Then our Intel Macs will be relevant for longer and we'll be less incentivized to get an Apple silicon one.
Got a tip for us?
Let us know
Become a MacRumors Supporter for $50/year with no ads, ability to filter front page stories, and private forums.
Apple Transitioning Macs to Custom Chips With 'Incredible' Performance Starting Later This Year
- Thread starter MacRumors
- Start date
- Sort by reaction score
Then our Intel Macs will be relevant for longer and we'll be less incentivized to get an Apple silicon one.
Most pros don’t need a Mac Pro anymore. In 2010, they did. A non-Pro machine isn’t only good for email, and a Mac Pro isn’t only for Hollywood studios or 3D animation.
If MBP or iMac aren’t powerful enough, Mac Pro is. Most would rather have MBP, and it’s not because they can save three or four thousand dollars over a Mac Pro.
If you haven’t read the transcript of the 2017 Mac Pro round table, you’re missing out on a lot of insight into how Apple thinks about pros and pro use of the Mac in general, and the Mac Pro in particular.
Transcript: Phil Schiller, Craig Federighi and John Ternus on the state of Apple's pro Macs | TechCrunch
You have already read the news. But we thought we would also use this opportunity to share a transcript of the interview with Phil Schiller, Senior Vice
techcrunch.com
Apple would be better off to stop selling all Intel Macs. They are worthless now that the transition is announced. Users don't need another worthless anchor on their desks. I lived through the last transition by not buying a new computer for 4 years. Advice here is the same.
These are tools and none of them stopped working with the last transition and they won’t this time around. The last transition was completed in about a year and a bit, this one won’t take forever either.
Apple needs to definitively keep selling tools for those who need them.
But the hardware transition took about 4 years, and 10.6.8 hung around on a significant number of machines for that time period.
People that jump to ARM will be replacing both their hardware and their software. And they will be renting a lot of that software (Adobe & Microsoft).
Why would that be? Its just a different processor platform, anything else remains the same
I’m not sure I’m understanding you right here. What do you mean by the transition taking 4 years? MacBooks, iMacs and Mac Pro were released in 2006 already and the announcement was 2005 that they’d switch to intel. Looks like less than two years to me, what am I missing?
The last PowerPC iMac was released in October of 2005. The first Intel iMac in January of 2006.
The last computer to move to Intel, the Mac Pro (née Power Mac G5), was released in August of 2006.
10.6 didn't boot PowerPC. The last OS to boot PowerPC was 10.5.8, released August 2009.
Given that they will typically be renting it, there isn't much of a replacement to do. You install a newer Adobe CC or Office 365.
Agreed - moving from a commodity CPU that can run any OS to yet another proprietary one (a tactic that that failed last time) without any tangible benefit seems crazy.
Having full control over release schedule, pricing, feature set seems quite a tangible benefit to Apple.
Likely much higher single-threaded perf seems quite a tangible benefit to us.
I watched the presentation after that and I like the idea of the iPad and Mac sharing apps. Kind of looking forward to this transition now.
You are missing that hardware needs software.
It took 4 years for the software to make the transition - and 10.6.8 hung around a lot longer than any other version of OSX with a LOT of people.
[automerge]1595281890[/automerge]
No, but it was the last version to run Rosetta. Which is why 10.6.8 hung on a lot longer than expected.
Some places rent software, but a lot of places don't.
Thank you for the additional context. You earlier on claimed that the hardware transition took 4 years hence my initial confusion. Software surely takes time at places but even if it took up to four years for some companies, that is a decade plus in the past.
Apple will turn this change to their advantage otherwise they wouldn’t attempt it.
The poster I responded to earlier suggested stopping all current sales which just doesn’t make sense to me. These are tools and you get them when you need them.
Most pros don’t need a Mac Pro anymore. In 2010, they did. A non-Pro machine isn’t only good for email, and a Mac Pro isn’t only for Hollywood studios or 3D animation.
If MBP or iMac aren’t powerful enough, Mac Pro is. Most would rather have MBP, and it’s not because they can save three or four thousand dollars over a Mac Pro.
If you haven’t read the transcript of the 2017 Mac Pro round table, you’re missing out on a lot of insight into how Apple thinks about pros and pro use of the Mac in general, and the Mac Pro in particular.
Transcript: Phil Schiller, Craig Federighi and John Ternus on the state of Apple's pro Macs | TechCrunch
You have already read the news. But we thought we would also use this opportunity to share a transcript of the interview with Phil Schiller, Senior Vice
It's not just performance, it's serviceablity. The Mac Pro is actually possible to quickly repair if it were to fail, the same is not true of the Macbook Pro or an iMac - if anything the Macbook Pro will end up in the trash heap with some holes drilled in it if something major were to fail in an enterprise environment.
The the move to non-serviceablity is 75% of the reason why my workplace has dropped the Mac - the remaining 25% is declining software support and compatibility.
Serviceability is so far down on my list of requirements that it doesn’t factor in. Sometimes it’s a choice of the frying pan or the fire.
But if the platform can’t meet your requirements, it makes sense to switch to something that can 👍
One more time people. "Proprietary" just means it's marketed under a registered trade name, i.e. not a generic/no-name brand product.
