS has been tried and failed by other vendors, Arm Chromebooks, and Arm Winsdows systems.

The place where Arm can compete with x86 is in the price system space, if Apple is willing to price the Air @ $499, it could be successful, but Apple has never been one to try and compete in the low end space based on price.

But just because they haven't really tried that before, doesn't mean they won't try it now.

If I could buy an Arm Air in the $499 price range that runs the desktop macOS compiled native for Arm, I think I would be a buyer for that type of system.

 

[Macbookprodude]

With each of Apple's previous transitions 68k->PPC PPC->x86 it required that the new CPU could run the old code at acceptable speeds while developers ported their code.

Arm can not run x86 code at anywhere near "acceptable" speeds.

Tho, it would seem Apple is going to go that way because they can fab their own Arm cpu's.

I don't think Apple will dump x86, likely the Air the Mini and the MacBook will get Arm, and maybe the low end iMac, while the Pro line will retain x86.

Apple will do it just to make more profit, not because Arm is better for desktop macOS, and for that reason I think it's a mistake.

Maybe ARM macs will be a lot cheaper possibly ? I think Intel was worse than PPC. Granted, a G4 Sawtooth in 1999-2000 was 1599.99 for the base model 350, 1799.99 for the 400, up to I think 2999.99 for the 500 mhz chip where the 2019 Mac Pro is over 6000.00 at base up to 35k ! Insane if you ask me, but who knows if I am still around in 5-10 years the 2019 Mac Pro will be around 1500.00

 

[DearthnVader]

Maybe ARM macs will be a lot cheaper possibly ? I think Intel was worse than PPC. Granted, a G4 Sawtooth in 1999-2000 was 1599.99 for the base model 350, 1799.99 for the 400, up to I think 2999.99 for the 500 mhz chip where the 2019 Mac Pro is over 6000.00 at base up to 35k ! Insane if you ask me, but who knows if I am still around in 5-10 years the 2019 Mac Pro will be around 1500.00

Inflation is a bitch.

Apple systems are price competitive with "Windows" offerings for desktop workstations with comparable hardware.

 

[Amethyst1]

This whole Arm for a desktop OS has been tried and failed by other vendors, Arm Chromebooks, and Arm Winsdows systems.

... because the systems (or at least the Windows ones) are too expensive given their performance and app availability issues IMHO.

 

[rampancy]

Maybe ARM macs will be a lot cheaper possibly?

I really don't think so, not unless there's strong enough market demand among Mac users for a low-cost headless/SFF machine, which the Mac mini was originally supposed to be (and I guess still is, sans the "low-cost" part). There's also the market for low-cost portables for education, but that's pretty much where the iPad is.

During the PPC-x86 transition, a lot of people were instantly assuming that Macs would become significantly cheaper, simply by virtue of x86 Intel CPUs being more widely produced and a perception that Apple wanted to appeal to traditional Windows customers. As we saw, Apple was just content to go with the status quo and continue to reap its high margins, for better or for worse. I don't really expect things to be any different with the x86-ARM transition, if/when it comes.

 

[Raging Dufus]

Haven't seen anybody mention this in here, but I thought it was a cool idea. How about a hybrid Mac?

Apple, ARM, and the coming of the hybrid Mac

There’s been a lot of debate about whether Apple would ditch Intel processors and start making Macs with ARM processors like the ones that power the iPhone and iPad. Those discussions have generally been about a binary choice between Intel and A...

sixcolors.com

There's no real reason for Apple to abandon x86. They can have the best of both worlds, and produce something fundamentally different from the rest of the industry that outperforms anything anyone else has to offer.

It's like the idea behind universal binary, except instead of emulating hardware in software, you have both types of hardware running inside the same computer. This would be a true innovation, and I'd like to think it's where they're headed. Would be even better if they shift their chip supplier for the x86/64 side from Intel to AMD.

Heck, I'd buy it!.... when it's about ten years old

 

[Bug-Creator]

Do we know why Apple switched from 68k to PPC?

Cos Moto pretty much mothballed the 68k line (and their own 88k RISC line) in favour of PPC.

The 68050 got canceled and the 68060 was a last hooray before turning existing costumers to ColdFire which never was 100% compatible.

The switch was so bad that there was a timeframe when a virtualized 68k Mac running on an 68060 Amiga was faster than anything Apple could offer (even if that was running the same app in PPC native).

 

[RogerWilco6502]

Cos Moto pretty much mothballed the 68k line (and their own 88k RISC line) in favour of PPC.

The 68050 got canceled and the 68060 was a last hooray before turning existing costumers to ColdFire which never was 100% compatible.

The switch was so bad that there was a timeframe when a virtualized 68k Mac running on an 68060 Amiga was faster than anything Apple could offer (even if that was running the same app in PPC native).

Oooof... ouch! That hurts that Motorola did that >.>

 

[Bug-Creator]

That hurts that Motorola did that

Did what?

25 years ago everyone was certain that CISC (x86/68k) was a deadend and Moto didn't have the resources/marketshare to let the 68k line defy that reality as Intel could.

Development of the 050 hit a few roadblocks and was delayed to the point that the team working on the 060 almost overtook them.

They could have furthered their 88k line, but that switch would have been as painful as the one to PPC.

So the joined IBM and Apple to form AIM and pushed for PPC.

The 601 used in early PPC-Macs were clocked 60MHz and upwards making them somewhat faster then the 68060 clocked at 50MHz.

But.....

Those early PPC-Mac had a rather slow RAM interface crippling it's real life performance.
The whole OS was still in 68k and the emulator running it was superslow.
Neither was Moto's fault.


Now when we get to why PPC was dropped a decade later, that can actually be laid at Moto's feet for not coming up with anything substantial after the G4 (and through countless name changes they never did till today).

 

[RogerWilco6502]

Did what?

25 years ago everyone was certain that CISC (x86/68k) was a deadend and Moto didn't have the resources/marketshare to let the 68k line defy that reality as Intel could.

Development of the 050 hit a few roadblocks and was delayed to the point that the team working on the 060 almost overtook them.

They could have furthered their 88k line, but that switch would have been as painful as the one to PPC.

So the joined IBM and Apple to form AIM and pushed for PPC.

The 601 used in early PPC-Macs were clocked 60MHz and upwards making them somewhat faster then the 68060 clocked at 50MHz.

But.....

Those early PPC-Mac had a rather slow RAM interface crippling it's real life performance.
The whole OS was still in 68k and the emulator running it was superslow.
Neither was Moto's fault.


Now when we get to why PPC was dropped a decade later, that can actually be laid at Moto's feet for not coming up with anything substantial after the G4 (and through countless name changes they never did till today).

Ohhhh, ok. Sorry, I thought you were saying that 68k could have been advanced further, Moto just chose not to do that.

On a side note, does the 68k emulator (for applications) still exist in Mac OS 9? It probably does given that I have to ask, I just want to know for certain