Originally published at: https://tidbits.com/2018/05/29/how-apple-is-chipping-away-at-intel/
For decades, Intel was the been the undisputed king of computer chipmakers, but in just a few short years, Apple has challenged the throne with its A-series chips. The Motley Fool’s Ashraf Eassa explains why.
Other than getting more powerful in their own sphere, how do ARM processors compare with Intel chips, say those used in Mac laptops? Not to mention the more powerful iMacs and soon to be debuted Mac Pros. The last I heard, for technical reasons, regular computer operating systems could not run on ARM chips, neither Windows nor the macOS. Has something changed to erase these technical difficulties? No offense, but no explication of these issues ever seems to show up in these macOS on ARM articles. Like, you know, if you don’t have an answer to a problem, just don’t bring it up.
The iOS was specifically designed to run on ARM processors. Due to their lower power requirements ARMs were ideal for cell phones and tablets. Now these chips and the devices that run on them are much more capable than they once were, but power is still the most important single consideration in any cell phone or tablet design.
At the same time, whatever problems Intel is having do not seem to have much affected Apple’s development of new laptop and desktop computers. Of course there may be problems behind the scenes that we don’t generally know about but they don’t show on the front end.
While I wouldn’t mind an ARM powered Mac, given equivalent performance with current Intel powered machines, I just don’t see that happening. Among other things, converting macOS to run on ARM would be a humungous undertaking with more than serious downside possibilities. It would be far simpler to convert iOS to run on Intel. With the unification of the file systems on iOS and macOS, compatibility seems to be improving.
Indeed, Adobe has developed a version of Lightroom that runs very well on iOS, with a comparable Mac version. But it’s still a poor imitation of Lightroom Classic CC. It’s designed for those for whom their primary camera is a smartphone with primary storage in the cloud.
Frankly I think this Mac on ARM notion is a chimera on a par with the cord cutting fantasies of those who think cable is on the way out. It’s not, for solid reasons I don’t need to go into here. While Josh is normally a level headed guy, mostly, nevertheless he is blowing smoke here.
Now I’m no expert and I remain to be convinced—as soon as a rational explanation of the compatibility problems with ARMs are explained. I have yet to see any hint of such an explanation. It’s like the bandwidth issue with streaming video; it just can’t compete with cable. So streaming fans just ignore it and make due with a lower quality picture. What you don’t know can’t hurt you—right? Until it does.
If I were a betting man, I’d happily bet money on Apple already having an internal version of macOS running on ARM, just as they had an Intel version for years. I don’t know anything that would prevent a desktop operating system from running on ARM, other than just performance, and Apple’s A-series chips are already “desktop class.” Something is surely holding the Mac lineup back, because outside of the iMac and MacBook Pro, the Mac lineup has only received nominal updates for the past few years.
There are a lot of reasons why Apple would switch to ARM: lower power consumption, less heat generation (imagine a MacBook Pro you could actually put in your lap), drastically better battery life, and most importantly, more control. But this ExtraBIT wasn’t really my commentary, but that of Ashraf Eassa, who knows way more about this stuff than I do.
And for what it’s worth, I’ve always had way better image quality from streaming than I have from any cable service.
Yes, software written in a compiled programming language needs to “target” a particular CPU architecture. But most software does not have to be written with that CPU in mind, they’re written in a higher-level language and its the compiler’s job to turn that into instructions specific to the CPU architecture.
Macs have already been through this multiple times, first transitioning from 68k to PPC processors and then from PPC to Intel processors. In the transition to Intel, software came as Universal or “fat” binaries, containing the code for both PPC and Intel architectures. An additional strategy was Rosetta, allowing PPC instructions to be translated to Intel instructions on the fly.
A big deal has always been made of of OS X/macOS, iOS and Apple’s other operating systems share underlying code, making the effort a less humungous undertaking.
There’s already a version of Windows 10 for ARM processors and includes Intel (x86) emulation though it might be limited to x86 applications written for the Universal Windows Platform (UWP).
And why should that be Intel’s fault or related to Intel’s CPU releases? Last I checked PCs came equipped with plenty powerful CPUs and routinely either their cores/clocks get improved or prices come down. It’s the Mac world where computers seem to have taken back seat. Nothing, I repeat absolutely nothing, has so far indicated Intel would be preventing Apple from improving Macs. It’s Apple’s lack of interest and attention that’s preventing Macs from being improved, not a lack of powerful CPUs. Plenty of those are available to the x86 world — at least those interested more in computers than say TV productions or hiphop music radio.
On Jun 5, 2018, at 3:24 AM, B. Jefferson Le Blanc <tidbits-talk@talk.tidbits.com
Other than getting more powerful in their own sphere, how do ARM processors compare with Intel chips, say those used in Mac laptops?
My guess, and as I don’t know much about battery technology is not a very educated one, is that maybe they need ARM for apps to run on iOS as well as Macs. I don’t know what kind of chips are in Home Pod, but it would make sense to have one kind if chip to rule them all. Since Apple is going full speed ahead with VR and AR, and there are rumors that they are developing glasses, ARM might be needed here as well. Devices that work seamlessly together is what Apple is all about.
