Many of us old beards on AI have been saying this for years. ARM Marklar makes sense. I don’t see Intel based Macs going away but I see them being primarily used for the high end.
In other words, I see Apple positioning the iPad Pro as the "ARM-based Mac".
We have had this debate a few times on this forum, and while "I" think it would be great to have an ARM "Mac" (I wouldn't call it iOS which is touch optimized), there are also reasons why Apple would not take this approach - Customer / market confusion. Even if called something other than a Mac, it will cause some issues with two laptop lines. Programs work on one, not another. - Potential Margin loss. While an ARM "Mac" should be cheaper - I wouldn't think Apple would sell it though below an iPad Pro - if margin % is similar to an Intel Mac, then Apple is still losing total margin dollars (30% on $799 vs. 30% on $1399). - Conflict with iPad. Apple is doubling down on iPads, as seen this past year. The iPad Pro's occupy the price point where these ARM Macs would play. How to reconcile the two? - A relatively flat "truck" (PC/Mac) market vs. expanding car market. PC market as a whole is shrinking. Mac market is holding steady mostly (ups and downs depending on year). Mobile devices (smartphones and tablets) are expanding. Is it better to focus on "more future"? Investment would still be in the Mac, but working on the Intel line.
So, while initially I was of the view of separating the Mac line into the consumer (ARM based) and pro (Intel based), I just don't think Apple will do that.
Is it possible that Apple could go ARM all the way? Move the entire Mac line to ARM? I am not a chip & OS detailed guy, so this is simply high level thinking/questions. - If the A-Series line can get enough processing power - perhaps with two (or more) chips per machine & with high clock rate - then would it be possible to get "good" performance on x86 compiled programs (a new Rosetta)? Benchmarks on the A11 show it outperforming Intel Core i5 unit, but many will say that is not a true apples-to-apples comparison. But is it getting there? - What is the difficulty in app developers porting to ARM? Does Apple perhaps have some tools that could help? Perhaps Mac App Store apps are already mostly there. If you could get Microsoft and Adobe onboard, is that enough? - In 2018, I believe Intel will make Thunderbolt available for other platforms/chipmakers to use (AMD, Qualcomm, ... and Apple). Perhaps this was a key item holding Apple back?
As for why Apple would do this? Not only would there be lower component costs (even 2x A-Series would be less - possibly much less - than Intel chips), but Apple could add native hardware functions as we see on the iOS devices. This dedicated silicon approach means improved functionality compared with other PC's, without having to rely on CPU/GPU brute force improvements.
Apple replacing Intel with ARM? Must be the first time ever we hear about that
What you missing here is stats on how Apple A-chips started to perform recently. Combining that with the fact that Intel is kind of stagnant and Apple doing its own R&D on their own A-chips, I think they will get there eventually. They are not happy with the fact that some company have their IP in Apple's products which they can sell to other companies too.
Does this mean going back to the days of most programs being incompatible with Mac like with the Power PC? No bootcamp anymore?
who cares about Window's programs? Very few.
If Apple releases an A-chip desktop for $199-$299 they will sell a boat load of them. And devs will be forced to bring over their Apps to A-chip desktop platform.
They're going to release a full desktop for less than their cheapest iPad? Your math is off.
most desktops now are in that range (no keyboard no mouse no monitor... ZeroClients).
But apple would never sell another device without an integrated monitor. So figure a 21" $699 iMac.
So, I was in an Apple Store this week. Number of employees I saw at desktop computers? Zero. Number with iPads or iPhones ... at least 50 percent. They were all working hard, too, and iPhone 8's were in stock at the store, although maybe not in the color or memory you wanted. Anyway, back to my point. People working hard on iPads.
The reasons why Apple is going to do this are obvious. They would control the entire stack.
Currently you don't just license an Intel Processor you have to license the motherboard as well. The cost is not trivial.
When you look at Mac hardware you see that Apple has essentially made them appliances. There are rarely PCI-Express slots, there are no M2 slots and other forms of connectivity. This makes it easy to migrate to another architecture because there are no legacy internal components to consider.
ARM vs Intel architecture from an application standard. As one poster said this is closer to the 68x - PPC transition than PPC to Intel and todays tools (LLVM and more) make the coding changes manageable. We'll know if Apple plans to make the transition because we'll start seeing AppKit and ARM get cozy in code.
As far as performance ARM has beefed up their architecture. You can now have up to 8 CPU in a cluster and you can cluster 32 of these Octoclusters together. Want more speed? Add more CPU.
