Become a MacRumors Supporter for $50/year with no ads, ability to filter front page stories, and private forums.
Don’t think of it as a cancellation. Apple is a conservative company. They developed these Mac Pros as a "Backup, Plan B” in case their high-end users could not get by with the Mac Studio. The Plan B was never needed.

Yup. Two more "no's" out of the thousand they do for every "yes".

We know Apple was working on an "Extreme" M1 series SoC and they were having issues with it. So it makes sense they would have a backup plan to update the 2019 Mac Pro with Comet Lake or Rocket Lake W-series Xeons, if necessary.

We know from the 2017 round table they now speak regularly with their "pro" customers so the feedback they were likely getting made offering a final Intel update was not necessary and that the M2 Ultra was "good enough" for those customers who were interested in an Apple Silicon Mac Pro.
 
The interesting "what if" is "what if Apple didn't mess up the 2013 Mac Pro", because that's probably when the exodus started making it even more niche than it already was.

Agreed. As slick-looking as it was, the 2013 Mac Pro started the move off the platform for a lot of those customers. The 2023 Pro was the final nail.
 
They constantly did it to themselves though. The Mac Pro made sense back in the late aughts and early teens when it wasn't stratospherically more expensive than its competition.

There has always been a desire for a more modular Mac that can dual boot Windows and play PC games as well as run the full MacOS experience in a separate environment. But the amount of people willing to pay over $10k just for the base model has never been enough.

And somewhat ironically, the dual boot situation is solved again since Windows on ARM is now fully officially supported. There's nothing stopping them from supporting video cards, Windows, and expandable RAM except their own decisions.

With intel macs you had the option to boot windows or linux, now you can't, Windows on ARM works only virtualized, linux on mac is pretty much limited.
 
  • Like
Reactions: phuklok1
They should’ve kept making the Mac pro with Intel for expandability. Apple silicon is amazing but it can’t compete in some areas, specifically expandability and high performance when power draw isn’t a concern

It would’ve been better for their ecosystem even if it wasn’t a huge seller. But it also could’ve been a big seller for specific needs
 
  • Like
Reactions: phuklok1
a factor in killing off the intel tower may of been the fact they pissed off nvidia by kicking them off hte platform years before. then nV said the same back when apple wanted updated drivers to run AI on nV GPUs. That was more than likely their "other use". a market where price didn't matter so much and who could afford 1TB+ of RAM @ 2022 prices.

then the power factor. a machine that could easily hit 2kw would max out the power available on a 20A circuit ! many homes only have 15A breakers & wiring, especially older ones. just not viable.

Not doing the extereme M3 is a big disappointment, but it may of crossed cost / power thresholds that didn't make it viable. they may also of looked at other things like ram speed and keeping the CPU fed with enough data that there was no economical design that made it work and it was idling waiting for data I/O. we'll never know 🙁
 
Windows on arm only works virtualized because of Apple’s choices. They could fix that if they wanted to
To be fair to Apple, Microsoft also made the choice to have an exclusivity deal with Qualcomm for Windows on ARM so for the longest time Apple couldn't offer BootCamp with Windows on ARM for Apple Silicon even if they wanted to. Apparently that exclusivity deal ended sometime in 2024 so Apple could now offer BootCamp.

While Apple would probably have been motivated to offer BootCamp for Windows on ARM back in 2020-2021 to help ease the Intel to Apple Silicon transition and maintain existing functionality just like Rosetta 2, the lack of Windows on ARM BootCamp clearly hasn't been a showstopper for most Apple Silicon Mac users. There probably isn't as much incentive to support Windows on ARM BootCamp now.
 
I find it extremely unlikely another Intel model was genuinely under consideration

Makes no sense

In an alternative reality where Intel had not screwed up their 14nm->10nm transition it might have made sense.

The Mac Pro 2019 arrived somewhat late waiting on Intel CPU (incrementally delayed) and AMD GPU (also delayed) components. The Mac Pro 2019 couldn't have been a Q1 2019 device because stuff wasn't ready.

Intel's "Ice Lake" Xeon W class chip was suppose to ship in 2020 on some of the early timelines Intel released (although Intel timelines were becoming more of a joke over time). It was based on 10nm which substantially slid on release date. That would have made sense for the Apple to use in late 2020 ( as a last gasp for Intel) or even maybe Q1 2021. That generation was largely ignored. Its 'paper launch' shipped in Q3 '21'.