Not to mention the more powerful iMacs and soon to be debuted Mac Pros. The last I heard, for technical reasons, regular computer operating systems could not run on ARM chips, neither Windows nor the macOS. Has something changed to erase these technical difficulties?
Though I haven’t heard this till today, I’ll bet that Apple had Macs in mind from the moment they realized they had to roll their own chip to bring iPhone to market. Intel, or any other manufacturer’s chips couldn’t handle what was needed for iPhone.Cost savings is another critical factor. And since cars are also a not terribly secret project, chips will be important here as well.
The iOS was specifically designed to run on ARM processors. Due to their lower power requirements ARMs were ideal for cell phones and tablets. Now these chips and the devices that run on them are much more capable than they once were, but power is still the most important single consideration in any cell phone or tablet design.
Power, size and weight. My MacBook Pro sits next to my gigundo, noisy G5 cheese grater. Faster, smaller and lighter are all important.
At the same time, whatever problems Intel is having do not seem to have much affected Apple’s development of new laptop and desktop computers. Of course there may be problems behind the scenes that we don’t generally know about but they don’t show on the front end.
One of the reasons Apple started developing chips was they needed more advanced technology than Intel was willing to develop in order to build iPhones. Apple had switched to Intel from IBM a few years before because RISC couldn’t handle OSX and IBM decided to exit the chip game.
Intel has been loosing an increasingly large amount of revenue since Apple started rolling their own. Their stock prices haven’t been doing well either.
While I wouldn’t mind an ARM powered Mac, given equivalent performance with current Intel powered machines, I just don’t see that happening. Among other things, converting macOS to run on ARM would be a humungous undertaking with more than serious downside possibilities.
Apple has had a big and constantly growing team of chip developers on staff since iPhone was a gleam in Steve Jobs’ eye. I’ll bet they’ve been working on the transition for Macs for years and years. There’s a growing battery technology group as well.
It would be far simpler to convert iOS to run on Intel. With the unification of the file systems on iOS and macOS, compatibility seems to be improving.
No way. Apple is all about their devices running seamlessly together, and it is one of their biggest selling points vs. Android and Windows. And adding to the big cash stash is critical to Apple’s future as well.
Frankly I think this Mac on ARM notion is a chimera on a par with the cord cutting fantasies of those who think cable is on the way out. It’s not, for solid reasons I don’t need to go into here. While Josh is normally a level headed guy, mostly, nevertheless he is blowing smoke here.
I’'m convinced Josh is right on the money here. Remember hundreds of millions of $$$$ Intel spent every year on proprietary and co-op print, broadcast and digital ads to reach the consumer and trade markets? “Intel Inside” is no longer a distinguishing factor because Intel is no longer the most advanced chipmaker, and manufacturers refused to pony up co-op dollars to pay to advertise something consumers no longer care about. The market PCs has been shrinking for years, and Intel bet on the wrong horse by not developing more advanced technologies, and they are just starting to acknowledge this:
Intel Outside: How The Chipmaker Is Pitching Itself Beyond The PC:
http://adage.com/article/podcasts/intel-make-stuck-inside-a-pc/313094/
Marilyn
If Intel isn’t making the chips they need for the Macs they’d like to build, then it’d make sense that they’re treading water until they can build what they want…
The problem with that logic is that it’s precisely the Macs that are being updated where you could argue Intel is lagging. There is no decent chipset for low power and large RAM, there was (until recently) no adequate 4-core low-power mobile CPU, etc. But ironically, for those Macs that have seen the fewest updates and have been almost entirely neglected (MP, Mm) there are many powerful CPUs available right now. Problem is, Apple chooses not to use them for lack of interest or commitment or whatever.
All this reasoning would have some merit if it were the MBs and MBPs that weren’t seeing any updates. And it’s there on low-power mobile where ARM might be most appealing. But it’s exactly those Macs that have seen (at least minor) modifications. However, on the desktop we have almost complete stagnancy (apart from the iMac Pro which is utterly niche) and that’s for sure not because Intel doesn’t offer powerful and competitive CPU solutions. It’s simply because Apple—by their own admission—chose to focus on other stuff.
Apple has boxed themselves into a corner. They refuse to lower the price points. So that would force them to update at least once annually to remain competitive. On the other hand, they seem so distracted with all kinds of other initiatives (that apparently haven’t led to any selling product) they simply haven’t committed the resources to really significantly advancing the Mac. And to make things worse they then set themselves goals that are neither in line with what many pro desktop users are looking for (to for example replace their aging cheese graters) nor with what more price-aware consumers are worried about. No pro has ever said I need my computer to come boxed with a screen that at most may be 1/8" thickness at the edge. And no Joe Sixpack has ever said I want to pay $500 extra so my screen is at most 1/2" thick. But if that is the corner Apple wants to work out of they would need to do far more to make it really interesting. Overly anorexic designs that can never be upgraded, OLED icon bars, or three-year old CPUs didn’t fool the pro Mac users nor entice PC users to switch to macOS en mass. Obviously Apple gave up on that to be content with PC users ditching their PCs for iPads. A real shame nobody thought to consider that there’s max $300 to make off that iPad sale, while a decent Mac would have made them three times that.