Mobile platforms have matured. There's only so much better the displays are going to get now that we have Super Retina and OLED. There's not going to be a big boom in sales based off of evolutionary updates.
It's time for the industry to take a look back at the desktop and focus on low power, near invisible options that provide, home automation, monitoring and more.
The ARM chip will probably become standard issue on Macs to handle things like Face ID, long before it takes over main processing duties for macOS apps.
Many of us old beards on AI have been saying this for years. ARM Marklar makes sense. I don’t see Intel based Macs going away but I see them being primarily used for the high end.
In other words, I see Apple positioning the iPad Pro as the "ARM-based Mac".
I think iOS will remain where it is, touch interface devices. Macs without touch interfaces will be of two types with and without intel, the mass market MacBooks and iMacs with ARM and Mac Pros, MacBook Pros and iMac Pros with either Intel or perhaps both ARM and Intel.
Did Apple learn from the PowerPC fiasco? Full (as much as possible) compatibility with the rest (95%) of the world (read, Windows) is a must. That requires Intel x86 inside Mac. Otherwise, we will be forced to switch to Windows. A shame for all!
Did Apple learn from the PowerPC fiasco? Full (as much as possible) compatibility with the rest (95%) of the world (read, Windows) is a must. That requires Intel x86 inside Mac. Otherwise, we will be forced to switch to Windows. A shame for all!
sorry bro.
but windozes in no where as important as it was in 1990.
Did Apple learn from the PowerPC fiasco? Full (as much as possible) compatibility with the rest (95%) of the world (read, Windows) is a must. That requires Intel x86 inside Mac. Otherwise, we will be forced to switch to Windows. A shame for all!
sorry bro.
but windozes in no where as important as it was in 1990.
As of a year ago...
Now add iOS to that graph........
I thought the discussion was about desktop/laptop and not smartphones. Adding iOS or Android would be out-of-place.
I don't think there's much doubt that Windows still rules the biz sector but I agree with Sog35 it's simply not that important to Apple's future that they ship computers with Intel Hardware.
Much as all Apple fans want MacOS on Apple ARM, ain't happening in near future. Switch possible in late 2020 or early 2021 if progress on ARM performance with low power matrix continue to improve. Intel while improving performance on every generation of X86 chips, also keeping TDP lower.
Does this mean going back to the days of most programs being incompatible with Mac like with the Power PC? No bootcamp anymore?
who cares about Window's programs? Very few.
If Apple releases an A-chip desktop for $199-$299 they will sell a boat load of them. And devs will be forced to bring over their Apps to A-chip desktop platform.
Apple would never bother with a laptop or computer as cheap as $299 .
Although I would love for therm to produce a iOSBook - one at $250 with polycarbonate body for young kids and a premium on at 899 and up,
Thought this wasn’t possible? Something about architecture?
This would be awesome and put Macs even further into the future. I read somewhere that the A11 is faster than MacBook Pros in some tasks.
I would assume Apple would start using their chips in MacBooks first but it wouldn’t make sense to have their laptop line incompatible with their desktop line so maybe we’re 5-10 years away when A-Series chips can compete with high end iMacs. This way they can upgrade the entire computer line with A-Series chips all at once.
1. What has Radio Patents got to do with Broadcom? If you mean WiFi chipset? Those standard are done by IEEE and are patents free. Apple has already shipped two generation of WiFi chip, the W1 and the new W2 in Apple Watch. Its own implementation of Bluetooth 4.2 and WiFi 802.11n. Needless to say, they are likely working on W3 with Bluetooth 5.0 and 802.11ax.
2. If you meant patents of LTE as in Qualcomm, well Apple is already having a battle with them anyway. Lets redefine what SEPs are and their fair price by court.
3. No one has the answer yet what the LTE baseband were on Apple Watch S3 LTE. It was suppose to be Intel according to rumors. But i have always doubt that because Intel simply dont have something so small and so advance ready. It doesn't support CDMA, which basically rule out anything Qualcomm. It doesn't support GSM, but it could be disabled. I am going to guess the LTE in Apple Watch is Apple's first implementation of LTE Baseband modem.
4. We wont get an Apple Modem in 2018. There is a different between Modem inside watch and iPhone. Next year it is likely to be 100% Intel Modem 7560, supporting 1Gbps speed, LAA, basically the same as Qualcomm X16. Fabbed with Intel's 14nm. Apple dont change modem often, so 7560 will likely stay until 2019. By 2020, Apple will have its own 5G modem.
