I continue to be excited by Apple's development of their arm chipset. I am also excited by the iPad and continuing maturation process of iOS. Clearly it is still not OS X, but let's be clear, it doesn't need to be. I think desktops with big screens will always be needed for certain tasks. For about 80% of what I do on a daily basis, however, the iPad already fits my needs. For some that number will be lower and for some it will be higher. But no more than 10 in either direction I suspect. As the chipset and the OS improves the number will move up slightly.
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.
iPad Pro's A9X Chip Rivals Performance of Intel's Low-End Core M Processors
- Thread starter MacRumors
- Start date
- Sort by reaction score
People really think Apple puts its best/most powerful ARM chips in iOS devices? That Apple is going to put all of its cards on the table in the form of a tablet device? Really?
Nothing stopping Apple producing a new ARM chip to rival Intel, but they're certainly not going to put such a chip in an iOS device.
Just because you can't buy a device right now with a very powerful ARM chip doesn't mean that Apple doesn't have such a chip in its labs.
Crazy if you think Apple can't match Intel.
When Apple releases a car, it will be a car that can match any other car on the market. It isn't going to launch a car from the early part of the twentieth century and then play catch-up for years.
Intel/carmakers have nothing that Apple can't emulate.
Nothing stopping Apple producing a new ARM chip to rival Intel, but they're certainly not going to put such a chip in an iOS device.
Just because you can't buy a device right now with a very powerful ARM chip doesn't mean that Apple doesn't have such a chip in its labs.
Crazy if you think Apple can't match Intel.
When Apple releases a car, it will be a car that can match any other car on the market. It isn't going to launch a car from the early part of the twentieth century and then play catch-up for years.
Intel/carmakers have nothing that Apple can't emulate.
There is no battle. Or it was lost a long time ago. Microsoft failed to establish a mobile OS presence and lost it's market position in the mobile space to Android. Without a viable phone OS there is no way to attract developers for tablets anymore. The Surface Pro is the same old non-mobile Tablet-PC technology, trying to compete with real tablets. Core M is merely the first Intel CPU to run without a fan, but it still consumes vast amounts of energy and is way too expensive. Another failing company. Intel and Microsoft both will now always remain PC companies and slowly vanish as the Post-PC era dawns. The Commodore and Amiga of our time.
Right and Jony Ive explicitly said Apple pencil was for making marks not UI navigation. Try harder.
Not so fast my friend. Let's see that iPro running engineering and scientific programs. Can it run my AutoCad and stress analysis?
In time it may be able to get their and I hope it does but for the time being it is nothing more than a large iPad.
There is no point in holding back production-ready technology. What's a bleeding edge ARM CPU today, is standard next year and almost obsolete the year after. And ARM chips are already superior on a performance per watt metric.
Except for the whole car part. Try to ride on your iPhone, if you think it's no different from a vehicle. Apple needs to partner with at least one car company or with a bunch of big suppliers.
I partially agree.
It can support Bluetooth peripherals.
What the A9x can't do is support the peripheral interfaces that are used on "real" computers.
There is no NVMe support, SAS, SATA, etc
There is no support for Thunderbolt or PCIe or an external GPU.
People are talking about performance and power consumption.
You cannot compare the performance of an laptop processor and take power into consideration because the A9x does not have the same number of peripheral interfaces that also draw power.
ARM != Intel
[doublepost=1453501513][/doublepost]
Tim Cook stated that there are two distinct ecosystems.
iOS is stripped down OSX with a different gui.
It's BSD under the hood.
People bring this up over and over again.
Multiple companies have found their way right out of business trying to compete with Intel at what they do best.
Here is your list of companies that just couldn't keep up.
AIM - Apple, IBM, Motorola with PPC
Motorola with 68K and 88K processors
Sun with SPARC
SGI with MIPS
AMD falling behind with x86.
While you might be able to run something like a ChromeBook, etc...
It will be a long time before you can put an ARM64 in the equivalent of a MacBook Pro 15 and have it competitive.
Yeah, the Intel M class processors are lowly, but that's not where Intel makes it's cash.
