Did anyone else feel hoodwinked when the Intel switch was announced ?

[B S Magnet]

I disagree.

Firstly, the CPU architecture is far less important than the software experience. For the first decade plus, Mac systems were significantly superior to the alternatives. Pricey, yes, but there really was no comparison. Many have argued that it wasn't until Windows 7 that there was a truly comparable user experience with ease of use and pleasure of use. (An option that I agree with.)

One could even give XP a pass here, but otherwise, yes, Windows 7 was a big step forward.

But since this topic is about hardware:

1. Macs started with 68K Motorola. This architecture was very competent. The Quadra / 68040 variants were notably faster than their Intel counterparts.

The Macs on 68040s were absolutely powerhouses for their moment. The next step beyond were either by levelling up to NeXT or SGI.

Even so, there were only a few major applications being released for a time to run on POSIX-based OSes (even Adobe got into it for a brief time with UNIX versions like for FrameMaker). So for a time, Macs were the only/main game for running audio/video/imaging-oriented commercial software. It would take a while for companies as Adobe, Quark, and Digidesign to begin churning out ports of feature-parity software for Windows (which, for their own reasons, were glitchy initially and, when released, would not come out for many months following the Mac release).

1. The PowerPC architecture was beast at the time it was released and for most of the the over 10 years that it was used. The majority of that time it outperformed Intel machines. Combined with the more pleasing and substantially easier to use Mac OS, PowerPC was a good thing.

It outperformed so long as Apple held up their end up the AIM alliance bargain. When Jobs came back, that alliance began fraying. Jobs chose favourites (Motorola for PPC7400, IBM for PPC970). This was the death knell for unified chip design and production by Motorola/Freescale and IBM. And yes, I do put this fraying at Jobs’s feet, as Jobs inherited the AIM alliance. As such, he treated it as his ginger-headed step-child.

1. What killed the PowerPC wasn't that it was "the wrong horse" but that the FreeScale / IBM alliance had worn out. IBM was no longer interested in devoting engineering to PowerPC chips for desktops and laptops because at that point Apple was the only major buyer and there wasn't enough demand from Apple to warrant IBM's continued engineering and manufacturing investment. The architecture was entirely capable of continuing, but the lack of improved laptop capable PowerPC chips was a killer. This was all the result of business choices, not any inherent problems with the PowerPC architecture itself.

See above.

The alliance “wearing out” was on Apple. I conjecture had Apple, under Jobs, been as committed an even partner to the continued cultivation and joint collaboration between the three partners, the post-PPC750 roadmap would have followed a different path (whereby IBM and Motorola/Freescale would have worked together, tightly, on this portfolio… which they didn’t). I also don’t think there would have been a PPC74xx and PPC970 as we now know it, but something else entirely not at odds with one another’s design purposes.

1. Apple could have switched to Intel *or* AMD. They chose Intel because of private deals offered by Intel that undercut AMD pricing.

Intel also had the longer track record, and for that, it came out of its extremely predictable journey from the 8088 to the i586/i686. They could, at that time, scale up production more readily and reliably than the younger, scrappier AMD. Jobs wanted predictable for his second act at Apple.

1. The drop in Apple prices because of the switch to cheaper commodity Intel hardware, something Apple heavily promoted and used as additional justification for the switch, was short lived. It was only a few years before Apple prices were back to the previous premiums, further compounded when Apple made the choices to eliminate use upgradeable RAM and storage, a tactic they had used many years before to ensure higher profits.

Sort of, but not quite.

The premiums arose out of the brand cachet they hammered into place after the Jobs-Ive makeover of the products as fashionable to see (and be seen with), post-1998.

Prior to, that quote-unquote fashionable element was privy mostly to people who were intimately aware with Apple and Macintosh already. During those years, one could risk having others roll their eyes at you for walking into a meeting with, say, a PowerBook and not a Pentium-based laptop. The same went for those handling the first Palm Pilot, versus those holding a Newton Messenger, ca. 1996–97.

Few offices beyond advertising, marketing communications, and sound studios had all-Mac or hybridzed workspaces. The presenting of Apple Macintosh hardware was a statement of commitment for being an outlier. It’ß why the “Think different.” campaign resonated with existing Apple product users.

As brand cachet improved, post-bondi blue iMac; as the industrial design began to take on a look not seen elsewhere (the Titanium PowerBook G4 in January 2001 comes to mind here); and as the quality of product build began to improve steadily in a long arc between 1998 and about 2011, Apple could (and sort of had to) raise prices.

For a time, Apple were also packing in more in-built hardware features than could be found on other products at time of introduction (integral AirPort, fibre-optic-backlit keys, hall-effect sensors, hybrid S/PDIF-analogue input-output ports, and so on). They were also, at long last, building enterprise-level hardware support and were beginning to tout the research facility benefits of their hardware. The cost of water jet and laser-cutting a solid billet of aluminium, come full-hilt by the unibody MB/Ps of 2008 (but no doubt probably saw first use with the Xserve’s front bezel), was another such example of up-front manufacturing cost, the trade-off being product rigidity and robustness unparalleled for its time.

What Apple did after 2011 with their hardwre and products is also a story of an Apple without Jobs.

Cook and his team recognized, like the MBA (Master of Business Administration, not MacBook Air) and industrial engineer he was, how one could elevate those identifiable, perceptually robust elements of the product hardware to a premium/luxury item, but no longer had to sell the product within on the extensibility which had been a bedrock of computing hardware from the start. This is also where Apple could begin to cut corners without lowering pricing (touting as “thinness”, an idea first popularized by Jobs, but not in the same spirit of “thin and feature-rich”, where “feature-rich” allowed for parts replacement and repairability/serviceability).

Charging as much or more for less material was a first step toward the Cook era, as was the rolling out of an operating system without bothering to meet principal milestones of completeness/stability of OS X, much less earn revenue on end-user licensing for that OS, after elevating Craig Federighi to oversee Mac OS X development from 2011.

This upscaling of an Apple product as a personal luxury item revealed itself first by the time of iPhone 4’s industrial design in 2010, and then brought over in mid-2012 with the retina MBPs which, to a casual eye not familiar intimately with Apple’s product lineage, might have thought the rMBP was an evolutionary step up in the long road of “more features, continued interoperability/interchageability, and with less material.” But it wasn‘t. It was where Apple began to take lessons learnt from the first couple of MBAs which made them possible as entry-level products (like soldering of RAM to save on design/procurement cost) and to “back-grade” higher-end products with those lower-end, cost-trimming workarounds.

The rMBP was the realization of making more money on a Mac unit by offering even less, whilst still making the marketing persuasiveness that nothing was being pulled back (which, as marketing is about bending clarity and integrity in the service of keeping consumers at attention), despite the fact that several aspects were being back-graded without a respective drop in pricing. Once components within Apple products could no longer be back-graded, Apple began reaching into new revenue streams, several of which having nothing to do with hardware or operating system software.