5. WiFi, Small Radio chipset are perfect fit for Global Foundry SOI technology, 22FDX now and 12FDX coming next year. I am hoping Apple will diversify its Fab usage to GF.
6. In typical Apple fashion, It is very likely Apple thinks 4.5 / 5G ( 1Gbps ) has reached a point where speed is no longer an issue. LAA / Small Cells, more band and lower latency are all done in spec already. Apple are now working its baseband modem because it can reduce cost, and more importantly minimize battery usage.
7. It is highly unlikely Apple decide to Drop Intel from Mac when they need Intel most on the modem side of things. Given iPhone is literally the whole of Apple, Tim would be foolish to do so.
8. Apple now has AMD to keep Intel honest. I seriously doubt if Intel does not lower its price to large clients like Apple.
9. Intel has already sent out words of warning on Emulation of x86 is like walking on a patent minefield. It truly is stupid, but Mircosoft are doing it too, Apple is not in a hurry. They can wait.
10. I believe Apple making their own Metal GPU makes much more sense, it include its AI and Neutral Engine inside, which is lacking on the Mac. There is still so much innovation inside GPGPU untapped, and Apple cant do much if those aren't opened up.
I think we will be seeing an iOS desktop very soon..
Hopefully in 5 years Apple is almost 100% vertically integraded and does not have to rely on any vendor to provide parts.
I hope not!!! That is an IOS desktop, such a device would be too likited. Now a Mac derived OS that could run iOS apps in a Window would be nice.
You see i actually want to see an ARM powered laptop. I just dont want to compromise with respect to all the capability Mac OS has. Mac OS is a very fine OS with nothing that really competes.
As for the other chips it has been known for a long time that Apple wants its own baseband solution. They will work on this because integration into the SoC makes sense.
Much as all Apple fans want MacOS on Apple ARM, ain't happening in near future. Switch possible in late 2020 or early 2021 if progress on ARM performance with low power matrix continue to improve. Intel while improving performance on every generation of X86 chips, also keeping TDP lower.
Most (very, very far from "all") Apple fans couldn't care less what chip is inside their Mac. Actually most Apple fans don't even own Macs probably.
Did Apple learn from the PowerPC fiasco? Full (as much as possible) compatibility with the rest (95%) of the world (read, Windows) is a must. That requires Intel x86 inside Mac. Otherwise, we will be forced to switch to Windows. A shame for all!
sorry bro.
but windozes in no where as important as it was in 1990.
As of a year ago...
Now add iOS to that graph........
I thought the discussion was about desktop/laptop and not smartphones. Adding iOS or Android would be out-of-place.
No it isn't.
I'm talking about an A-chip desktop that runs iOS and modified OSX.
Hardly anyone uses windoes now days, execept at work.
And gaming. And higher education. Surface devices are becoming quite popular on college & university campuses.
Much as all Apple fans want MacOS on Apple ARM, ain't happening in near future. Switch possible in late 2020 or early 2021 if progress on ARM performance with low power matrix continue to improve. Intel while improving performance on every generation of X86 chips, also keeping TDP lower.
Most (very, very far from "all") Apple fans couldn't care less what chip is inside their Mac. Actually most Apple fans don't even own Macs probably.
You can tell the older Mac users by how stubborn they cling to Intel. I remember the days of Macs getting their PPC clocks cleaned by Intel computers half the cost. There is a fear that moving away from Intel will lead us back there but times have changed. What we see in mobile devices with regard to integration other co-processors is going to come to Mac OS. You're going to see specialized AI chips, chips to handle realtime video and fast GPU to do heavy compute.
Desktops have stagnated not because people love mobile devices so much but because all the innovation got sucked out the platform when everything went Wintel. Neither Intel nor MSFT would take any real risk thus we have been subjected to a decade of evolutionary "ho hum" upgrades.
I don't know if Apple is going to play disruptor but they are uniquely positioned to change how desktop computing is done for the next generation.
Comments
We have had this debate a few times on this forum, and while "I" think it would be great to have an ARM "Mac" (I wouldn't call it iOS which is touch optimized), there are also reasons why Apple would not take this approach
- Customer / market confusion. Even if called something other than a Mac, it will cause some issues with two laptop lines. Programs work on one, not another.
- Potential Margin loss. While an ARM "Mac" should be cheaper - I wouldn't think Apple would sell it though below an iPad Pro - if margin % is similar to an Intel Mac, then Apple is still losing total margin dollars (30% on $799 vs. 30% on $1399).