Also Apple is a very close partner with Intel and knows the roadmaps for mobile CPUs.
The MacBook is a very specific vertical market laptop.
The MacBook Air and the Pro are what you want to compare an ARM to and lets face it an ARM64 is not i5, let alone an i7. Forget about a quad core i7.
Last edited:
I'm not seeing any GPU comparison in that article--did I miss it?
It's interesting, because BareFeats did a similar comparison a couple months ago.
Their CPU results (based entirely on Geekbench, so obviously not as comprehensive as this comparison) were similar--the A9X was actually challenging the MacBook Air in single-core scores, while it was in the ballpark of the 12" MacBook in multi-core. So the result would be similar--it's getting up in the range of the previous-gen Broadwell Core M in speed.
But in that same BareFeats article, the graphics test results were astounding. Looking at offscreen rendering in GFXBench, the iPad pro was just as fast as a current-gen 15" MBP with a 2GB Radeon R9 M370X. It was 3-4 times the speed of the MacBook and around double the MacBook Pro 13" with Intel HD 6100.
If those results are even remotely accurate, that means that Apple's fanless, 5W SOC (AnandTech claims that the entire iPad Pro, sans display, draws around 5W, on par with a Broadwell Core M) can challenge a relatively recent mid-tier dedicated mobile GPU that by itself has a TDP of at least 15W. Which is pretty astounding, no matter how you look at it.
Depending on use case, that makes the A9X a pretty formidable SOC, and given the GPU performance increase curve that Apple A-series and other ARM CPUs have been following recently, might make the next generation or two particularly interesting.
Aside: It was very difficult to work out a benchmark for comparison, but in absolute graphics performance it looks like the current A8-based AppleTV has roughly equivalent performance to a PS3. Yes, that's a very old game system at this point, but it's pretty impressive given that the ATV is a little $150 fanless media brick using a previous-generation phone SOC.
It's interesting, because BareFeats did a similar comparison a couple months ago.
Their CPU results (based entirely on Geekbench, so obviously not as comprehensive as this comparison) were similar--the A9X was actually challenging the MacBook Air in single-core scores, while it was in the ballpark of the 12" MacBook in multi-core. So the result would be similar--it's getting up in the range of the previous-gen Broadwell Core M in speed.
But in that same BareFeats article, the graphics test results were astounding. Looking at offscreen rendering in GFXBench, the iPad pro was just as fast as a current-gen 15" MBP with a 2GB Radeon R9 M370X. It was 3-4 times the speed of the MacBook and around double the MacBook Pro 13" with Intel HD 6100.
If those results are even remotely accurate, that means that Apple's fanless, 5W SOC (AnandTech claims that the entire iPad Pro, sans display, draws around 5W, on par with a Broadwell Core M) can challenge a relatively recent mid-tier dedicated mobile GPU that by itself has a TDP of at least 15W. Which is pretty astounding, no matter how you look at it.
Depending on use case, that makes the A9X a pretty formidable SOC, and given the GPU performance increase curve that Apple A-series and other ARM CPUs have been following recently, might make the next generation or two particularly interesting.
Aside: It was very difficult to work out a benchmark for comparison, but in absolute graphics performance it looks like the current A8-based AppleTV has roughly equivalent performance to a PS3. Yes, that's a very old game system at this point, but it's pretty impressive given that the ATV is a little $150 fanless media brick using a previous-generation phone SOC.
Last edited:
They better get there before Microsoft.... oh wait.
[doublepost=1453502467][/doublepost]
Unless you have something I don't, no iOS or touch on MacBook.
Yes it can, it may however not be the best device for that purpose. But that doesn't matter, because AutoCAD is not that kind of market shaping killer application.
And the iPhone 6+ is nothing more than a large iPhone 6, but it's build on a sound hardware and software foundation and supported by a healthy developer community and the Surface Pro is not.
I share the same sentiments that the iPad Pro running OS X would be something exquisite. But I do think the iPad line is solid even without the full OS. iOS is a pretty solid system even without the main draws of a full OS, but what really makes the iPad shine IMHO is the apps.
