Become a MacRumors Supporter for $50/year with no ads, ability to filter front page stories, and private forums.
People in the industry pretty much know what the chips are. There’s just less information directed at the general public.
By "what they are" do you mean the architectural details, or just that they're very specialized ARM architectures? If the former, could you share a link?
 
The Year-old A11 still outperforms Android’s flagship CPUs. And not by a small amount. Pretty sad...

I kinda feel bad for Apple’s software team, considering that they have absolutely zero excuses with how amazing and powerful the hardware is that they have to work with. It would be interesting to see the relationship between the hardware and software teams at Apple: The hardware team sitting back in their chairs, relaxing as they watch the software team frantically sweat and toil trying to reach the same standards of quality. Or, perhaps, it is the exact opposite, and maybe that’s why Apple is in the current situation they are now in.

Yea, well.... the Mac Mini update team's offices are in the bathroom, apparently.
 
  • Like
Reactions: Delgibbons
Why do people still bring up the "PowerPC to Intel transition" what makes people think it's gonna be that easy to go from Intel to an A12?

We bring this up as evidence that Apple knows how to do these transitions, and knows how to do them well. Apple is many things, stupid is not one of them. If they decide to transition from Intel to ARM, you can be damn sure they'll do it as seamlessly and painlessly as possible... Just like the transition from PowerPC to Intel.
 
  • Like
Reactions: bluecoast
The Year-old A11 still outperforms Android’s flagship CPUs. And not by a small amount. Pretty sad...

I kinda feel bad for Apple’s software team, considering that they have absolutely zero excuses with how amazing and powerful the hardware is that they have to work with. It would be interesting to see the relationship between the hardware and software teams at Apple: The hardware team sitting back in their chairs, relaxing as they watch the software team frantically sweat and toil trying to reach the same standards of quality. Or, perhaps, it is the exact opposite, and maybe that’s why Apple is in the current situation they are now in.
I expect a deep emoji search function will fully exploit that brutal cpu overcapacity, at some point
 
Last edited:
Finally, a true tech journalist. This is what a tech article should be, talking about tech, not just yelling about beautification and price. Anandtech makes all the tech youtubers look like bumbling idiots from tabloids trying to talk about tech.

This is where Apple innovates. They are betting on mobile since the first iPhone, and they starting to see the fruits of their labor. Nobody else will be able to match this as most other OEMs only care about short term sales and channel stuffing with little to no real innovation. Only few realize they cannot rely on off the shelf hardware (Samsung, Huawei, and Xiaomi) but they are behind.

Agreed. I think it’s clear that a major focus for iPhones will be with imagining - for the photo & video cameras, Face ID and then their application with AVR. Oh and Siri now that they have a new AI chief.
 
The purportedly increase of performance in older devices is accomplished by overclocking the CPU faster rather than any underlying core code optimization. This means that battery life will be impacted negatively for those older devices... It all plays in the favor of selling new devices.. :)
 
  • Like
Reactions: Lankyman
They are not cheaper even for Apple. It all has to due with how the chip is designed. As per the Anandtech article, the A12 probably costs about 50% more to fab than what Qualcomm sells the SD845 for. The reason they can afford to do that is because they know it is going in a very expensive device where they can recoup their costs. Basically Apple can completely ignore the cost of the chip and build it to the best specifications which may not be a valid business plan for a company who sells chips that will go in a multitude of devices.

I am not lacking about ARM to ARM comparisons. I am talking about an ARM cpu designed by apple and produced by TSMC, compared to a x86 intel cpu developed and produced by intel. High end intel cpu’ s can run as high as $1000+, and because apple can buy them in bulk, it will of course be a little cheaper.

I am sorry but no way that a cpu builder like intel, with a huge staff for design, marketing , fabs , real estate, sales representatives has lower overhead costs, than an apple design team working together with an engineering team from TSMC and have TSMC do their core businesses of producing chips.
And apple can make better deals with TSMC because they are already producing 200 million socs a year for apple, instead of the 16-20 million cpu’ s that apple buys from intel in a year.
 
We bring this up as evidence that Apple knows how to do these transitions, and knows how to do them well. Apple is many things, stupid is not one of them. If they decide to transition from Intel to ARM, you can be damn sure they'll do it as seamlessly and painlessly as possible... Just like the transition from PowerPC to Intel.
So because a company did something once that means they can do it again?
 
I guess there are a lot of pro’ s opposing the idea of arm based macs.
I see a lot of advantages if apple would do this for the entire apple ecosystem.