- Conflict with iPad. Apple is doubling down on iPads, as seen this past year. The iPad Pro's occupy the price point where these ARM Macs would play. How to reconcile the two?
- A relatively flat "truck" (PC/Mac) market vs. expanding car market. PC market as a whole is shrinking. Mac market is holding steady mostly (ups and downs depending on year). Mobile devices (smartphones and tablets) are expanding. Is it better to focus on "more future"? Investment would still be in the Mac, but working on the Intel line.
So, while initially I was of the view of separating the Mac line into the consumer (ARM based) and pro (Intel based), I just don't think Apple will do that.
Is it possible that Apple could go ARM all the way? Move the entire Mac line to ARM? I am not a chip & OS detailed guy, so this is simply high level thinking/questions.
- If the A-Series line can get enough processing power - perhaps with two (or more) chips per machine & with high clock rate - then would it be possible to get "good" performance on x86 compiled programs (a new Rosetta)? Benchmarks on the A11 show it outperforming Intel Core i5 unit, but many will say that is not a true apples-to-apples comparison. But is it getting there?
- What is the difficulty in app developers porting to ARM? Does Apple perhaps have some tools that could help? Perhaps Mac App Store apps are already mostly there. If you could get Microsoft and Adobe onboard, is that enough?
- In 2018, I believe Intel will make Thunderbolt available for other platforms/chipmakers to use (AMD, Qualcomm, ... and Apple). Perhaps this was a key item holding Apple back?
As for why Apple would do this? Not only would there be lower component costs (even 2x A-Series would be less - possibly much less - than Intel chips), but Apple could add native hardware functions as we see on the iOS devices. This dedicated silicon approach means improved functionality compared with other PC's, without having to rely on CPU/GPU brute force improvements.
most desktops now are in that range (no keyboard no mouse no monitor... ZeroClients).
But apple would never sell another device without an integrated monitor. So figure a 21" $699 iMac.
Currently you don't just license an Intel Processor you have to license the motherboard as well. The cost
is not trivial.
When you look at Mac hardware you see that Apple has essentially made them appliances. There are rarely
PCI-Express slots, there are no M2 slots and other forms of connectivity. This makes it easy to migrate to
another architecture because there are no legacy internal components to consider.
ARM vs Intel architecture from an application standard. As one poster said this is closer to the 68x - PPC transition
than PPC to Intel and todays tools (LLVM and more) make the coding changes manageable. We'll know if Apple
plans to make the transition because we'll start seeing AppKit and ARM get cozy in code.
As far as performance ARM has beefed up their architecture. You can now have up to 8 CPU in a cluster and you can cluster 32
of these Octoclusters together. Want more speed? Add more CPU.
Mobile platforms have matured. There's only so much better the displays are going to get now that we have Super Retina and OLED.
There's not going to be a big boom in sales based off of evolutionary updates.
It's time for the industry to take a look back at the desktop and focus on low power, near invisible options that provide, home automation, monitoring
and more.
Sog35 it's simply not that important to Apple's future that they ship computers with
Intel Hardware.
70% of their Rev comes from iOS.
Although I would love for therm to produce a iOSBook - one at $250 with polycarbonate body for young kids and a premium on at 899 and up,
This would be awesome and put Macs even further into the future. I read somewhere that the A11 is faster than MacBook Pros in some tasks.
I would assume Apple would start using their chips in MacBooks first but it wouldn’t make sense to have their laptop line incompatible with their desktop line so maybe we’re 5-10 years away when A-Series chips can compete with high end iMacs. This way they can upgrade the entire computer line with A-Series chips all at once.
You see i actually want to see an ARM powered laptop. I just dont want to compromise with respect to all the capability Mac OS has. Mac OS is a very fine OS with nothing that really competes.
As for the other chips it has been known for a long time that Apple wants its own baseband solution. They will work on this because integration into the SoC makes sense.
You can tell the older Mac users by how stubborn they cling to Intel. I remember the days of Macs getting their PPC clocks cleaned by Intel computers half the cost. There is a fear that moving away from Intel will lead us back there but times have changed. What we see in mobile devices with regard to integration other co-processors is going to come to Mac OS. You're going to see specialized AI chips, chips to handle realtime video and fast GPU to do heavy compute.
Desktops have stagnated not because people love mobile devices so much but because all the innovation got sucked out the platform when everything went Wintel. Neither Intel nor MSFT would take any real risk thus we have been subjected to a decade of evolutionary "ho hum" upgrades.
I don't know if Apple is going to play disruptor but they are uniquely positioned to change how desktop computing is done for the next generation.