Mac ARM models will flop. A pointless and dumb exercise. It took Apple 20 years to unify on Intel ... trying to be different and claiming PowerPC was faster didn't get them anywhere.
...an ARM processor would cut back on virtualization solutions that the "pro" market sometimes rely on...
Absolutely this. I've been able to replace my company-provided crappy HP and Dell machines with Macs over the years because I'm able to virtualize Windows for certain corporate applications that require it, as well as for being able to compile code that targets the Windows Server environment. I also virtualize Linux frequently as my development needs require. The loss of Intel CPUs powering Macs will render virtualizing impossible, at least for Windows, and we'll have to see how well the Apple processors are able to run emulation in place of virtualization.
My guess is - they will sell different machines for a time. This is a little different from what they did in the past transitions (and I think this need of accessing Windows will slowly dwindle in the future) but, tell me if you think I'm wrong, wouldn't that be a good differentiate between a PRO Mac and a non-Pro mac?
Remember when Ming-Chi Kuo changed jobs, and wasn't going to be reporting on Apple any more? Doesn't seem to have stopped him from throwing stuff at the wall and waiting to see what sticks.
Maybe he isn't a person, Ming-Chi Kuo is an alias used to conceal the real identity of one of the industry's most elite analysts. You know, like James Bond. Every few years the old one retires and gets replaced by a new one. Imagine the furore if they ever picked a female Ming-Chi Kuo.
Anyway, for Apple to go to fully ARM in the Mac, I think something's going to have to happen with Intel. The whole industry is caught up in x86 compatibilty. Apple could move their entire line of computers to ARM tomorrow, provided they have processors with the necessary power (which the A Series aren't yet), except for providing continued support to a Windows based Enterprise world. One of two things needs to happen - either Windows needs to cease to be relevant as either an applications platform or a server platform (yes, I know you have Linux servers, but not everyone does and some services require Windows), or everyone needs to abandon the zombie x86 platform and shift all their code base to ARM. (Given how much luck MS have had with shifting their own Office suite, that seems unlikely in the short term.)
I think what's actually going to happen (and I have no actual knowledge or inside information to back this up) is that the T Series (or its succcessor) ARM is going to take over more and more of the front end OS, freeing up the x86 for heavier application loads. Eventually, the x86 will only be necessary for the high end apps, and will essentially be treated as a remote application server by the OS. Ultimately, they might push all of the Pro apps into the Cloud, so that a local x86 (or successor) CPU isn't needed. And we end up back with a Terminal-Server model by stealth.
That's not a bad theory. I think it will be a differentiator in the future between Pro and non-Pro Macs.
Well, yet more "walled garden" treatment from Apple. While moving Macs to "A" series processors may be good for Apple, it certainly doesn't do much for those of us who actually use real software tools and virtualization in our work. How long do you think it would take (if at all) for suppliers of professional software tools (e.g. Adobe CS suite, Blackmagic video, etc.) to port over their applications with any sort of stability?
Sure, casual consumer users content to rely on Apple's own software (Photos, Safari, Pages, etc.) wouldn't see much difference, but those who actually use Macs to make a living may very well have huge problems. Of course Apple is doing everything it can to demonstrate they no longer care about pro users anyway, so maybe it's no big deal to them. Oh well, Microsoft and many PC suppliers have made it clear they DO care about the pro market, so there are alternatives.
Well, yet more "walled garden" treatment from Apple. While moving Macs to "A" series processors may be good for Apple, it certainly doesn't do much for those of us who actually use real software tools and virtualization in our work. How long do you think it would take (if at all) for suppliers of professional software tools (e.g. Adobe CS suite, Blackmagic video, etc.) to port over their applications with any sort of stability?
Sure, casual consumer users content to rely on Apple's own software (Photos, Safari, Pages, etc.) wouldn't see much difference, but those who actually use Macs to make a living may very well have huge problems. Of course Apple is doing everything it can to demonstrate they no longer care about pro users anyway, so maybe it's no big deal to them. Oh well, Microsoft and many PC suppliers have made it clear they DO care about the pro market, so there are alternatives.
I think you are mistaken MCK's prediction with Apple Keynote?
I expect Apple is more interested in transportation as a service, rather than a car to sell to an individual. There’s a lot of room for Apple to develop services around such an offering as a relates to the ease of use their clientele expect.
The reason I don’t use Uber, Lyft, taxis or buses is the same reason I like having my own car. I’m an adult who can afford it and using any of those alternatives usually involves a dirty, smelly or slow vehicle.
I’m hoping Apple offers lease or buy luxury autonomous vehicles.