If you have powerful arm based macs, there would be a lot more scalability in apps between devices .
For example: the Mac based app will have a different way to control the app(keyboard/mouse) , nicer graphics, more options and faster data processing.
But the developer could also release an iPad version, but with touchscreen control, tuned down graphics, slower operation or data processing , etc without changing the source code. And they could alter the app for the iPhone , etc, etc.
Is a great possibility for better and more complex apps on your iPhone, iPad, etc.
And with the yearly hw updates of the iPhone and iPad , all the developers have to do is to add a hw profile that makes more use of the newer , more powerful hw and close the gap between the Mac app and iPad app(faster, better graphics, handling of more data streams at the same time, etc)

Could also work the other way around , though, and have apps on your Mac, that don’ t take full advantage of the more powerful Mac.

But the scalability of software between macOS and IOS , without the need of compiling, etc, sounds very interesting.
 
Last edited:
  • Like
Reactions: bluecoast
We bring this up as evidence that Apple knows how to do these transitions, and knows how to do them well. Apple is many things, stupid is not one of them. If they decide to transition from Intel to ARM, you can be damn sure they'll do it as seamlessly and painlessly as possible... Just like the transition from PowerPC to Intel.
Rosetta? Horrid experience & phase.
 
I guess there are a lot of pro’ s opposing the idea of arm based macs.
I see a lot of advantages if apple would do this for the entire apple ecosystem.

If you have powerful arm based macs, there would be a lot more scalability in apps between devices .
For example: the Mac based app will have a different way to control the app(keyboard/mouse) , nicer graphics, more options and faster data processing.
But the developer could also release an iPad version, but with touchscreen control, tuned down graphics, slower operation or data processing , etc without changing the source code. And they could alter the app for the iPhone , etc, etc.
Is a great possibility for better and more complex apps on your iPhone, iPad, etc.
And with the yearly hw updates of the iPhone and iPad , all the developers have to do is to add a hw profile that makes more use of the newer , more powerful hw and close the gap between the Mac app and iPad app(faster, better graphics, handling of more data streams at the same time, etc)

Could also work the other way around , though, and have apps on your Mac, that don’ t take full advantage of the more powerful Mac.

But the scalability of software between macOS and IOS , without the need of compiling, etc, sounds very interesting.

Great post. I’d add that to those that oppose Marzipan apps...

...Imagine the Mac in 5 years time when (let’s say) the iPad has got way more powerful and (presumably) trackpad/mouse support & the iPad’s market share has got to be higher than the Mac.

Who then is going to develop Mac apps when you could stay on iOS and hit the iPhone, iPad & the watch etc and the mythical Apple glasses?

Apple maybe. Adobe & Microsoft I guess plus all of the stalwart app devs who’ve been on the platform for decades.

But no one is going to develop new Mac apps.

The TLDR version of this is that Apple is developing marzipan to save the Mac.

If Apple doesn’t develop marzipan, the Mac in 5 years time is basically going to be a legacy platform (for most users) and a highly specialised one for pro users.

However, Apple makes mass market products. It just won’t be worth their while to keep developing the Mac platform.
 
Or maybe also use the power of the chip for Smart HDR, faster browsing in Safari due to processing javascript, rendering videos with imovie. I'm sure nobody uses the features.:rolleyes:
Well I certainly don’t, I have a large screen 4K TV for things like that, a PC/Mac and who the hell uses Safari? I gave up on Safari years ago.
 
The purportedly increase of performance in older devices is accomplished by overclocking the CPU faster rather than any underlying core code optimization.

This is false.
[doublepost=1538918796][/doublepost]
For example: the Mac based app will have a different way to control the app(keyboard/mouse) , nicer graphics, more options and faster data processing.
But the developer could also release an iPad version, but with touchscreen control, tuned down graphics, slower operation or data processing , etc without changing the source code. And they could alter the app for the iPhone , etc, etc.
Is a great possibility for better and more complex apps on your iPhone, iPad, etc.

None of this has to do with CPU architectures.
 
  • Like
Reactions: Feenician
Ok, so you think x86 macOS is compatible with ARM based IOS?

Porting your app to a different architecture isn't that big a hurdle any more. Architecture-specific optimizations in apps have become rare.

The big obstacle is designing UIs that make sense on a smartphone, a tablet, and a laptop/desktop computer. A CPU architecture isn't going to make that any easier. A framework like Marzipan might.
 
  • Like
Reactions: Feenician
Great post. I’d add that to those that oppose Marzipan apps...

...Imagine the Mac in 5 years time when (let’s say) the iPad has got way more powerful and (presumably) trackpad/mouse support & the iPad’s market share has got to be higher than the Mac.