If an app, a well coded app, is on the iPad, and I use it daily and/or exclusively, then I can see the iPad replacing my day to day Retina Macbook. As of now, that's not the case.
The iPad Pro is delicious. Nice to use, nice to work on, and extremely nice to carry around and pull out and is even faster (at least by my uses) than my 12" Retina. I just need like . . . . three more apps to be fully functional on the iPad, and a little bit more flexibility with how apps interact with each other.
If an app, a well coded app, is on the iPad, and I use it daily and/or exclusively, then I can see the iPad replacing my day to day Retina Macbook. As of now, that's not the case.
The iPad Pro is delicious. Nice to use, nice to work on, and extremely nice to carry around and pull out and is even faster (at least by my uses) than my 12" Retina. I just need like . . . . three more apps to be fully functional on the iPad, and a little bit more flexibility with how apps interact with each other.
Given that an A9 or A10 is substantially lower cost than any Intel chip, I would guess the first ARM Mac will be dual core and have outsized memory to offset.
The biggest problem I see is it will be mostly a cloud software confuser. Limited local software.
On the other hand it will support the latest LTE, Bluetooth, wifi.
The biggest problem I see is it will be mostly a cloud software confuser. Limited local software.
On the other hand it will support the latest LTE, Bluetooth, wifi.
Don't forget that Apple hasn't attempted to release a high-performance CPU yet. What would an A-series CPU with the power budget of a 15" MBP or iMac look like?
We don't know. It's conceivable that Apple is *already* competitive with some of the higher performance CPUs from Intel with a comparable power budget.
Intel's CPUs will always be better, they just have more power, their GPUs suck but that is why you buy an Nvidia GPU.
I've been programming ARM chips for 30 years - I worked on the original OSs - Arthur and RISC OS. Even in 1986 the ARM chip was very fast, even though it was only running at 8 MHz (that's not a misprint - I mean 8 million cycles per second, not the modern 3-4 billion) so I am quite surprised that a modern ARM chip can't wipe the floor with a CISC chip. I think the reason is that so far ARM development has been for mobile devices and so power usage is paramount, with a desktop, or even laptop, there is mains power or larger batteries, that could drive a much more powerful ARM chip.
Imo this shows how underperforming the Macbook is, not how powerful the iPad is. Being the same price as the Macbook Pro and having the processing power of an iPad thats $500 cheaper makes the Macbook a tough buy.
First of all, the A9X isn't going to be making it's way into a Mac. It'll be an A10, or A11. The nice part is that with the Mac App Store, USB C, and the fact that OS X was built to be portable things are looking pretty good.
Keep in mind, the iPad Pro can already edit 4K pretty well. there are very few consumers who need more than the iPad Pro can already push out. And Apple is in a very different place with Mac software developers than it was when they switched from OS 9 to OS X and from PPC to Intel. They might write an x86 emulator, but realistically I imagine that they won't. They'll just require that all apps submitted to the Mac App Store be compiled for ARM.
What were the big hiccups in the last few transitions? Quark, they don't matter anymore, Microsoft, they already have all their office apps on iOS with Xcode and have hopefully been working on Office for Mac in Xcode now, and Adobe ... that's what selling Intel Macs for the next two years will be for.
However even if Adobe does blow off Apple's ARM adventures, Affinity Designer and Photo, (and likely Publisher by then) can be heavily promoted by Apple in the mean time to pressure Adobe to get their act together. But while back in the early 2000s Apple was focused on home, education and a few niches like desktop publishing, video and music. Apple is now a HUGE brand in education and home, still huge in music, growing in enterprise, doesn't give a damn about desktop publishing. All the iLife and iWork apps will be ready on launch day, Logic, MainStage, FCPX, and Motion will be available shortly after. And within six months nearly all major third party Mac App Store apps should be ready to go.