What does being an "adult who can afford it" have to do with owning an automobile? Cars have a wide range of prices, including teenagers being given one or paying just a few hundred dollars for one and they probably can't afford to use cabs or ride sharing services too often. Would someone who lives in a major city, like NY, not be an adult or have less money than you because they don't drive in the city because there's no convenient place to park? What about those that take trains in Europe? What about people that responsibly will take a cab or ride share to meet friend for drinks instead of driving while potentially over the legal limit? I'd say these people were "adult" for being responsible and willing to fork over the extra cost to being driven over driving their own vehicle.
I agree with you Soli, but I just want to point out one thing which I know you are aware of, that in most cases, owning your own car is much more expensive than using other forms of transport as you outline. Also, in many parts of the civilized world using public transport is not a 'class thing', it is an environmentally friendly decision that has the added convenience of not having to deal with a car all day. Not a lesser expensive decision, just more practical.
I'm pretty that if you live in a place where you need to commute a long distance everywhere you go that taking a cab or ride share every time would be a lot more expensive. Say, 10 miles to college 4 days a week, 8 miles to work 5 days a week. Then going out and running errands in your average US city. In NYC I will often take a cab or ride share as opposed to the subway because I can afford it and it will save me time. What wouldn't save me time in NYC is having to drive myself and then look for parking not to mention the stress involved.
If the "average American driver puts more than 1,000 miles on their car every month" and you're a college student that was gifted a car for your HS graduation, and you need to pay for fuel for that 1000 miles at current prices and with the average MPG for city driving, plus annual registration and insurance, how far would that take you if you only used Lyft? I don't think it would get you anywhere close to a 1000 miles per month.
And what if I wanted to do a road trip. Would ride sharing be less expensive than owning an automobile? A less expensive option if live in a dense city with great public transportation, no reasonable place to park, and excessive fees for storing a vehicle would be to rent, but for the average American owning an automobile is going to the most sensible choice to cover a lot of distance over ground.
There are simply too many scenarios for Spam to state that public transportation and ride sharing is only for poor children.
To the naysayers, consider this. Cars used to be greasy, smokey, smelly, overly complicated, explosion-containing machines. With the evolution of battery power, sensors, cameras, and computers, they have become much more like a Personal Transport Device than the scaled-down version of a fire-breathing steam locomotive that conventional cars are. It’s not such a reach that Cupertino could put its Apple-y secret sauce on such a device. Tesla and others are doing it in California from scratch without the Detroit model. Simple, safe and user-friendly, with all kind of wonderful tech baked in. I’ll buy one.
Remember when Ming-Chi Kuo changed jobs, and wasn't going to be reporting on Apple any more? Doesn't seem to have stopped him from throwing stuff at the wall and waiting to see what sticks.
Maybe he isn't a person, Ming-Chi Kuo is an alias used to conceal the real identity of one of the industry's most elite analysts. You know, like James Bond. Every few years the old one retires and gets replaced by a new one. Imagine the furore if they ever picked a female Ming-Chi Kuo.
Anyway, for Apple to go to fully ARM in the Mac, I think something's going to have to happen with Intel. The whole industry is caught up in x86 compatibilty. Apple could move their entire line of computers to ARM tomorrow, provided they have processors with the necessary power (which the A Series aren't yet), except for providing continued support to a Windows based Enterprise world. One of two things needs to happen - either Windows needs to cease to be relevant as either an applications platform or a server platform (yes, I know you have Linux servers, but not everyone does and some services require Windows), or everyone needs to abandon the zombie x86 platform and shift all their code base to ARM. (Given how much luck MS have had with shifting their own Office suite, that seems unlikely in the short term.)
I think what's actually going to happen (and I have no actual knowledge or inside information to back this up) is that the T Series (or its succcessor) ARM is going to take over more and more of the front end OS, freeing up the x86 for heavier application loads. Eventually, the x86 will only be necessary for the high end apps, and will essentially be treated as a remote application server by the OS. Ultimately, they might push all of the Pro apps into the Cloud, so that a local x86 (or successor) CPU isn't needed. And we end up back with a Terminal-Server model by stealth.
That's not a bad theory. I think it will be a differentiator in the future between Pro and non-Pro Macs.
I think the ultimate end-game of this, at least as I've envisioned it, is that there is no differentiator between Pro and non-Pro hardware. Pro becomes a collection of services running on high performance hardware in a data centre near Cupertino (or wherever else they have them). Like iCloud on steroids. They may offer an enterprise solution where they sell or rent the hardware to house in your own data centre, but (for example) Final Cut Pro becomes a thin client connecting to a workflow on an application server somewhere. (Maybe they'll sell a box you can buy from the Apple store to plug into your network, and access via your MacBook.)
Basically, I'm just extrapolating from trends in high end computing in the enterprise. Apple may never go this way at all. They might announce a 7nm ARM based processor capable of full x86 emulation with desktop performance tomorrow. Or Intel might get their act together and start shipping some new CPUs. It's really just me thinking out loud about what Apple could do assuming things stay pretty much as they are.