HP , Dell , Lenovo ... all largely. skipped them. Those vendors largely all went with AMD for workstation processors in that era. Apple pragmatically didn't have a choice but to stick with Intel... who fumbled that specific generation extremely badly.

Additionally, there was no M1 Mac Pro. Mac Pro didn't start until M2. Something in late 2020 would fit. Mac Pro in 2019 , 2021 , 2023. would have put Apple on a regular 2 year iteration schedule for the Mac Pro. After pulling a Rip van Winkle for 6 years and pragmatically 4 year before that , it would gone a long way to healing their reputation.

[ Aside. the 6800 MPX cards came in 2021. So a 2021 Intel model that wasn't a fumbled 10nm implementation would have paired better with those. It was a quasi new Mac Pro in 2021 with the new MPX modules. However, the RDNA2 architecture that of 6000-series had been used in other 'last gasp' Intel Macs also. ]


All that said.... Apple should have had access to early "ice lake" Xeon W and if it was garbage ... that project shouldn't even had proceeded past that initial stuff being horrible. I can see Apple start the prep for that, but how badly Intel was executing at that point put it at risk from the start.

Apple's 'Extreme" chip didn't make much economic sense either. So a gap filler Intel at the very top end made some sense. If Intel had produced an extremely good chip on time it would have made even more sense.
 
Windows on arm only works virtualized because of Apple’s choices. They could fix that if they wanted to
Windows for ARM only works virtualized because Windows for ARM is not the same thing as Windows for Apple Silicon. Apple Silicon may use the ARM instruction set but the hardware is otherwise radically different to anything that Windows supports.

Microsoft would have to write new drivers for everything - including accelerated graphics/media engine/NPU/storage controller etc. Under virtualziation, the "drivers" can just be a comparatively simple translation layer between Windows and the native MacOS drivers.

Apple don't actually block 3rd-party OSs from booting - just don't help or support it. Hence Asahi Linux can boot - but they have to reverse engineer everything and its still a long way from being a "shrink wrap" product or supporting anything past M2.

Even if Apple shared information with MS, MS would have to put in a significant amount of work to develop and support it - and they have zero incentive to do that. A good native implementation of Windows on a MBP, Mini or Studio would run rings around the current Snapdragon stuff - but unless Apple start selling Silicon to HPDelnovo, PC makers wouldn't be able to make competing machines. Meanwhile, MS can still make good money off Mac users by selling Office 365 subscriptions & windows licenses for virtualization.

(That's apart from the rumours that MS had some sort of exclusive deal with Qualcomm...)
 
I'm sorry that at least one of these models never saw the light of day. I've used Mac tower machines for practically my entire professional IT life, and I was very sad when Apple discontinued the latest edition of the Mac Pro. I would like to see it return to our professional niche in some form, and I think it was Apple's own fault that these machines were in the shadows. With a little more attention and a different pricing policy, they could still be extremely successful. In addition, these are still the last professional Macs to offer ECC memory, which is extremely important and a dealbreaker in some businesses.
 
  • Like
Reactions: phuklok1
That intrigues me. Would they have continued an Intel macOS branch for this machine?

The Intel Mac Pro was somewhat 'renewed' in 2021 with new GPUs ( W6800 w6900 ) ... did that get an extenion? Nope. If somehow Intel had made a very good 10nm chip in 2020-2021 then basically would have got same time line as have now. Some folks did buy new Mac Pros in 2021-2022 and things ended 'early' for them. Apple just would have chopped the support window on those also.

Left it on a specific OS but offered updates and patches just for that model?

It is extremely unlikely that one model of Mac can generate enough funds to support keeping a forked OS around. No one is paying directly for on going support or directly for the OS. So number of units sold is a key factor in amortizing the development and support costs. The Mac Pro was 'single digit' (most likley 1% or rounded up to 1%) of the Macs sold even back in late Intel era.

[. macOS on the Axx Pro chip is very minor deviation. Same baseline microarchitecture as the M-series , with bigger gaps in some of the I/O ( USB 3 and 2 instead of full Thunderbolt certification. ) . Also in terms of units sold order (or 2-3 ) magnitude difference also relative to Mac Pro. So even if some increased costs, more folks to amortize those costs over. ]


I understand why certain professionals would find this hugely attractive (the issues with card compatibility, Boot Camp, and probably very decent expansion options).