They'd start with the low end, and then start making higher power variants for the MacBook Pro and 5K iMac. The nice thing is that since the ARM SOCs are so power efficient they could pack in four of them (more likely just make more cores) at which point sure you might only be able to do that filter in Photoshop or FCP so fast, but you'll be able to do a lot of them at once. And realistically that's what people care about. What they don't want to see is their overall system slowing down. They don't mind reading some email while something else renders. (Well some people do but most people I think don't.) Plus if software designers all know how fast this one year's Macs will render something, that means that they can ensure that whatever they're doing will feel fast on that generation's hardware.
The good part is that it means instead of getting random updates whenever Intel feels like it, we'll get yearly updates with processors as fast as they could make them. It'll also mean that this year's MacBook Pro processor will be next year's MacBook Processor. It'll probably mean much less BTO options for Macs.
Keep in mind, the iPad Pro can already edit 4K pretty well. there are very few consumers who need more than the iPad Pro can already push out. And Apple is in a very different place with Mac software developers than it was when they switched from OS 9 to OS X and from PPC to Intel. They might write an x86 emulator, but realistically I imagine that they won't. They'll just require that all apps submitted to the Mac App Store be compiled for ARM.
What were the big hiccups in the last few transitions? Quark, they don't matter anymore, Microsoft, they already have all their office apps on iOS with Xcode and have hopefully been working on Office for Mac in Xcode now, and Adobe ... that's what selling Intel Macs for the next two years will be for.
However even if Adobe does blow off Apple's ARM adventures, Affinity Designer and Photo, (and likely Publisher by then) can be heavily promoted by Apple in the mean time to pressure Adobe to get their act together. But while back in the early 2000s Apple was focused on home, education and a few niches like desktop publishing, video and music. Apple is now a HUGE brand in education and home, still huge in music, growing in enterprise, doesn't give a damn about desktop publishing. All the iLife and iWork apps will be ready on launch day, Logic, MainStage, FCPX, and Motion will be available shortly after. And within six months nearly all major third party Mac App Store apps should be ready to go.
They'd start with the low end, and then start making higher power variants for the MacBook Pro and 5K iMac. The nice thing is that since the ARM SOCs are so power efficient they could pack in four of them (more likely just make more cores) at which point sure you might only be able to do that filter in Photoshop or FCP so fast, but you'll be able to do a lot of them at once. And realistically that's what people care about. What they don't want to see is their overall system slowing down. They don't mind reading some email while something else renders. (Well some people do but most people I think don't.) Plus if software designers all know how fast this one year's Macs will render something, that means that they can ensure that whatever they're doing will feel fast on that generation's hardware.
The good part is that it means instead of getting random updates whenever Intel feels like it, we'll get yearly updates with processors as fast as they could make them. It'll also mean that this year's MacBook Pro processor will be next year's MacBook Processor. It'll probably mean much less BTO options for Macs.
This argument needs to be retired.
Didn't Steve say "who wants a stylus?" when he was showing off the first generation iPhone?
Apple was simply selling a new product and introducing multi-touch technology to the masses. The only thing that compared at the time were devices like Palm Pilots, which required a stylus in order to work. Also, the styluses of 2007 and the wacoms and apple pencils of today are in totally different leagues.
And Steve Jobs wasn't afraid to change his mind:
Last edited:
Yeah, I question the comprehension skills of people who constantly cite the stylus comment in relation to the stylus on the iPP. Completely different use case, if Jobs was alive today he'd still maintain capacitive screens + finger are infinitely better for general UI interactions on a phone/tablet than a stylus, which is hard to disagree with. For accurate, precise activities, stylus >>>> finger.
Of course, Apple change their tune with the wind, whatever is most appropriate to the current line of products they're selling, the company's stance will reflect it. I don't understand why people have such a problem with this, Apple are a money making enterprise and represent the pinnacle of capitalism and effective marketing. If you to look to them for moral guidance, consistency and truthfulness....well, I really don't know what to suggest beyond seek help.
Of course, Apple change their tune with the wind, whatever is most appropriate to the current line of products they're selling, the company's stance will reflect it. I don't understand why people have such a problem with this, Apple are a money making enterprise and represent the pinnacle of capitalism and effective marketing. If you to look to them for moral guidance, consistency and truthfulness....well, I really don't know what to suggest beyond seek help.