And since about 2016, Apple have optimized that back-grading to about as far as they can across their entire hardware product line. Some of that back-grading was such a rollback on improving feature function (remember butterfly keys and too-thin MacBook Pros?) that the back-grading got set aside for proven scissor-key mechanisms and a current crop of portables whose form factor, superficially, echo the unibody-era MBPs of 2008–2016).

I think Cook was a useful, interim captain after Jobs stepped down. But Apple’s greatest hardware/software leaps forward came from the times when that piloting was led by someone who, frankly, was either disruptive (Jobs) or had a creative background in architecture and a track record for collaboration (Sculley).

I do think Cook is successful at one thing: extracting a maximum of profit and consumer brand devotion with the least to offer in terms of future-proofing that which Apple sell. Re-framing Apple as a luxury marque vis-à-vis a Cartier, Yves Satint-Laurent, or Maserati, but “let do it with tech”, was a Cook speciality. With that shift upward into luxury pricing and cachet comes a little less accessibility for consumers across all income levels. With introduction of Mac mini, MBA, iBook G3, and iPod shuffle, Apple at least attempted to offer something for everyone at a wider range of budgets. With the Performa line of the late Sculley era, Apple at least attempted the same, however ham-fisted it ended up being under Spindler.

This has been a response which turned into an essay. It’s one of my specialities around here.

I would say that looking at the history of Apple CPUs 1984-2005, they didn't back a series or wrong horses.

I mostly concur, save for the way Jobs mishandled the AIM alliance, post-1997.

He had no business picking favourites in that alliance by shutting out the other party in the process. Doing so only escalated strain in the partnership, stunted innovation in that initial, balanced alliance, and ultimately left Apple in the position of having to save face and fall to a Plan B. This frankly, may have always been Jobs’s plan, owing how he knew from 1997 how NeXTstep/OpenStep was always cross-platform-capable by core design of its POSIX bones; how the same could apply with Rhapsody; then with OS X; internally, could continue to be compiled for little-endian MIPS-based Intel (and probably AMD64) architecture; and finally, was ready-made for prime time come the 2005 Developer Transition Kit running on a Pentium 4 CPU.

 

[mode11]

When Jobs came back, that alliance began fraying. Jobs chose favourites (Motorola for PPC7400, IBM for PPC970). This was the death knell for unified chip design and production by Motorola/Freescale and IBM. And yes, I do put this fraying at Jobs’s feet, as Jobs inherited the AIM alliance. As such, he treated it as his ginger-headed step-child.

Great post in general, but I do question this a little. It's reminiscent of Amiga fans who imagine their platform could have beaten the PC, if only Commodore's management had done things differently. The 68K and PPC architectures were fine, the issue always seemed to be in raising their clock speeds. The PC market was much bigger, so Intel had much more cash, and could therefore afford to build advanced new fabs based on smaller process nodes. Which provided the usual triple-win of lower manufacturing costs, lower power consumption and higher clock speeds.

A clever / efficient architecture with great IPC is great, but if the competition can consistently scale their clocks higher, they will always wind up overtaking - and likely for lower cost. This relegated the Mac to niche status for a long time, until they threw in the towel and leveraged the price / performance of x86 like everyone else.

Apple Silicon is interesting, as it leverages an architecture that leads the (massive) smartphone industry. Unsurprisingly, it's an excellent fit for laptops, though likewise appears to have sounded the death knell for expandible desktop Macs.

Re-framing Apple as a luxury marque vis-à-vis a Cartier, Yves Satint-Laurent, or Maserati, but “let do it with tech”, was a Cook speciality.

Or indeed, Hermes. I tend to think of Apple as more BMW than Bugatti though.

 

[B S Magnet]

Great post in general, but I do question this a little. It's reminiscent of Amiga fans who imagine their platform could have beaten the PC, if only Commodore's management had done things differently.

My understanding at the time — and since — is Apple wanted a quick roll-out of the AltiVec version of the SIMD extension, as designed by Motorola, to be able to make that a valuable selling point for Mac products shortly after Jobs returned.

This meant rushing out a 32-bit solution — in effect, a PPC750 design with the SIMD extension appended. This was a solution in Motorola’s immediate interest, insofar as they could sell this CPU design not only to Apple, the primary customer, but also to customers of low-power embedded systems (including, well, gaming consoles). IBM, which were just as interested in an SIMD extension solution, wanted to prioritize its inclusion for a 64-bit chip design (and would, as VMX, by the time of POWER4). This would require a different approach — one with a longer-term horizon of planning and implementation.

This push by Apple, however, to roll out something with SIMD support ASAP, as Apple were the chief PowerPC customer and third leg of the AIM alliance, put strain on the alliance by 1998. To roll out something ASAP meant having to settle on an existing, 32-bit product, such as the jointly-designed PPC750. This meant grafting AltiVec onto a PPC750 CPU to create the PPC7400. It was short-term gain with long-term pain, as the appending of SIMD as AltiVec would constrain design ability to ramp up clock speeds as quickly as needed. Over the next five years, this put strain on Apple and their initial leap forward over Pentium iterations, until Intel and AMD marched past and left Apple in the crow-eating situation of ditching PowerPC for Intel by 2005.

Consequently, IBM walked away from the AIM alliance Somerset joint design facility in 1998, whose sole purpose was to serve Apple’s CPU needs. Even so, IBM stepped up to manufacture Motorola’s PPC7400 CPUs for first-gen Power Mac G4s in 1999 and 2000, when Motorola couldn’t keep up with production demand (probably under Apple’s insistence, not out of benevolence by Big Blue). It wouldn’t be until a few years later before the AltiVec/VMX extensions were (re)united under the Power ISA 2.03 standard. [On a personal note, I never inspected the November 1999 PPC7400 in the Power Macintosh G4/350 I bought then, because at the time I was unaware there were 7400s from both Motorola and IBM.]

So yes, it does appear the controlling/persuading partner of the AIM alliance, in how and when the SIMD extension was integrated into PowerPC, was that of Apple, which refused to wait for a joint solution to emerge from Somerset. IBM, being the biggest of the trio, could afford to cut losses and move on more easily than either Motorola or Apple. At least, given literature and reports available back then and what can still be sussed now from articles from back then, Apple’s role (under Jobs) is what tends to come through when reading carefully between those lines.

The 68K and PPC architectures were fine, the issue always seemed to be in raising their clock speeds. The PC market was much bigger, so Intel had much more cash, and could therefore afford to build advanced new fabs based on smaller process nodes. Which provided the usual triple-win of lower manufacturing costs, lower power consumption and higher clock speeds.

This is generally true. The counterweight to that capability was its root architecture was descended directly from 8088s from 25 years earlier. It necessitated a direct competitor of Intel, AMD, to shake that up with x86_64.