If folks were offering $500/yr per Mac Pro for Intel support and Apple would gross a lot more than several 10's of millions per year doing it. But Apple isn't doing it for 'free'. "find hugely attractive' but don't want to pay is a mismatch.

A substantive amount of chatter on Mac Pro in these forums are from folks who bought them used. So Apple saw no money from those at all. That isn't paying for longer support windows.

When we consider how long people had Mac II's and IIfx's—with all 8 slots filled and max memory—running in production environments (I'd be shocked if there aren't a couple still running somewhere), this sounds like the sort of machine that would have filled the same purpose. This would have been a fun hobbyist machine on the aftermarket in 5+ years.

I highly doubt Apple had ever considered that the primary purpose of the Mac Pro. Ever expanding workloads where will outstrip the Machine in 3-6 years is likely more closer to the target market. Folks who going to sit and squat with largely static workloads isn't really the Mac Pro space. It just 'happened' to fall into the also useful for that category , but that likely never was the target market.

the Mac Pro 2023 somewhat drift more in that direciton ( folks want new home for large set of legacy cards of fixed bandwidth).... but that lead to the end of the line... not a healthy product.
 
While Apple would probably have been motivated to offer BootCamp for Windows on ARM back in 2020-2021

"BootCamp for Windows on ARM" is a meaningless concept unless Microsoft produce an Apple Silicon version of Windows.

"Bootcamp" on Intel only works because Intel Macs were a gnats whisker away from being "PC compatible" clones and regular x86 Windows almost-but-not-quite "just worked".

Some Macs (e.g. the original Mac Pro) could just install and run Windows XP straight off the standard installation media. Other Macs needed some horsing around to make the installer work - plus, you ran the risk of the Windows installer hosing your Mac OS system.

"Bootcamp" (full name Bootcamp assistant) was just ab point-and-drool helper tool for partitioning the disc, downloading the (mostly) third party Windows drivers for generic hardware and patching the Windows installer to work properly on all Macs. When that was done, Bootcamp went away and left an installation of standard Windows with the correct drivers set up.

ARM systems may all use the same instruction sets, but don't have the highly standardised architecture of PC hardware - and Apple Silicon is like nothing on earth (well, except the iPad), with virtually everything on the SoC being proprietary Apple hardware. You can't get regular Windows-on-ARM running with just a few tweaks and off-the-shelf drivers.
 
  • Like
Reactions: phuklok1
J170 was apparently a new Intel-based Mac Pro, which is notable because Apple had already completed its transition to Apple silicon by the time the project was in active development.

This is quite likely not true. Apple could not have completed their transition to Apple Silicon without doing a Mac Pro on Apple Silicon. After shipping a Mac Pro on M-series there likely would not have been a co-appearing or after completion Mac Pro Intel model.

Could Apple has shipped a Mac Pro on Intel after they "about two years" deadline. Again extremely unlikely.
Could Apple have dropped a Mac Pro in 2021 to match the MPX AMD 6000 series? Yes. But Apple's transition would not have been complete at that point. The Intel Mini was not really complete either in 2021. 2021 was still under their two year target window.

There is a fanciful "completed' where the Mac Studio M1 Ultra replaced the Mac Pro , but it didn't. It was only Apple's transition replacement for the large screen iMac.

Gurman did not specify what "certain uses" the Intel model was meant to address, but its existence suggests Apple explored keeping an x86 option alive for specific professional workflows even years into the Apple silicon era.

The Mac Pro Ultra didn't appear until 2023. So it was 'years" from 2020 to 2023. If Intel Xeon W "Ice Lake" wasn't garbage in 2021 that could have ridden out the last two years better than the 2019 era CPU did. After 99% of the Mac market went to Apple Silicon there is no viable market left for a forked OS with low units and relatively low revenues.

Mac Pro got new GPUs when other much higher volume Macs got new GPU architecures. when those all leave no viable driver market either.


J190, meanwhile, was said to be an M3 Ultra-equipped Mac Pro that was supposed to launch alongside the M3 Ultra Mac Studio in early 2025, but Apple shelved it before release. It would have paired the same chip found in that Mac Studio with the Mac Pro's unique-to-the-lineup PCIe expansion slots.

Unique PCIe slots but if still PCI-e v4 and same subset card coverage ... how many would they have sold?
Apple doesn't have enough M3 Packages for Studio sales now. They'd have even less if splitting limited inventory with a Mac Pro product.