See, it's really easy to do this whole Tech Punditry thing. Someone should hire me.
Car? I don’t think so. There is something so cool in a company that can sell high margin items in tiny little boxes that can be shipped via ups and fed ex and returned the same way vs something that has so many liabilities, needs tires, upholstery, brakes and servicing in garages and all the headaches that go with that. Yeah no way. This is the Apple TV all over again.
Let’s remember Ford is stopping car production to focus mostly on trucks and commercial vehicles. Tesla is not making any money really. A car factory? Where? China? The US?
It just sounds like a terrible idea all the way around.
Can't wait. A12 is up there with Skylake server performance per core, and NOT normalized per clock, just core vs core.
I hope Apple takes the opportunity to bring segmented Intel features like an AVX-512 equivalent down to more products. ARM has variable length SIMD so they can scale from 128 up to 2048.
I think Apple are way ahead on this and have application-specific silicon already for video encoding and web rendering etc. They’ve been using custom private ISA to complement ARM for some time. They only highlight the obvious, like the Bionic neural engine, and keep everyone guessing on the other stuff as to why their cores are so fast at low clock/power.
If Apple does move to A Series on the Mac, loss of virtualization becomes a real concern for the 3-5% of Mac users who need it. Could Apple offer a Pro model with a dual CPU configuration? Probably. But it would greatly complicate the MacOS needed to run such a thing. Would such an OS be worth Apple’s time and effort to maintain? Not at all sure about that.
They could offer partial solutions in place of virtualisation for legacy corporate apps; i.e. .NET core/framework ARM native with emulation for the Windows executable.
If Apple does move to A Series on the Mac, loss of virtualization becomes a real concern for the 3-5% of Mac users who need it. Could Apple offer a Pro model with a dual CPU configuration? Probably. But it would greatly complicate the MacOS needed to run such a thing. Would such an OS be worth Apple’s time and effort to maintain? Not at all sure about that.
Where is the 3-5% figure coming from? Your estimated guess?? My estimated guess is that it is at least 30%.
I am the owner of a software development company and in case I can no longer buy an Intel based Mac, the share of Macs in the company will drop from 50% to a single machine. A Mac that cannot run Windows or Linux in a virtual machine is a useless Mac for my company
Mac ARM models will flop. A pointless and dumb exercise. It took Apple 20 years to unify on Intel ... trying to be different and claiming PowerPC was faster didn't get them anywhere.
That is because PowerPC wasn’t faster at the things that counted. In fact at the time of the switch power PC could barely offer half the integer performance of Intels chips. There is a reason to Steve’s emphasis on floating point performance.
ARM will not be anymore of a flop than the current Mac line up. It really comes down to Apple pulling their heads out of the clouds and start making Macs people want. There are many use cases where an ARM based Mac makes sense. An important one is the power savings possible, especially for always on uses. Another is the rapid changes I expect to see in AI / ML code and thus the importance of Neural Engine.
ARM on Mac is critical to Apples future because we are quickly moving past the time where the CPU is important to overall system capability.
Not only that, cars spend on average something like 95% of their time parked in the driveway. The cost of car ownership, whether you can afford it or not, is wasteful. Driving and parking is a drag if you live in an urban setting. If you really love driving, I mean for real... go to a race track. The most fun that can be had on four wheels
The freedom to get up and go whenever I want to, without having to wait for someone else, hope they are skilled (and sober) enough to keep me safe, are honourable enough to stick to the deal, and have a vehicle (or body) that doesn't stink, makes owning my own car, as the commercial used to say, "priceless". In addition, it allows me to live away from anyone else and still be able to visit those I care about whenever the mood strikes me.
I expect Apple is more interested in transportation as a service, rather than a car to sell to an individual. There’s a lot of room for Apple to develop services around such an offering as a relates to the ease of use their clientele expect.
The reason I don’t use Uber, Lyft, taxis or buses is the same reason I like having my own car. I’m an adult who can afford it and using any of those alternatives usually involves a dirty, smelly or slow vehicle.
Nah, you just made that up. There are age requirements and I’ve never had a smelly or dirty car. They’d be voted out.
So because you haven't experienced it, it doesn't happen at all.
I have ridden in smelly Ubers and Lyfts both. Maybe eventually they are voted out, but how many votes does that take? And until it happens, there's still a chance you'll get one.
I would say it will start with Mac mini. Since there are rumor about mini pro line up, Base mini can be start $399 with A12X or better in it.
Some people already use only iPad for all their task at home. Unless they are heavy PC gamer.
Watch media, write some note, listen music, playing lite game all these can done by tablet. Only thing miss is bigger screen.
Headless Mac which mini is good start since AppleTV does not support keyboard and mouse.
I don't think ARM Mac will kill AppleTV either. It is different market. Since cable company start adapt AppleTV as their set top box, it has market to sell.