Also the idea that Apple uses "commodity" CPUs in any regular sense of the term (aka "off the shelf items that can be interchanged with similar parts") is pretty ridiculous. Most models have a soldered CPU, and those that do have socketed CPUs (27" iMac, Mac Pro, maybe iMac Pro?), are limited to what they support (either due to the physical socket or the EFI). You may as well say that your Toyota has a "commodity" engine because you could go to town and rip out the engine, to replace it with a Honda engine, if you try hard enough.
I thought it was a valid use of proprietary.
Intel sells those CPUs for the express purpose of putting them in any PC. Toyota or Honda don't sell those engines for the purpose of putting them in any car.
How? What definition of proprietary fits an Apple-designed CPU using the ARM instruction set, but doesn't fit an Intel CPU using the i386/amd64 (to use Linux kernel nomenclature) instruction sets?
My point is that some of the Intel CPU's apple uses being available to buy separately means 0 in the general sense of the term "commodity hardware" when it comes to Macs specifically, because they're mostly soldered and those that aren't are EFI limited, or socket limited.
I agree in the sense that the average 'non-pro' machine is now capable of doing some fairly high-end stuff, just not as quickly. The gap has narrowed, at least.
The problem is more around long-term reliability and heat dissipation (and stuff like environmental noise, if that matters to you). Any machine Apple makes besides the iMac Pro and Mac Pro aren't designed for longer-term heavy-crunching type stuff.
Again, I think we're down to the definition of Pro. I'm talking about a class of type of work and quality of the tools, not whether one gets paid vs home-user. (The latter definition is silly, given my illustration of a lawyer and a chrome-book, or the very example here of a hobbyist doing 3D rendering.)
Apple seems to be more running on the latter these days too. Look, here's a 'pro' laptop with a fancy Touch Bar and some higher specs that can barely be used, rather than something actually usable by the higher-end pro. (Ok, a bit hyperbolic, but makes the point, hopefully.)
While the hardware transition could happen quickly (in theory), the software one won't. People will need Intel Macs for some time yet, and even if the new Macs are 2x performance, it doesn't suddenly make the Intel Macs worthless (even if for resale, but not for usefulness).
But aren't they already mostly renting the software? I've got a couple of packages still on the x-time-period release, but the majority of my stuff is subscription. (To be fair, the couple non-subscription are the higher-end, more expensive stuff, besides Revit which I don't yet own.)
Yeah, less than 4 years. I suppose that's what people are worried about here, too. It kinda sucks to buy a new machine and then have it essentially EoL in under 4 years. More so for home users, hobbyists, prosumers than businesses, but still.
It depends on what kind of speed gains we see. If Apple just stays like 1.2x the performance of Intel, then I'd say you're right. If Apple starts heading into the 2x to 4x range, then it's well worth it.
That might be the case for some, but I think it was more due to the rough OS releases that followed Snow Leopard. They also ditched a lot of other functionality/features like a true server.
While I'm not sure I could go back and lose some of the features we now have, I would in a heart-beat in terms of OS quality. Snow Leopard was so popular because it was the last quality OS for the Mac Apple produced. I hope we'll get back there some day in a current OS update.
The big difference today, is that people can't realistically sit on an old OS any longer. The tech behind browsers and security is too important to keep current (unless it is just a machine to run some app, separate from the main-use machine) not to keep up on updates.
So, a new, non-compatible OS release (which software packages follow), will obsolete a perfectly capable piece of hardware. (This wasn't so true a decade or two ago.)
Think of it more big-picture than the ability to swap a chip. Apple is going from a machine that can run almost any piece of software, to one where you'll need to run software specifically designed for the new platform. It makes sense in that context.
My fingers were typing hardware while my brain was thinking software.
I am sure this will be an advantage to Apple - customers, not so much. This is about cost control, not performance. As an added bonus, it is one more step to only being able to buy software through the Apple storefront.
AFA Intel Macs, anyone paying full price for an obsolete piece of hardware isn't very smart. Intel Macs literally have no future.
It is why I am now a typical Apple customer - I dumped my macs for a Ryzen based system that will outperform a 7,1 Mac Pro & kept the phone & iPad.
I wouldn’t necessarily call Intel CPUs commodity, but wouldn’t argue with someone who did. However, they’re certainly Commercially-available Off The Shelf (COTS) and there are non-Intel alternatives. Whether or not the CPU is soldered or socketed is a red herring, and irrelevant.
However, your definition of proprietary is much too restrictive. Apple is moving to a proprietary—yes, that is exactly the correct word—solution that is 100% exclusive, and hence proprietary, to Apple. Keep in mind Apple’s silicon is not only a CPU but many other co-processors as well.
Last edited:
And how likely do you think it is that Apple can beat INTEL, the primary designer and manufacturer of CPUs in the world, at their own game by 2x to 4x?
Just having smaller instructions in the CPU that can each execute quicker doesn't make the overall computation faster, necessarily (since you end up needing many more operations to get the same thing done). RISC has been around a LONG time and is pretty well understood. I'm not sure it's the panacea some here seem to think it is. Although I will admit to being intrigued at the possibility of iPad and Mac apps being somewhat interchangeable.
Every computer ever made literally has no future -- It's just a matter of time frame. An Intel Mac purchased today easily has about a three year window of viability, I think. Could be longer, but it almost certainly won't be shorter than that.
Maybe the primary, but certainly not the best.
As to your question:
2x - quite.
4x - maybe.
Of course depending on what your definition of performance is.