Who then is going to develop Mac apps when you could stay on iOS and hit the iPhone, iPad & the watch etc and the mythical Apple glasses?

Apple maybe. Adobe & Microsoft I guess plus all of the stalwart app devs who’ve been on the platform for decades.

But no one is going to develop new Mac apps.

The TLDR version of this is that Apple is developing marzipan to save the Mac.

If Apple doesn’t develop marzipan, the Mac in 5 years time is basically going to be a legacy platform (for most users) and a highly specialised one for pro users.

However, Apple makes mass market products. It just won’t be worth their while to keep developing the Mac platform.

"If Apple doesn’t develop marzipan, the Mac in 5 years time is basically going to be a legacy platform (for most users) and a highly specialised one for pro users."

As the iPad Pro becomes more powerful and takes on more tasks designed for the Mac, the future of the Mac is as a specialized device for Pro users. I could be wrong, but I still Marzipan as nothing more than a way to run iOS / iPad apps on your Mac.
 
These Apple Ax chips have been amazing for years.

One thing I don't understand though - many speed tests on youtube actually show modern android phones (i.e. S9) launching apps faster than the iPhone.

Any idea how this is possible? Is iOS that bloated? It seems like Android should have no chance given much slower cpu's and the overhead of java.
 
I don't understand why a smartphone needs so much computing power. Is it because of mobile gaming?
[doublepost=1538935182][/doublepost]
I guess we will see soon apple laptops with A13 ??
That will end usage of apple hardware in IT (except for ios devs, of course).
 
So because a company did something once that means they can do it again?

Uh, they have done it, let's see, how many times:
- Mac: M68K to PPC to Intel (and two 32 bit to 64 bit, once PPC, once Intel)
- iPhone: 32 bit to 64 bit (now 64 bit with PAC)
- aWatch: 32-bit to 64-bit (by using ILP32 and bitcode they didn't even need a developer rewrite)

Not to mention ISA-like changes you aren't even aware of, like GPU changes (upgrades from the same company, and nV vs AMD), or ISA-like changes (like dynamic compilation of JS).

Yeah, I think they kinda know the in's and out's of pretty much every permutation of moving ISAs...
 
"If Apple doesn’t develop marzipan, the Mac in 5 years time is basically going to be a legacy platform (for most users) and a highly specialised one for pro users."

As the iPad Pro becomes more powerful and takes on more tasks designed for the Mac, the future of the Mac is as a specialized device for Pro users. I could be wrong, but I still Marzipan as nothing more than a way to run iOS / iPad apps on your Mac.

I totally agree, that’s going to be a key use for the Mac. As Jobs memorable said, it’ll be a ‘truck’.

My point is is that I feel that it’s goong to be hard for Apple to justify continuing investment in the Mac unless it shares the ARM tech stack that it’s built out wit the iPhone.

So sure, I suspect that in 5 (ok maybe 5-10 is probably more realistic for what I’m about to say) years time, pros (creatives, scientific researchers & business people) will be the prime users of the Mac - but they’ll be using pro software on ARM macs.
 
Not sure exactly what you mean by this - a hackintosh is basically non-Apple hardware running MacOS. I guess the tablet equivalent concept would be to install iOS on a non-Apple tablet. Constructing a functioning hackintosh requires matching Apple Mac hardware closely enough to run MacOS on it. That is no simple task even with Intel based CPUs. To install iOS on a non-Apple tablet would require at minimum that the tablet be ARM processor based. If by some miracle you could get iOS to install on such a tablet, it would still accomplish nothing toward comparing iOS on ARM with iOS on an Intel platform. You would only be comparing different ARM processors running the same OS. This is the issue which is rather muddied when saying that "A12 outperforms a Skylake CPU". Perhaps that is true using certain benchmark measurements, but it doesn't reveal much about how iOS on the A12 compares with Windows on Skylake (or any other Intel platform). If iOS could be installed on Intel hardware, and Windows on Apple ARM hardware, there would be more ground for comparisons in terms of the user experience. The primary motivation for building a hackintosh is to get MacOS to run on cheaper non-Apple hardware, and that's only possible since Macs and PCs both run on the Intel platform. I don't see that happening anytime soon with tablets. A more reasonable comparison, and one which is already done, is to compare iPhones with Android phones, as they both are ARM based systems.

A Hackintosh is hardware running MacOS without Apple's approval/blessing. So an iPad would be included.
 
Register on MacRumors! This sidebar will go away, and you'll see fewer ads.