The car, Apple does not need to manufacture the car itself. They just work with car company to manufacture it. CarOS will be introduce.
Incorporate Siri, Map, other Apple service in car dashboard. Amazon recently start to add on the car.
Apple needs to doing this.. I would like to see dash cam incorporate it. I think dash cam is necessary these days.
Owning car can be headache.. Most problem is insurance. Always screw by bad driver and insurance company.
Why am I pay more premium when I have no accident last 15 years or more?
Unless you like to travel lots of place then owning car does make sense. Other wise driver-less car service is good choice. No auto insurance and maintenance cost will cover all car service and more.
...an ARM processor would cut back on virtualization solutions that the "pro" market sometimes rely on...
Absolutely this. I've been able to replace my company-provided crappy HP and Dell machines with Macs over the years because I'm able to virtualize Windows for certain corporate applications that require it, as well as for being able to compile code that targets the Windows Server environment. I also virtualize Linux frequently as my development needs require. The loss of Intel CPUs powering Macs will render virtualizing impossible, at least for Windows, and we'll have to see how well the Apple processors are able to run emulation in place of virtualization.
That would be one MORE reason why Apple will introduce the A series into the lower end MacBooks rather going product wide. The -Pro line will likely retain Intel. For the time being.
To me, from a technical standpoint, it makes little sense that the low-end MacBooks haven't already been shifted over to the A series. (Hand wringing over software, drivers, etc. from Mac loyalists are mostly nonsense rationalizations)
To me, the most likely reason it hasn't already been done is supply: can TSMC provide enough high end, 7nm A series processors to power another product line?
While I fully believe the scenario about Apple injecting the A series into low end Macs, let me offer a different possible scenario:
1) Apple keeps the Mac line alone stumbling along with Intel processors.
2) Apple creates a third line (or perhaps modifies the iPad line) to compete with the Microsoft Surface line.
Actually, from a business standpoint that makes sense: The iPad line is essentially stagnating at lower volumes with little signs that its market will grow -- while the combination of tablet and laptop as used by the Microsoft Surface opens up a whole new, untapped market for Apple.
After all: Who wants to buy one device for games and social media and another for doing homework if one device can do both EQUALLY WELL?
Comments
“Sometimes I'm right and I can be wrong“
Sure, casual consumer users content to rely on Apple's own software (Photos, Safari, Pages, etc.) wouldn't see much difference, but those who actually use Macs to make a living may very well have huge problems. Of course Apple is doing everything it can to demonstrate they no longer care about pro users anyway, so maybe it's no big deal to them. Oh well, Microsoft and many PC suppliers have made it clear they DO care about the pro market, so there are alternatives.
If the "average American driver puts more than 1,000 miles on their car every month" and you're a college student that was gifted a car for your HS graduation, and you need to pay for fuel for that 1000 miles at current prices and with the average MPG for city driving, plus annual registration and insurance, how far would that take you if you only used Lyft? I don't think it would get you anywhere close to a 1000 miles per month.
And what if I wanted to do a road trip. Would ride sharing be less expensive than owning an automobile? A less expensive option if live in a dense city with great public transportation, no reasonable place to park, and excessive fees for storing a vehicle would be to rent, but for the average American owning an automobile is going to the most sensible choice to cover a lot of distance over ground.
There are simply too many scenarios for Spam to state that public transportation and ride sharing is only for poor children.
I think the ultimate end-game of this, at least as I've envisioned it, is that there is no differentiator between Pro and non-Pro hardware. Pro becomes a collection of services running on high performance hardware in a data centre near Cupertino (or wherever else they have them). Like iCloud on steroids. They may offer an enterprise solution where they sell or rent the hardware to house in your own data centre, but (for example) Final Cut Pro becomes a thin client connecting to a workflow on an application server somewhere. (Maybe they'll sell a box you can buy from the Apple store to plug into your network, and access via your MacBook.)
Basically, I'm just extrapolating from trends in high end computing in the enterprise. Apple may never go this way at all. They might announce a 7nm ARM based processor capable of full x86 emulation with desktop performance tomorrow. Or Intel might get their act together and start shipping some new CPUs. It's really just me thinking out loud about what Apple could do assuming things stay pretty much as they are.
See, it's really easy to do this whole Tech Punditry thing. Someone should hire me.
ARM will not be anymore of a flop than the current Mac line up. It really comes down to Apple pulling their heads out of the clouds and start making Macs people want. There are many use cases where an ARM based Mac makes sense. An important one is the power savings possible, especially for always on uses. Another is the rapid changes I expect to see in AI / ML code and thus the importance of Neural Engine.
ARM on Mac is critical to Apples future because we are quickly moving past the time where the CPU is important to overall system capability.
I have ridden in smelly Ubers and Lyfts both. Maybe eventually they are voted out, but how many votes does that take? And until it happens, there's still a chance you'll get one.