A clever / efficient architecture with great IPC is great, but if the competition can consistently scale their clocks higher, they will always wind up overtaking - and likely for lower cost. This relegated the Mac to niche status for a long time, until they threw in the towel and leveraged the price / performance of x86 like everyone else.

Again, had Apple — or rather, Jobs — been a more collaborative, active partner in the AIM alliance, ca. 1997–98, the stress fractures leading to IBM’s relinquishing of the Somerset facility would have been, if not avoided, then offset by several more years. Apple were, of course, in the throes of a course correction after being just metres from hitting the hard floor of no return in early 1997, and Jobs was thinking of near-term, impactful selling points — less so the long game — to bring to his keynotes and the company’s marketing communications arm.

Alliances, of course, mean all partners must have the discipline to check their egos at the door and to negotiate solutions. For a short while — at least from 1992 to 1998, AIM did just this.

Apple Silicon is interesting, as it leverages an architecture that leads the (massive) smartphone industry. Unsurprisingly, it's an excellent fit for laptops, though likewise appears to have sounded the death knell for expandible desktop Macs.

It didn’t have to be that way, and it still doesn’t have to be that way.

What I see in Silicon is the fruit of Apple spending years of buying out start-ups for a slew of novel, competitive technologies they could then dump into their own, ARM-based designs — namely and chiefly, the Ax line used for iPhones and iPads.

The pre-emptive absorption of start-ups which would have prompted improved competition, in effect, quashes future competition. That’s what we, the end users, lose. In the end, some core users now must look elsewhere, as customers who would, for example, be the Mac Pro’s base market, find the limitations of SoC (within what was an inherently modular hardware platform) to look to other manufacturers for future replacements.

Although it may work for many, at least for now, the more Apple make Silicon and macOS ever-more proprietary, closed, and dependent on their increasingly closed ecosystem, may be what pushes them into a corner as their remaining competitors manage to do similarly whilst not being locked into a vertically closed ecosystem.

Or indeed, Hermes. I tend to think of Apple as more BMW than Bugatti though.

This is really the splitting of hairs and meaningful only to, idk, subscribers of AdWeek and Communication Arts (of which I was once one).

 

[rampancy]

This meant rushing out a 32-bit solution — in effect, a PPC750 design with the SIMD extension appended. This was a solution in Motorola’s immediate interest, insofar as they could sell this CPU design not only to Apple, the primary customer, but also to customers of low-power embedded systems (including, well, gaming consoles). IBM, which were just as interested in an SIMD extension solution, wanted to prioritize its inclusion for a 64-bit chip design (and would, as VMX, by the time of POWER4). This would require a different approach — one with a longer-term horizon of planning and implementation.

For folks who want to know more about the fine-detailed differences between the PPC 750 and the PPC 7400, Ars Technica still has Jon "Hannibal" Stokes' excellent article on the evolution of Apple's use of the PowerPC ISA and CPUs from 2004.

For me, no, I hadn't felt hoodwinked. Murmurs and reports of Marklar (Apple's "secret" original x86 OS X project) had been swirling for years on the rumor sites, and the precedent set by Star Trek (which another poster mentioned) was widely known among Apple afficionadoes. It really couldn't have come sooner. I've seen so many Great White Hope projects that were supposed to have "killed Intel" and made the PowerPC reign supreme on the desktop come and go, or be nothing more than vapourous rumors (the x704, the 750VX/VXe, the 7457-RM, the MPC 7500, the PA6T-1682M).

Apple needed to keep pace with the market, and they had a viable software platform to make the steps they needed to take. Plus, there were all of the important lessons they learned from the 68K-PPC and System 7-OS X transitions.

 

[B S Magnet]

For folks who want to know more about the fine-detailed differences between the PPC 750 and the PPC 7400, Ars Technica still has Jon "Hannibal" Stokes' excellent article on the evolution of Apple's use of the PowerPC ISA and CPUs from 2004.

For me, no, I hadn't felt hoodwinked. Murmurs and reports of Marklar (Apple's "secret" original x86 OS X project) had been swirling for years on the rumor sites, and the precedent set by Star Trek (which another poster mentioned) was widely known among Apple afficionadoes. It really couldn't have come sooner. I've so many Great White Hope projects that were supposed to have "killed Intel" and made the PowerPC reign supreme on the desktop come and go, or be nothing more than vapourous rumors (the x704, the 750VX, the 7457-RM, the MPC 7500, the PA6T-1682M).

To be fair, that last one was Apple’s doing for buying out PA Semi, arguably hostilely, before PA Semi could do much more with their line of planned, low-power multi-core CPUs. But at least the PA6T-1682M (7W at 2GHz for dual core CPU was nothing to sneeze at!) ended up existing in products.

Still, were there a way to compile a build of OS X, whether Leopard or SL-PPC, to boot and function on the PA6T, that would be a sight to see for the sake of, “Wow, it can be done.”

 

[rampancy]

To be fair, that last one was Apple’s doing for buying out PA Semi, arguably hostilely, before PA Semi could do much more with their line of planned, low-power multi-core CPUs. But at least the PA6T-1682M (7W at 2GHz for dual core CPU was nothing to sneeze at!) ended up existing in products.

I was deeply skeptical of PA Semi, just like how I was skeptical of Exponential Technology and the x704. As a newly-minted startup, I felt like it was very unlikely that they'd have the chops to deliver competitive PPC CPUs at a timely pace for Apple. There was speculation that Texas Instruments would be the fab to produce these; would they also suffer from the same production issues that plagued IBM and Motorola in the past? Maybe, maybe not. But at the time I would have had a lot more confidence had, say, Apple just bought out PA Semi (which they did anyway) and gotten TSMC, Global Foundries or IBM to make the chips for them.

The PA6T-1682M wound up in the AmigaOne X1000; between that and all of the custom silicon it had, I daresay that has to be one of the most unique personal computers that I'd ever learned of.

Still, were there a way to compile a build of OS X, whether Leopard or SL-PPC, to boot and function on the PA6T, that would be a sight to see for the sake of, “Wow, it can be done.”

I'm almost certain there has to be a hardware prototype somewhere with a custom build of 10.5 or 10.6. It definitely would be cool to see it in the Computer History Museum alongside the pink-striped Exponential prototype 9500.

 

[Amethyst1]

I'm almost certain there has to be a hardware prototype somewhere with a custom build of 10.5 or 10.6.

Sorry for going off-topic, but I do wonder when full OS X began running on ARM. 2010, on the A4, later — or earlier?

 

[rampancy]

Sorry for going off-topic, but I do wonder when full OS X began running on ARM. 2010, on the A4, later — or earlier?