Sharing the same chip is a dual edged sword. If both products move on from the M2 Ultra then that package is dead ended. (no more products to "hand me down" into. ). So make Ultras and then throw them away in the trash can every 12-18 months. Economically that is not likely to work and the unit run rates Apple is consuming.
[ The 'AI server' outlet is beyond quirky in terms of volume. Maybe once Apple gets 'Apple Intellgence' up to a much more broad and consistent usage level. ]


Apple had also developed "Extreme" chip variants, processors with twice the CPU and GPU cores of Apple's Ultra chips, for a future Mac Pro before winding down those projects over cost and demand concerns.

If Ultras are two expensive to throw in the trash can every 12-18 months... "Extreme" (twice as big package) would even more so economically detached from that practice.
 

There were rumours of an Intel Ice Lake Mac Pro for 2022. With Ice Lake's Deep Learning Boost instructions and support for 4TB of RAM alongside RDNA3 GPUs, maybe one of those "certain use cases" was machine learning/AI development?

But Ice Lake was largely a 'paper launch'. It wasn't a good chip. Most of the features hyped by Intel didn't get around the problems it had with excess heat and being very late ( as a 10nm chip) and not particuarly competitive (to AMD's contemporary options). HP, Dell, Lenovo didn't really use the chip either. It wasn't just Apple that skipped it.


Or an Intel Mac Pro after the Apple Silicon transition (ie. alongside or after the M2 Ultra Mac Pro in 2023) would point to a Intel Sapphire Rapids based Mac Pro which with 56 CPU cores and also 4TB RAM support would have been very interesting.

Sapphire Rapids was another Intel bungled, delayed roll out. 12 steppings iterations until they chased out the bugs there. Once 'fixed' Sapphire Rapids wasn't too bad ( still not competitive in several ways to AMD Threadripper). But this is WAY too late after the rest of Mac line up had completely walked away from 3rd party GPUs. If don't have a viable GPU to pair with the CPU then there is no viable system there.

The Mac Pro by itself isn't a viable market to get a set of new generation drivers ported.
 
  • Like
Reactions: atonaldenim
The moment the first Mac Studios came out with Ultra chips, the fate of the Mac Pro was sealed imo. Then the lousy m2 Ultra update came out which simply was to complete the transistion, didn't have much going for it, except PCIe. It was just a Mac Studio with an insane amount of cooling and larger form factor.

The Trash Can was their idea of what they wanted for a "pro" level machine in 2013 and they suceeded with that to some extent with the Mac Studio years later. I agree with others, the market for these were so low because there was anything differianting from the Studio.

For high performance builds and high I/O, there's just much better value with AMD threadripper imo.
 
  • Like
Reactions: waz0wski
  • Memory needs over 192GB (the last Intel Mac Pro could do 1.5TB).

Capacity isn't as big of a disconnect as lack of ECC. Gigantic capacity with no ECC is still 'inelgible system' for lots of data integrity critical workloads.

  • Any software and processes that cannot run in the Apple hypervisor, and required direct execution on an Intel system (anything that needs AVX, for example).

Rosetta 2 currently covers several Intel specific extensions. The initial Rosetta 2 didn't cover AVX but Apple did add most of that in over time.

"... Rosetta translates all x86<wbr data-v-880086a2="">_64 instructions, including ones from the AVX and AVX2 instruction set, but it doesn’t support the execution of AVX512 vector instructions. If you include these unsupported instructions in your code, run them only after verifying that they’re available. ..."

Intel consumer chips dropped AVX512 so no big mismatch there.

If looking at it from a 'decades of support" viewpoint then Rosetta 2 is going away. It was always presented as a scoffold to making the transition; not as 'the solution forever'.


As for the hypervisor. It is more so that Apple doesn't really support 3rd parties in the kernel. That is coming to macOS in general. So ancient IOKit drivers would have had a problem long term on Intel also. DriverKit places 3rd party drivers is a in-between zone between kernel and general user space. Similar to issue with 32-bit Intel code. I That code category is gone from the supported library frameworks.



  • Anyone who's using apps that won't run properly and Apple Silicon and I no longer getting updates.

Abandoned apps and drivers are not really in Apple's "skate to where the puck is going" approach to most things.
If it wasn't viable for these developers to spend money on it , why is it going to viable for Apple to do it. ( other than take money from a large group of folks who do pay and hand to do groups of folks who don't pay.)
 