I am not so sure about that. The Love/hate relationship between Adobe and Apple isnt working, plus Adobe has a lot of customers on the Windows platform. What happens if Apple switched over to ARM and no Adobe tools will work? Would Adobe invest in rewriting their complete Suite? So they have to keep up 3 versions (ARM OSX, OSX, Windows). I think it would kill the iMac instantly for any designer who using Adobe seriously.
With that said, Adobe has Cuda hungry tools witch Apple can not deliver with there AMD setup. Its a shame, but these days, the Apple machines are not your best choice anymore for creatives who need some Nvidia muscle or a more high end CPU like an i7 6 or 8 core. Those fast clocked CPU's are very populair in high power PC's and dont need an expensive Xeon / ECC memory. It looks like i have to leave my loved OSX in my next hardware update this year and move over to a power machine with Windows 10.
If they go ARM all the way, I think they will focus on the normal mainstream user and start with a laptop or something. The home (fun) user for internet, music, netflix, etc. It could be an very interesting time but i tink they kill there last bit of connection with their creatives. And maybe, the tablet / laptop migration will enter a new fase with very fast ARM chips.
Intel can use some competition!
You may be right, but just to note, in every single one of those examples the competitor didn't start in a strong market position, and the competition was in specific areas with limited markets, and they often did compete for quite a while.
ARM may well never match desktop-class or heavy-laptop-class x86 chips. But the scales are not the same:
Today--as in 4Q 2015--there were a total of about 75.7 million PCs shipped worldwide (of which Apple made 5.6 million), if Gartner is correct. This number has been steadily decreasing for a while. Given that Intel has about an 80% market share, that means they shipped around 60 million Core-type CPUs in 4Q 2015. Unless something dramatic changes, that number will be lower this year.
Apple hasn't announced results for the most recent quarter yet, but last year for the same three month period, Apple--alone--sold 74.5 million iPhones and 21.1 million iPads. Ignoring iPod touch and ATV sales, that's 94.6 million computing devices sold in the year-ago holiday quarter running A-series CPUs. The numbers might be somewhat lower in the Q4 2015 quarter, but they're probably at least in the general ballpark.
Assuming an unrealistically high iPhone marketshare of 20% and iPad share of 33% (both are actually lower) that puts global non-Apple ARM-based phone sales in the same quarter at 300 million and non-Apple ARM-based tablets somewhere around 40 million. Those numbers are, thus far, still increasing.
So, in absolute number of CPUs shipped, in a good holiday quarter Apple--by itself--is making and shipping something like 50% more A-series CPUs than Intel is selling x86 PC CPUs. They are selling something like 25% more A-series CPUs than all x86 PC CPUs, period. Non-Apple ARM CPU shipments are now easily in the range of 4 times higher than total x86 CPU shipments. x86 CPUs, of course, cost tens if not hundreds of times more each than ARM CPUs, so Intel is doing pretty well on revenue, but that's not necessarily a long-term advantage, and in absolute numbers the entire x86 architecture is now a minority platform by a wide margin.
So comparing ARM and A-series CPUs today to PPC, 68K, SPARC (niche high-end workstations and servers), MIPS (same), and even AMD with their ~20% marketshare using a compatible instruction set to ARM today is a tough sell given that with the possible exception of 68K briefly nobody has ever had the kind of sales volumes ARM has today, both in absolute terms and relative market share.
The bottom line is, Apple is already selling more A-series CPUs than Intel is selling x86 CPUs. They are not an underdog with a niche market anymore, they already have a dominant market position.
This probably has a bit to do with the fact that Apple could literally buy Intel at market price with their cash on hand and still have a few billion left over.
I there with you..be interesting to see how this does long term...I have the iPad air 2..thats the most I will spend for
an Ipad..I doest''t make sense FOR ME...with it's limited capabilities
to buy the big iPad and not have OS X does not make sense..if this could run OS X...SWEET
I messed with the the touchscreen Windoz machines..they suck but for the $$ with the touchscreen will win most
I want apple to have that same thing..BIG IPad, with OS X come on Apple give the people what they want!