The first serious rumours surfaced in 2018, and the benchmarks of what had to have been M1 prototypes showed up in Geekbench in 2019. But there were rumours going back as 2012 (enough for Ars Technica's John Siracusa to talk about it on an episode of Hypercritical that still holds up today).

So I'd speculate that yeah, 2010 sounds about right. In fact I'd wager that the work to get macOS on ARM CPUs may stretch back to the beginnings of iOS, since Jobs made such a big deal of the first iPhone "running OS X".

 

[Amethyst1]

In fact I'd wager that the work to get macOS on ARM CPUs may stretch back to the beginnings of iOS, since Jobs made such a big deal of the first iPhone "running OS X".

Hm. I interpreted that as a placeholder for what would eventually be called i[Phone]OS. The first version identifies as Darwin 9.0 (uname -sr).

 

[akator]

The premiums arose out of the brand cachet they hammered into place after the Jobs-Ive makeover of the products as fashionable to see (and be seen with), post-1998.

I'm onboard with everything except this. The Apple premiums were well established in the 80s.

My Apple IIc was more expensive than contemporary alternatives, as was my Mac Plus, and Mac IIsi — and all of those were purchased on educational discount. The Apple "memory premium" was insane when I bought my IIsi. Apple wanted just under another $1000 to increase the RAM, the same memory available for under $400 when purchased elsewhere. I obviously did that instead.

Fortunately, after that I worked at universities, purchased their Macs, and used their Macs so I didn't have to buy one again out of pocket for years. Evem so, it was a constant battle to justify because commodity PC hardware could be purchased for significantly less.

When the first PowerPC boxes were released the battles got easier because the prices were closer to PCs. When I filled a computer lab with 6100's it was far less trouble than I had experienced previously because of the price.

The G3 iMac was the first Mac that I didn't have to go to battle over because the price was close enough to a similarly speced PC. After that I moved to the private sector and didn't have to worry about all of that anymore, but I did have to start buying my own Macs and PCs again.

I loved that temporary price drop with early Intel Macs. I didn't love the massive decrease in hardware quality and reliability (which I already mentioned), as I still owned much older Macs that hadn't even once had any hardware issues.

 

[Clix Pix]

Coming in really late to this fascinating and educational thread! You guys have filled in some gaps in my knowledge about the reasons behind the shift to Intel from PPC and why that was so important. Thanks!

I bought my first Mac (G5 iMac Rev B) right after the keynote in which Steve introduced the first iMac with a built-in iSight camera. I had already been leaning in the direction of buying a Mac and getting away from Windows but was still hovering on the brink, so to speak. Steve introduced the new iSight iMac and the thought flashed through my mind, "I don't care about a camera, I don't need that, I like the other iMac!" BOOM! So, yes, the very next day I was in the Apple Store buying my very first iMac (G5, rev B and, yay, on sale now, too!) and being nervous as heck about this. Was I going to the real Dark Side or was Windows truly the Dark Side and I was finally coming into the Light?

Didn't take me long to decide that I'd finally come into the Light. I fell in love with my new iMac in very short time. So much was intuitive and so much easier than wrestling with Windows! I was amazed. Unsurprisingly, not long after that I bought myself a companion Mac: a G4 PowerBook because I was getting ready to travel and knew that if I didn't want to use Microsoft Windows at home I sure wasn't going to want to do so away from home, either. So both my old Windows desktop and laptop were stashed, huddled together in a spare room, now unloved and unused as I enjoyed and took delight in my two new Macs. Eventually I wiped both of those Windows machines and donated them to a charity organization.

That trip I mentioned? Uh....actually, yeah, it was in January 2006 and the destination? San Francisco.....MacWorld. !! That is a trip I'll always remember. Anyway, so I'm sitting there at the keynote presentation and the guy in the bunny suit comes out and then Steve is announcing the first Intel Macs. Whoa, WOW!!!

I wasn't all that bothered that hey, quite recently I'd just spent a fair amount of money on two Macs which were both on the PPC platform and now suddenly within just two or three months the PPC machines were being pushed aside to make room for the brand-new Intel platform-based machines. ??!!! I loved my two PPC machines and was happy with them. Only later did it belatedly occur to me how angry some people probably were, though, if they'd bought a new iSight iMac or a new G4 PowerBook, expecting that this was the latest for a while yet to come, and -- BAM!

That really was a stunning and (from a customer's viewpoint) abrupt move on Apple's part and while the replacement of the G4 Powerbook with the first-gen MBP might not have been so much of a slap in the face to customers and long-time Mac users, I'm sure that the announcement of the Intel-based iMac just a couple of months after the big fuss Steve had made over the first G5 iMac with iSight camera was something which really did sting. I have no idea of how many people had eagerly jumped to buy that G5 iMac with iSight, which did seem new and revolutionary, only to see something even more new and revolutionary arrive on the Apple / Mac computer scene just a short time later, but can guess that many were very unhappy. Ouch!

As for me? I continued enjoying my two new PPC machines for a while, but of course was also curious about the whole Intel platform version of Macs as well. An opportunity at work came up for me to sell my Powerbook to a colleague and I happily wasted no time in doing just that so that I could justify buying myself an MBP when my Powerbook was still relatively new. I kept my beloved G5 iMac rev B for a while longer -- several years -- and continued to appreciate what it could do for me, too.

Interesting times, 2005 and 2006.....

 

[rampancy]

Interesting times, 2005 and 2006.....

It really was, wasn't it? On the hardware side, the original Core Duo-based MacBook Pro was a monster of a machine, and that was compared to the mighty DLSD/Hi-Res PowerBook G4, which was itself an excellent laptop.

I also had a thought that I didn't touch on earlier posts, but have articulated elsewhere. PPC never really "died"; it's just that IBM and Motorola/Freescale's goals for making the PPC a profitable platform stopped aligning with Apple's goal to make and sell products with an industry-competitive general computing CPU. The PowerPC has been used and continues to be used in stuff ranging from car microcontrollers, to networking routers, to space probes to fighter jets.

 

[B S Magnet]

I'm onboard with everything except this. The Apple premiums were well established in the 80s.

They were, yes. What they weren’t, for the most part, were fashionable-as-luxury pieces.

They were prized in higher education and in the visual design industry because they could do, at scale, what no other product could. That was their chief premium value proposition. Even as major universities purchased a few NeXTstations and NeXTcubes for advanced research in the hard sciences, the same institutions also had entire computer labs and libraries chock-full of nothing but Macs.

To wit, my first physical contact with, quote-unquote, “the internet”, happened on a Mac workstation inside the principal library of a major university in 1993 (it may have been as late as February 1994, though). Without question, that university owned a dozen or more NeXTstations peppered across their physics, chemistry, compsci, and astronomy departments (and possibly more in their media & technology school).

The fashionable-as-luxury-piece component, however, came into force after Cook took over the ship.