This is sad. It suggests Apple struggled internally with the decision to discontinue the Mac Pro. The simple truth, from Apple’s perspective, is that the Mac Pro was simply too niche to continue developing and marketing.

'Too niche' probably isn't an accurate description. Apple is reportedly building a custom Private Cloud Compute (PCC) service package just for certain Apple Intelligence workloads that don't fit on the locally run models. That is a very narrow workload; niche. The rest of Apple cloud services are probably going to stay running on something else.

The problem more so is that the Apple Silicon group has one , and only one, customer: Apple. It can only go into an Apple product ( or with server chip an Apple Service). The chips are not being sold to other system vendors so there is a fixed number of systems can go into for a fixed amount of time (if Apple churns the products regularly... chips get kicked out of products at a steady rate. If no place else to go ... then into the 'trash can'. ).

The Mac Pro used a substantively different product line then the rest of the Mac line up in the Intel space. Xeon Workstation chips were different die and different support chipsets from the consumer stuff that Apple used. Intel has tons of system build customers so the serve ( and workstation variant) dies went into hundreds of designs , and systems worth 100's of millions of dollars collectively. So the costs for doing the chips was spread out over a large group.

Apple is skipping the 'Ultra" on generations also.

The serve chips may work if Apple starts charging money for them. ( Some Apple Intelligence Pro features are being coupled to paying a high enough price of iCloud+. ) . I suspect that will expand as folks see more value in more of the PCC services. If not, the server chip probably will die approximately as fast as the Apple silicon Mac Pro did.

The Mac Pro never fully supported the R&D costs of its components development ( CPU , GPU , PCI-e switch , etc). There was a ecosystem of other systems that all helped defray the costs over a much larger set of systems. Going to Apple Silicon means detaching from large chunks of the x86 ecosystem. Similarly when rest of Mac ecosystem completley duimps discrete GPUs... all opportunity for shared costs go out the window.

The 'decoupling' is what is the driving force there. not niche.


Even the Mac Studio makes up a low sing-digit percentage of total Mac sales, and that’s as far as Apple’s willing to extend itself to accommodate the pro desktop market.

In the Intel era the Mini was largely handicpapped by the Intel CPU+GPU offerings. The "Mini Pro" is not. And Apple needs it to be not handicapped. They need other product(s) for the Mn Pro chip to go into beside just the Macbook Pro . Throwing Pro and Max class chips in the trash can every 12 months likely doesn't work economically either.

Intel , AMD, Nvidia ... no one of them are tossing medium large chips they developed in the trash can every 12 months. They may introduce new 'more shiny' chips to buy with 12 month regularity but they will not stop selling the older stuff completely.

[ If look at the iPhone side Apple doesn't either. The iPhone gets sold for 3-4 years with that chip in it. As the 'An Pro' chip diverged to being a seperate die ... surprise , surprise it starts showing up in other products. iPad Mini , Mac Neo . etc. ]
 
  • Like
Reactions: hieranonymous
And now dgx station is available from nvidia partners. 72 cpu cores, lots of unified mem. Perfect for local ai. People pay anything to get hold of those.
M5 extreme mac pro could have been more or less that. 4x18 cpu cores and 4x40 gpu cores. Above 1200 tb bandwidth (2000?) etc etc
Why is Apple content to not be on the forefront of the current computing trends and instead just cater to the consumer and prosumer of what was hot 5 years ago (movie clip production for the attention economy)
 
  • Like
Reactions: phuklok1
Capacity isn't as big of a disconnect as lack of ECC. Gigantic capacity with no ECC is still 'inelgible system' for lots of data integrity critical workloads.



Rosetta 2 currently covers several Intel specific extensions. The initial Rosetta 2 didn't cover AVX but Apple did add most of that in over time.

"... Rosetta translates all x86<wbr data-v-880086a2="">_64 instructions, including ones from the AVX and AVX2 instruction set, but it doesn’t support the execution of AVX512 vector instructions. If you include these unsupported instructions in your code, run them only after verifying that they’re available. ..."

Intel consumer chips dropped AVX512 so no big mismatch there.

If looking at it from a 'decades of support" viewpoint then Rosetta 2 is going away. It was always presented as a scoffold to making the transition; not as 'the solution forever'.