Prior to, Jobs wasn’t aligning Apple products with blue-blood/nouveau riche brands one finds typically (and only) on North Michigan Avenue, Rodeo Drive, the Mink Mile, and Bond Street. In urbanist-speak, these are hyper-gentrified areas. Further demonstration of this shift toward luxury-first has come with where Apple are siting post-Jobs Apple retail store locations: less in shopping malls and upscale shopping centres peppered across a major city and more in those aforementioned districts whose names need no explanation to those for whom money is their plaything. They are the kinds of districts wherein even for the biggest cities of the world, it is rare for there to be more than one of them for any given city.

As with the Apple products (such as those adopting the Frog Design language) produced before Jobs’s return, the Jobs-Ive products were still very much in the realm of “this is a premium computing product, because this also gets done the work for which it excels.” They were premium for what they could do in terms of visual-oriented productivity.

Cook’s chief contribution has been to render the Apple marque as that luxury piece. His emphasis toward that end has made the way Apple’s products are designed and built to take second chair. The most extreme attempt to that end has given us butterfly mechanisms, staingate, five-figure Apple Watches, and “gold” aluminium.

This is no longer “premium” territory. This is wealth trophy-level, high-fashion, posh stuff. For many places, the conspicuous display of an iPhone now is the flashing of, say, a Dolce & Gabbana pair of eyeglass frames or Louis Vuitton wallet from as recently as fifteen years ago: from one end, that iPhone is entry-level stuff for which all the higher-end stuff is accessible to you (because, bluntly, you’re doing very well or have inherited such), or else that current iPhone is, practically, the closest one is likely to touch in their life to that taste of wealth.

And yet, there’s a sense Cook wants it both ways: as top-shelf luxury (i.e., very durable goods), yet in making the equipment’s components impossible to replace due to soldering consumables and extreme cryptographic locking, as fast-fashion (i.e., disposable, sub-durable goods). Held in this light, it begins to make some sense why Cook’s oversight has moved Apple hardware products into becoming cryptographic safes with no possibility for parts interchangeability. In cryptography’s absence, Apple products would be stolen at orders far greater than than they were just fifteen years ago when my own aluminium MBP was stolen or, say, a couple of years earlier when I watched someone literally grab an iPod someone was using, boldly wrest it from them, and flee through a subway door, escaping without being stopped.

For more reasons than is topical to this discussion, I hate this new normal.

My Apple IIc was more expensive than contemporary alternatives, as was my Mac Plus, and Mac IIsi — and all of those were purchased on educational discount. The Apple "memory premium" was insane when I bought my IIsi. Apple wanted just under another $1000 to increase the RAM, the same memory available for under $400 when purchased elsewhere. I obviously did that instead.

Even so, the Apple stuff you were buying then was not anywhere on the radar of people who were in the market for Bentley automobiles and bespoke fur coats from, idk, Christian Dior. (And hey, at least you could buy third-party RAM!)

Apple stuff then were in the realm of nerdy people who made stuff, recognized their utility and, thus, their use-value. Artists who were fortunate enough to have patrons, for example, fit that bill as early as 1986 when the Macintosh Plus went on sale. But the people well-to-do enough to buy that art? Apple, as a brand, wasn’t even a blip on their radar. If it was, then it was little more than regarding an all-in-one Macintosh, like the Plus or SE/30, as a curio.

Fortunately, after that I worked at universities, purchased their Macs, and used their Macs so I didn't have to buy one again out of pocket for years. Evem so, it was a constant battle to justify because commodity PC hardware could be purchased for significantly less.

Yup.

When the first PowerPC boxes were released the battles got easier because the prices were closer to PCs. When I filled a computer lab with 6100's it was far less trouble than I had experienced previously because of the price.

Yup. By then, Spindler recognized how outfitting PowerPC Macs with commodity parts and ports like PCI and IDE (and not, say, Nubus and SCSI) were effective ways to cut costs per unit and also to improve their use-value for folks who entered the modern computing realm on a steady diet of IBM PC clones. This was also when Apple outsourced essential peripherals (like displays) to the companies which could deliver what they needed (such as Sony’s Trinitron displays in Apple displays and the colour laser printing engine already in use by Canon’s colour laser copiers).

Even so, those devices used their own, industry-idiosyncratic serial bus and display ports. USB, in that sense, was also a price-cutting game-changer, and Jobs recognized that when making sure the first iMac shipped with it. So too was the eventual adoption of DVI-oriented ports.

The G3 iMac was the first Mac that I didn't have to go to battle over because the price was close enough to a similarly speced PC. After that I moved to the private sector and didn't have to worry about all of that anymore, but I did have to start buying my own Macs and PCs again.

…

I loved that temporary price drop with early Intel Macs. I didn't love the massive decrease in hardware quality and reliability (which I already mentioned), as I still owned much older Macs that hadn't even once had any hardware issues.

The early–early Intel Macs were rushed out the door, and there were teething issues. They were products for which the enclosures within were designed for a different architecture with different thermal demands. Apple managed to get the balance right from around 2007 (with the first, unibody-styled, purpose-designed-for-Intel iMacs), then carrying that idea over to the unibody laptops of 2008, then over to the Mac minis in 2010.

For that brief moment, 2007 to 2012, Apple, still under Jobs’s watch (though with Cook waiting in the wings and chomping at the bit), they got the balance right.

This balance would be relatively short-lived.

 

[dmr727]

By then, Spindler recognized how outfitting PowerPC Macs with commodity parts and ports like PCI and IDE (and not, say, Nubus and SCSI) were effective ways to cut costs per unit and also to improve their use-value for folks who entered the modern computing realm on a steady diet of IBM PC clones. This was also when Apple outsourced essential peripherals (like displays) to the companies which could deliver what they needed (such as Sony’s Trinitron displays in Apple displays and the colour laser printing engine already in use by Canon’s colour laser copiers).

Your display and printer examples predate Spindler. Apple's earliest (external) Macintosh displays were already using Trinitron, and the original LaserWriter used the same Canon engine as HP's LaserJet. Doesn't change your point about PCI and IDE, but I was a huge fan of Sony's Trinitron back in those days, and was really pissed at my Dad for pairing our shiny new IIcx with some off-brand inferior display.

 

[B S Magnet]

Your display and printer examples predate Spindler. Apple's earliest (external) Macintosh displays were already using Trinitron, and the original LaserWriter used the same Canon engine as HP's LaserJet. Doesn't change your point about PCI and IDE, but I was a huge fan of Sony's Trinitron back in those days, and was really pissed at my Dad for pairing our shiny new IIcx with some off-brand inferior display.

Spindler took over in 1993.

The ColorLaserWriter 12/600 PS arrived in 1995. The lion’s share of LaserWriter models and pretty much all of the advanced LaserWriter models (like those capable of printing to ledger — to 11x17 — format, in colour, and/or the multi-tray models like the LaserWriter Pro 810) went on sale either during or in the years after 1993. Otherwise, yah, every LaserWriter-branded printer made, going back to 1985, used either the Canon laser engine or the Fuji/Xerox engine. All but two of the StyleWriter inkjet models showed up after Spindler took over. A majority of the OneScanner flatbed scanners were sold during Spindler’s years. And so on.

Although there was the Trinitron-based 1987 13-inch Apple High-Resolution Color Display, it was the expansive variety of CRTs, many of them Trinitron-based, with Apple branding to arrive after 1993, and these would carry the company to the LCD era (Apple AudioVision 14, some of the Apple Multiple Scan models, the ColorSync displays, and the Studio Display CRTs).

 

[dmr727]

Spindler took over in 1993.

The ColorLaserWriter 12/600 arrived in 1995. The lion’s share of LaserWriter models and pretty much all of the advanced LaserWriter models (like those capable of printing to ledger — to 11x17 — format, in colour, and/or the multi-tray models like the LaserWriter Pro 810) went on sale either during or in the years after 1993. Otherwise, yah, every LaserWriter-branded printer made, going back to 1985, used either the Canon laser engine or the Fuji/Xerox engine. All but two of the StyleWriter inkjet models showed up after Spindler took over. A majority of the OneScanner flatbed scanners were sold during Spindler’s years. And so on.

Although there was the Trinitron-based 1987 13-inch Apple High-Resolution Color Display, it was the expansive variety of CRTs, many of them Trinitron-based, with Apple branding to arrive after 1993, and these would carry the company to the LCD era (Apple AudioVision 14, some of the Apple Multiple Scan models, the ColorSync displays, and the Studio Display CRTs).

Um...right. I'm aware of all of that. Guess I don't understand your point - what did Apple's expanding of their monitor and printer lineup have to do with Trinitron and 3rd party printer engines, as both were being used far earlier?

 

[B S Magnet]

Um...right. I'm aware of all of that. Guess I don't understand your point - what did Apple's expanding of their monitor and printer lineup have to do with Trinitron and 3rd party printer engines, as both were being used far earlier?

The point was Spindler took considerable stock in significantly expanding those merchandising options for the end user.

Prior to, those options were supplemental at most and — often — special, collaborative efforts between Apple and the other party.

Spindler took that, genericized some of it (such as maintaining external enclosures for those products across new product revisions), and mostly got out of the way as product internals got updated whenever the host company’s tech within was updated (helping to explain some of the arcane model numbers for peripherals, such as the Apple Color OneScanner 600/27, the Apple Color StyleWriter 6500, the LaserWriter 12/640 PS, and so on).

This, alongside the adoption of standardized ports and buses, were all cost-cutting measures which — to the degree they worked and were survived by Jobs-era Macs which carried forth those measures — also diluted the brand and risked it to the point of near-extirpation by 1997.

 

[dmr727]

The point was Spindler took considerable stock in significantly expanding those merchandising options for the end user.

Prior to, those options were supplemental at most and — often — special, collaborative efforts between Apple and the other party.

Ah, I see what you're saying. I tend to agree that these efforts diluted the brand - I was not a huge fan of that era.

 

[TheShortTimer]

I agree with your entire reply and I've expressed similar views in other posts/threads. I've selected these segments because I wanted to focus on a couple of specific points.

The early–early Intel Macs were rushed out the door, and there were teething issues.

64 bit CPU's with 32 EFI's. Downgraded capabilities in some respects compared to preceding PPC models. I'll return to this topic later...

For that brief moment, 2007 to 2012, Apple, still under Jobs’s watch (though with Cook waiting in the wings and chomping at the bit), they got the balance right.

This balance would be relatively short-lived.

The peak Intel years, in my view and yes - the decline was sharp. Glued batteries, permanently soldered RAM and SSDs etc.

On the hardware side, the original Core Duo-based MacBook Pro was a monster of a machine, and that was compared to the mighty DLSD/Hi-Res PowerBook G4, which was itself an excellent laptop.

Disagree on the 2006 MBP. My model has one FireWire 400 port whereas the PB G4 has 800 and 400 ports. The optical drive is a single-layer writer whereas the PowerBook features a dual layer one. My brother and I were stumped as to why his MBP failed to burn dual-layer DVDs and on a hunch, I checked the drive's specifications and discovered the reason. This was very disappointing and a strange decision. When I upgraded the RAM and HDD on mine, replacing the optical unit was a top priority.

I also had a thought that I didn't touch on earlier posts, but have articulated elsewhere. PPC never really "died"; it's just that IBM and Motorola/Freescale's goals for making the PPC a profitable platform stopped aligning with Apple's goal to make and sell products with an industry-competitive general computing CPU. The PowerPC has been used and continues to be used in stuff ranging from car microcontrollers, to networking routers, to space probes to fighter jets.

Absolutely. I have five* video games devices that use PowerPC CPUs and yes, the machines are no longer current platforms but it's still indicative of just how prolific their presence has been - and as you've pointed out, continues to be - right across the board despite Apple and IBM parting ways.

Very interesting to learn that the F-35 uses a G4 set-up and that the F-22 and Rafale-18 have G3's.

*The Nintendo GameCube, Wii and Wii-U all use enhanced G3's. As I've already mentioned, Microsoft enlisted IBM to design a PPC CPU for the Xbox 360 and finally, the IBM-Toshiba-Sony developed 64-bit PPC Cell processor powers the PlayStation 3.

 

[radellaf]

I'm not sure why, but this has been an extremely interesting thread. I saw it in a "MacRumors Forums - See what you have missed" email, and had to do a double-take seeing a 2024 post about PPC->Intel. Stuck with it and enjoyed the discussion.
I was PC for the 386 & 486 (6809 before). I used 68x and PPC macs at my first job, and the way the transition was handled seemed masterful.

They were too expensive for home, until Steve and my beloved bondi iMac, followed by the iMac DV, G3 macbook, G4 eMac (what a CRT!), and the 2008 black macbook (Penryn? 32 bit x86) which was a dismal disappointment in terms of software support. Dumped within 4 years, yet the hardware STILL works fine. Given how well the PPC transition was handled, I was hoping for much better for the software-support-life of my first Intel. The G4 eMac also "died" from slow software decay, with the hardware being absolutely fine past the end.

Whereas, a 2011 MBP literally desoldered itself and died in the middle of a game, in addition to having some sort of badly designed (intentionally?) IDE bus handling such that non-Apple 2TB HDDs would park-load themselves to death unless I booted into Linux off USB first in order to use some utility to set a volatile parameter.