As for the hypervisor. It is more so that Apple doesn't really support 3rd parties in the kernel. That is coming to macOS in general. So ancient IOKit drivers would have had a problem long term on Intel also. DriverKit places 3rd party drivers is a in-between zone between kernel and general user space. Similar to issue with 32-bit Intel code. I That code category is gone from the supported library frameworks.





Abandoned apps and drivers are not really in Apple's "skate to where the puck is going" approach to most things.
If it wasn't viable for these developers to spend money on it , why is it going to viable for Apple to do it. ( other than take money from a large group of folks who do pay and hand to do groups of folks who don't pay.)
A few notes:

ECC is a big one, that should have been on my original list, thank for pointing it out. However, capacity is a big considering of some tasks. I saw many people stating at the time that they couldn't upgrade until Apple offered more memory for computational tasks.

Yeah, I should have elaborated that it was later added, but at the time Apple was likely working on this last Intel Mac Pro, it was probably a consideration. So, timing is key here.

I would not ever consider Rosetta a long term options, but many people may have wanted one last Mac Pro as a hold over (little did they know that the Mac Pro as a product was doomed, anyway).

More so the issue with mentioning the hypervisor is that the Apple Silicon Hypervisor is made for virtualizing AArch64 only, not x86/x64. This is why I have a 2013 Mac Pro siting on the other end of my desk from my Mac Studio.

"Abandoned apps and drivers are not really in Apple's "skate to where the puck is going" approach to most things.
If it wasn't viable for these developers to spend money on it , why is it going to viable for Apple to do it. ( other than take money from a large group of folks who do pay and hand to do groups of folks who don't pay.)"

I'm not 100% sure what you mean by this, so I'm going to reply based on what I infer: that is a valid statement from a business philosophical perspective, but in reality high end professionals aren't just buying computers to upgrade to the new thing, or because it's feeling old, like consumers do, they're replacing faulty ones or expanding their infrastructure, and many times certain critical software or drivers for expansion cards and peripherals may not be available, or available yet (since we're talking a transitional machine here), so for many, moving to Apple Silicon wasn't an option, due to a significantly larger investments prior to the announcements to Apple Silicon, so a transitional machine would have been very important. Again, we saw this with the move to the G5 and the removal of native OS boot support.
 
This fits the public perception we have that there was pretty clearly a power struggle within Apple over the Mac Pro, and the people who wanted it gone clearly won.

I really want to know who developed that Intel Mac Pro, and who was the driving force behind its cancellation.

I have a feeling Tim Cook had no strong feelings either way, and let his lieutenants fight it out and bring him their decision.

Either that, or his only question was to make a financial case for it. It seems engineering/marketing created a machine with no financial case.

Hope we get a book about this and the rest of Cook's tenure some day soon.

Internal struggle within Apple seems like a plausible read of the situation. I’d guess Ternus was on the side of continuing the Mac Pro, the keynote presenter at the 2019 Mac Pro’s unveiling. Competing internal priorities explains the conundrum that vexed me in 2022 before we knew if the Mac Pro would ever get updated again: why would Apple have put SO much engineering effort into designing the 2019 Mac Pro as a kickass tower workstation (including the custom MPX slots cooling solution), if they knew Apple Silicon was right around the corner and were going to abandon the new Pro “shredder” design after only 1-2 generations? Why not follow up the 2013 trash can Mac Pro with the same slot-less concept of a “2019 Mac Studio” instead?

Gurman’s report answers: yes, (some parts of) Apple did intend for the new shredder Mac Pro to be a long-lived platform designed with enough future expansion capability to support many future model revisions and even architecture transitions, just as the Power Mac G5 design became the Intel Mac Pro with only a few little tweaks. But broader company and market developments ultimately curtailed that long term vision from becoming a reality.

deconstruct60 makes the good point that the axed Intel Mac Pro was probably a speed bump update to the 2019 model, that was delayed by Intel supply to the point where it became untenable to sell within the context of the overall Apple Silicon transition. It wasn’t until M2 generation that Apple stopped selling the higher end Intel Mac Mini for example. Similarly a 2020-21 speed bump to Intel Mac Pro could have still been defensible within the architecture transition, given the niche customer base paying attention to this model.

But beyond a certain point, Apple as a company has long forced people to embrace the future by abandoning support for the past, ever since Jobs’s return and the success of the all-USB, no floppy disc iMac. And continuing to offer a newly updated Intel-based Mac Pro for very long after the launch of the Apple Silicon Mac Studio could have significantly undermined the success of the ARM transition among the market of pro customers, if it gave pro software developers an excuse to drag their feet on supporting Apple Silicon at all.

The M3 Ultra Mac Pro may also have been victim to similar supply chain limitations as the Intel Mac Pro 2020-21, as others noted. Good point ChrisA that keeping these options in the back pocket is a wise strategic imperative even if never ultimately deployed.

Apple Silicon Mac Pro’s inability to buy base model now, upgrade most components later, is a death blow to the prosumer market that represents most people posting here on MR Forums, myself included. $6000 for a computer is a very big stretch for most enthusiasts like us, but we can rationalize the initial expense if we know we’ll be able to add more RAM, CPU, GPU, internal storage later on down the road, at lower 3rd party or even used prices. Maybe in a few years we’ll be able to piecemeal our way to a $10,000 system that Apple would have charged $50,000 for.

But if we have to finance the whole desired system config upfront at Apple retail prices, sorry no that is beyond enthusiasts like us. So that shrinks the market somewhat to only the corporate buyers who can afford Apple’s full prices, which admittedly still probably accounts for the majority of the Mac Pro’s revenue to Apple anyway. But the hearts and minds of the fan base are lost, so the Apple Silicon Mac Pro goes out with a whimper. M3 Ultra Mac Studio with faster Thunderbolt 5 PCIe expansion capabilities eliminates the last remaining justification for a Mac Pro with internal PCIe slots, once GPU expansion is off the table. TB5 allows for 100 gigabit networking and basically anything else people might need a PCIe slot for.

Anyway, thanks Gurman for confirming that those of us holding out for more Mac Pro updates weren’t entirely deluded, they really did exist within Apple. (Except for the people wishing for a dream Mac Pro built with stackable “compute module” blades, yall were always dreaming!)
 
Last edited:
  • Like
Reactions: CarAnalogy
Windows on arm only works virtualized because of Apple’s choices. They could fix that if they wanted to
Whoever has the problem (which one wants have to fix it) at the end is irrelevant, the truth is that macs only boot Mac OS, Apple Silicon doesn't works with other Operating System (fully), has been 5 iterations of apple silicon and doesn't seems to be a future in where you can boot Windows and MacOS on the same computer as with x86 Macs.
 
To be fair to Apple, Microsoft also made the choice to have an exclusivity deal with Qualcomm for Windows on ARM so for the longest time Apple couldn't offer BootCamp with Windows on ARM for Apple Silicon even if they wanted to. Apparently that exclusivity deal ended sometime in 2024 so Apple could now offer BootCamp.

While Apple would probably have been motivated to offer BootCamp for Windows on ARM back in 2020-2021 to help ease the Intel to Apple Silicon transition and maintain existing functionality just like Rosetta 2, the lack of Windows on ARM BootCamp clearly hasn't been a showstopper for most Apple Silicon Mac users. There probably isn't as much incentive to support Windows on ARM BootCamp now.

Nvidia SOC will boot windows, the problem is UEFI, macs doesn't use UEFI, BootCamp was their way to "fake" UEFI, so until apple and microsoft work together (I doubt they will) macs will only boot MacOS (limited to some linux variants).
 
Whoever has the problem (which one wants have to fix it) at the end is irrelevant, the truth is that macs only boot Mac OS, Apple Silicon doesn't works with other Operating System (fully), has been 5 iterations of apple silicon and doesn't seems to be a future in where you can boot Windows and MacOS on the same computer as with x86 Macs.

Nvidia SOC will boot windows, the problem is UEFI, macs doesn't use UEFI, BootCamp was their way to "fake" UEFI, so until apple and microsoft work together (I doubt they will) macs will only boot MacOS (limited to some linux variants).

Apple. It's Apple that doesn't want to fix it. For better or worse.

As much as I'd like it, I can't really blame them.

Technical issues aside, they're at a point now where they feel they don't need Microsoft anymore, not for this.

They're probably right.
 
Apple. It's Apple that doesn't want to fix it. For better or worse.

As much as I'd like it, I can't really blame them.

Technical issues aside, they're at a point now where they feel they don't need Microsoft anymore, not for this.

They're probably right.
That's correct, they won't make Windows easy to install, they want users to keep using MacOS.
 
Register on MacRumors! This sidebar will go away, and you'll see fewer ads.