I guess I never had an upgradeable, PC-like, Mac. The blue G3 desktop on the job was a work of art. My own machines gave me OS9 and X instead of the then-crappy Windows, but were nearly impossible to upgrade across the board. And I can't call 2011 part of the Intel glory years, given the thermal disaster of the 2011 MBP. (The 2012 models had USB3 and seemed to last). I guess I bought the two worst-fated Apple laptops, ever. And got very used to a stack of external HDDs.

Then I bought the worst-timed iMac ever, the late 2015 in 2017. An old design updated months later, but with a dead MBP... what could I do. Still, it ran for a long time, still works, but wow am I happy with the pre-ordered maxed out M3 iMac (no pro or max CPU option). The aesthetics of both iMacs are compromised by "having" to put air filters on the bottom grate, given you have FANS but no ability to open the case to blow out the DUST. Anyway, another CPU transition well done, apart from losing QuickTime 7 Pro for video editing. I do wonder what the future is, though. Good -Apple has full control and can optimize everything including GPU/CPU. Bad - Apple can't benefit from anything that happens in the PC space, and seems to have a desire, lurking under the surface, to turn the Mac into an iPad.

Mouse Pointers. I had a strong opinion for OSX 1 and 2 that it was horrible.? I had bad carpal tunnel, and compared to OS 9, it really hurt my wrists for reasons I couldn't fathom. But, now? MacOS vs Win7,10,11? Somehow, Windows just seems ever so slightly touchy, and the Mac is just smoooth. Also, having a magic trackpad on the left and a logitech mouse on the right is just the best of everything. Maybe Win11 would be better if the Logitech mouse driver was useable (it doesn't play well with NVIDIA Optimus). IOW, MacOS mouse pointer and scrolling? Superior.

 

[B S Magnet]

Whereas, a 2011 MBP literally desoldered itself and died in the middle of a game, in addition to having some sort of badly designed (intentionally?) IDE bus handling such that non-Apple 2TB HDDs would park-load themselves to death unless I booted into Linux off USB first in order to use some utility to set a volatile parameter.

RaaaaaaadeonGate

The SATA HDD park-loading bit is one I’m not familiar with, as one who’s owned and used, non-stop, two 13-inch 2011 MBPs (whose SATA bus logic is no different from the 15- and 17-inch variants). Then again, one of the very first things I did with my early 2011 13-inch MBP in August that year, was to pull out the SuperDrive, put in a HDD caddy, install a 1TB HDD there, and install a much smaller SSD in the usual HDD location as a boot volume.

I guess I never had an upgradeable, PC-like, Mac. The blue G3 desktop on the job was a work of art. My own machines gave me OS9 and X instead of the then-crappy Windows, but were nearly impossible to upgrade across the board. And I can't call 2011 part of the Intel glory years, given the thermal disaster of the 2011 MBP. (The 2012 models had USB3 and seemed to last). I guess I bought the two worst-fated Apple laptops, ever. And got very used to a stack of external HDDs.

The 2008 BlackBook4,1 would have been legendary had Apple bothered to set up those and the WhiteBook4,1 with 64-bit EFI and a GPU better than the GMA X3100. That would have to wait until 2009, with the return of the WhiteBook5,2 (but, alas, no BlackBook counterpart). There have been a couple of instances where BlackBook enthusiasts used the case of a 4,1 to stuff the WhiteBook5,2 (2009-era) contents into it, which enabled them to use a BlackBook with OS builds beyond a (patched) Mountain Lion. (It does require building up spots on the internals for where anchoring grommets were moved between 4,1 and 5,2, so it’s not a simple, one-for-one transplant.)

The 15- and 17-inch MBPs of 2011, meanwhile, were the giant sore in an otherwise good period of products. As a 2011 13-inch MBP user to this day (including making this reply!), GPU notwithstanding, the unibody models in those sizes were sturdy, solid, and all other components were outstanding. Parts availability was quite good. The unibody MacBooks of 2010 were also so much better designed and built than the predecessor, taking its design and assembly cues from the aluminium unibody.

Nevertheless, my dream Mac was the late 2011 17-inch quad-core i7 with an anti-glare display. Once RadeonGate revealed that how the failing GPU problem would never be fixed by AMD (in the way the GeForce 8600M GT of the 2007–08 MBPs were by Nvidia), they became absolute albatrosses unless one made the firmware (or hardware) tweak to get the system to use the iGPU — the Intel HD 3000 Graphics — exclusively. Sure, there is the mid-2012 15-inch MBP (and it is quite nice!), but Apple chose not to produce the 17-inch for that series. They were also Ivy Bridge models, which don’t natively run Snow Leopard (a big deal-breaker for me and my work).

Then I bought the worst-timed iMac ever, the late 2015 in 2017. An old design updated months later, but with a dead MBP... what could I do. Still, it ran for a long time, still works, but wow am I happy with the pre-ordered maxed out M3 iMac (no pro or max CPU option). The aesthetics of both iMacs are compromised by "having" to put air filters on the bottom grate, given you have FANS but no ability to open the case to blow out the DUST. Anyway, another CPU transition well done, apart from losing QuickTime 7 Pro for video editing.

At least with your 2015 iMac, you can still have access to QuickTime 7 Pro if you have a partition on there which can boot into Mojave. But otherwise, yah.

I do wonder what the future is, though. Good -Apple has full control and can optimize everything including GPU/CPU. Bad - Apple can't benefit from anything that happens in the PC space, and seems to have a desire, lurking under the surface, to turn the Mac into an iPad.

Apple, if they’re not careful, risk locking themselves into a nook of a niche and risk becoming an evolutionary dead-end. Which would, historically, not be the first time Apple almost ended up there without making a necessary course-correction.

 

[TheShortTimer]

RaaaaaaadeonGate

The SATA HDD park-loading bit is one I’m not familiar with, as one who’s owned and used, non-stop, two 13-inch 2011 MBPs (whose SATA bus logic is no different from the 15- and 17-inch variants).

I'm not familiar with this either and my 13 inch 2011 operates as a daily driver and in the years that I've owned it, the computer has remained constantly switched on.

The 2008 BlackBook4,1 would have been legendary had Apple bothered to set up those and the WhiteBook4,1 with 64-bit EFI and a GPU better than the GMA X3100. That would have to wait until 2009, with the return of the WhiteBook5,2 (but, alas, no BlackBook counterpart). There have been a couple of instances where BlackBook enthusiasts used the case of a 4,1 to stuff the WhiteBook5,2 (2009-era) contents into it, which enabled them to use a BlackBook with OS builds beyond a (patched) Mountain Lion. (It does require building up spots on the internals for where anchoring grommets were moved between 4,1 and 5,2, so it’s not a simple, one-for-one transplant.)

Some time ago I embarked on this project and really need to finish it.

A poster on ThinkClassic provided a pretty instructive guide on accomplishing this modification.

Apple, if they’re not careful, risk locking themselves into a nook of a niche and risk becoming an evolutionary dead-end. Which would, historically, not be the first time Apple almost ended up there without making a necessary course-correction.

This time around, the odds are that there won't be a Jobs like figure to swoop in and lead them out of the rut.

 

[mode11]

This push by Apple, however, to roll out something with SIMD support ASAP, as Apple were the chief PowerPC customer and third leg of the AIM alliance, put strain on the alliance by 1998. To roll out something ASAP meant having to settle on an existing, 32-bit product, such as the jointly-designed PPC750. This meant grafting AltiVec onto a PPC750 CPU to create the PPC7400. It was short-term gain with long-term pain, as the appending of SIMD as AltiVec would constrain design ability to ramp up clock speeds as quickly as needed. Over the next five years, this put strain on Apple and their initial leap forward over Pentium iterations, until Intel and AMD marched past and left Apple in the crow-eating situation of ditching PowerPC for Intel by 2005.

OK, but Apple didn't really have much choice there. If they had held off on the G4 until they had a 64 bit chip, Macs would have fallen even further behind x86 in the early 2000's.

So yes, it does appear the controlling/persuading partner of the AIM alliance, in how and when the SIMD extension was integrated into PowerPC, was that of Apple, which refused to wait for a joint solution to emerge from Somerset. IBM, being the biggest of the trio, could afford to cut losses and move on more easily than either Motorola or Apple. At least, given literature and reports available back then and what can still be sussed now from articles from back then, Apple’s role (under Jobs) is what tends to come through when reading carefully between those lines.

Apple was the principle customer of consumer PPC, so it's natural they dictated that product. IBM's POWER chips were / are a separate product, with no application in Macs, and a big market on their own; I expect IBM shrugged when Apple moved on. Motorola had no chance of keeping up with Intel in terms of process node. All told, this didn't seem like an alliance that had a long term future, even if Jobs had played nice.

This is generally true. The counterweight to that capability was its root architecture was descended directly from 8088s from 25 years earlier. It necessitated a direct competitor of Intel, AMD, to shake that up with x86_64.

Lots of architectures date back a long way; the ARM v1 came out in 1985. In any case, as transistor counts increased, ISAs got increasingly abstracted. Whereas RISC held a significant advantage in the mid-90's, this became less so later. x86_64 added 64-bit instructions, but the P4 was already running out of steam by that point and the Athlon 64's architecture was strong generally. Intel had already been working on Core since 2001, originally for mobile use as a development of the Pentium M, then for everything after they abandoned Netburst.

The pre-emptive absorption of start-ups which would have prompted improved competition, in effect, quashes future competition. That’s what we, the end users, lose. In the end, some core users now must look elsewhere, as customers who would, for example, be the Mac Pro’s base market, find the limitations of SoC (within what was an inherently modular hardware platform) to look to other manufacturers for future replacements.

Good point.

Although it may work for many, at least for now, the more Apple make Silicon and macOS ever-more proprietary, closed, and dependent on their increasingly closed ecosystem, may be what pushes them into a corner as their remaining competitors manage to do similarly whilst not being locked into a vertically closed ecosystem.

The iPhone rules Apple. Apple Silicon is largely based on the A-series and can't stray too far (e.g. PCIe GPUs) whilst still being able to benefit from the investment there.

So I'd speculate that yeah, 2010 sounds about right. In fact I'd wager that the work to get macOS on ARM CPUs may stretch back to the beginnings of iOS, since Jobs made such a big deal of the first iPhone "running OS X".

It was to be expected that Apple base 'iPhoneOS' on OS X rather than create a new OS from scratch. When the iPhone came out, it used a Samsung ARM chip; Apple had needed a low-power processor suitable for a smartphone. At the time, Intel were dominating everyone with the Core series and Apple had no track record of designing an ARM CPU. I doubt Apple thinking of an ARM transition for the Mac at that point. I expect it was only after a) Apple's chip development proved very successful and b) Intel's chip development stalled, that Apple started thinking seriously about the transition.

I have no idea of how many people had eagerly jumped to buy that G5 iMac with iSight, which did seem new and revolutionary, only to see something even more new and revolutionary arrive on the Apple / Mac computer scene just a short time later, but can guess that many were very unhappy. Ouch!

It's not just that a better model came out so soon - there was always something better on the horizon, especially in the early 2000's. It was that after years of promoting the superiority of PPC, the 'megahertz myth' and generally lampooning Intel, it was quite some U-turn to announce a transition to their processors. Especially when the Intel Macs proved to be so much faster.

 

[B S Magnet]

OK, but Apple didn't really have much choice there. If they had held off on the G4 until they had a 64 bit chip, Macs would have fallen even further behind x86 in the early 2000's.

They did, but arguably, they passed on it.

IBM announced the 64-bit POWER3 two weeks before Motorola announced the 32-bit PPC7400. Both were designed around the PowerPC ISA.

Had Apple pursued this, given power consumption ratings, POWER3-based CPUs in Macs might have been designated “G4/64” for Power Mac (and Cube) workstations, whereas lower-power, laptop/AIO-oriented PPC7400-based CPUs would have been designated as “G4/32” for those models. Meaning — a 64-bit Mac, provided there was a ready-made OS to make use of 64-bit instructions — could have been on offer as early as 1999. (I have no idea whether NeXTstep or OpenStep was ever considered or tested as a 64-bit build before Apple’s acquisition of NeXT.)

At this point, this is nothing better than conjecture for a lively conversation.

[An amusing thing is IBM added Motorola’s AltiVec to their PPC750 plans, but ultimately never implemented them, at least in whole, for a finished chip product.]

Apple was the principle customer of consumer PPC, so it's natural they dictated that product. IBM's POWER chips were / are a separate product, with no application in Macs, and a big market on their own; I expect IBM shrugged when Apple moved on. Motorola had no chance of keeping up with Intel in terms of process node. All told, this didn't seem like an alliance that had a long term future, even if Jobs had played nice.

Even so, POWER — POWER4 — was adopted for use in Macs, as the AltiVec-enabled PPC970 for the G5s.

The iPhone rules Apple. Apple Silicon is largely based on the A-series and can't stray too far (e.g. PCIe GPUs) whilst still being able to benefit from the investment there.

Let’s just hope they don’t back themselves into a corner they can’t escape should sales of their Macs begin to go — and stay — relatively flat.

 

[mode11]

Even so, POWER — POWER4 — was adopted for use in Macs, as the AltiVec-enabled PPC970 for the G5s.

That's true. Though its big-iron roots made it completely unsuitable for laptops, and when down-clocked and running as a single CPU for iMacs, it was barely faster than a G4 of similar clock speed ( https://www.reddit.com/r/VintageApple/comments/8vswn1 ). Conversely, the PPC970 was only used by IBM briefly, in their low-end stuff. So the G5 was ultimately a stop-gap, before the inevitable switch to x86.