June 5
jcenters:
If Intel isn’t making the chips they need for the Macs they’d like to build, then it’d make sense that they’re treading water until they can build what they want…
The problem with that logic is that it’s precisely the Macs that are being updated where you could argue Intel is lagging.
The problem with this logic is that Macs would be doomed to failure if Apple did not keep them tightly integrated in the Apple ecosystem. I’ve been reading for at least the past year or two that over 1 billion iPhones are currently in use globally. A quick search turned up the fact that the global Mac user base in 2017 was only about 100 million.
One of the essential elements of the Apple ecosystem is that hardware and software play nicely with one another and systems and services are as integrated as technology will allow. This is critical for consumers, and a huge incentive for developers to build apps first and foremost for Apple. The announcement that apps will soon be available for iOS and Mac OS was long overdue, and probably the reason that Apple News for Mac OS wasn’t announced till yesterday. Going forward, having chips that Apple designed and engineered for iOS/Mac OS that will run faster, better and more efficiently is simply better business and a smarter development strategy.
There is no decent chipset for low power and large RAM, there was (until recently) no adequate 4-core low-power mobile CPU, etc. But ironically, for those Macs that have seen the fewest updates and have been almost entirely neglected (MP, Mm) there are many powerful CPUs available right now. Problem is, Apple chooses not to use them for lack of interest or commitment or whatever.
Intel is even dedicating more and more of its product development and production away from laptops/desktops. Please check out the recent article and podcast and interview I linked to in my earlier post.
Apple has boxed themselves into a corner. They refuse to lower the price points. So that would force them to update at least once annually to remain competitive.
Sales of Macs started to decline about two years ago; they have been tanking for years longer among lower profit margin PC manufacturers. IBM sold its PC hardware division over a decade ago, and Sony about 3-4 years ago. HP split off its PC business, Gateway died a quick and painful death, etc. Years ago Dell repositioned its self as an enterprise focused company, and has zeroed in on IT rather than consumer and small business sales. They were all selling many billions more PCs for a LOT less money than Apple was with Macs. Low margin on low commodity products = low or no profit, which was something Apple learned the hard way in the past.
Except for upgrades to Windows, Microsoft is putting its weight in acquiring companies like LinkedIn and Github (which was strategically announced the day before Apple’s conference) and also becoming more of a cloud and business services company.
Again, Josh is right on the money here.
None of any of that requires ARM over Intel. Integration has never required a common ISA as Apple has demonstrated themselves for the past decade. Just because one thing makes sense on mobile does not mean it has to make sense on a MP. Intel and x86 will be with us for many years to come in the performance computing area no matter what happens in mobile doohickey world. And if this ARM thing does actually happen on the Mac, it would either mean ARM and Intel Macs will continue to co-exist (possibly even within the same product) or Apple would essentially have to abandon the pro market (and likely desktops) entirely. The latter makes zero business sense. Especially long term.
We all know that Apple has done an amazing job making ARM processors that run with desktop performance in mobile devices that are designed to be extremely power efficient. Why do we not think that Apple can also design ARM processors that have even more performance because they do not require so much battery efficiency?
My guess is that ARM-based Macs would be using not the same A11 that’s in the iPhone X (or presumably A12 in this year’s iPhones), but something designed instead to be used specifically in more powerful battery-based notebooks or always-powered desktops.
June 5
None of any of that requires ARM over Intel.
Profitability is one just one of the many critical elements in the equation. They can make smaller, lighter, more efficient and faster chips for a lot less money then whatever they can buy them from Intel. AFIK Apple does not own Intel stock or have have any fiduciary responsibility to the company, but they do to Apple shareholders, the people who own the company.
Integration has never required a common ISA as Apple has demonstrated themselves for the past decade. Just because one thing makes sense on mobile does not mean it has to make sense on a MP.
To grow and increase profits in the future, Apple must innovate way beyond all the their innovations of the past decade. People are buying fewer and fewer PCs and taking a lot longer to replace the PCs they have. They are using mobile devices more and more to do more and more things. Robo vehicles, AR, VR, streaming entertainment, etc. are areas of tremendous potential growth. Mobile is now the predominant type of computing and PCs are being used less and less in business and in home.
Unfortunately, Intel has not been known for innovation for years. And when the press covers the Consumer Electronics Shows (which I attended for over 25 years when they were strictly a trade show) why do they no longer talk about the latest model PCs from Dell, Lenovo, etc.?
As I mentioned, there are currently 100 million active Mac users and 1 billion active iPhone users. That’s at the very least 900 million potential primary targets to switch to Mac. They are also good targets for Watch, Apple TV, iPad, Music, Air Pods, Home Pod, Pay, iCloud. the branded credit card Apple is scheduled to release soon, Books, and the as yet to be announced streaming service, etc. App compatibility between Apple devices will be a huge selling point for Macs among the 1 billion + 100 million + hundreds of billions globally who have never owned an Apple device, as well as an incentive for owners of older Macs to upgrade.
Size does matter when it comes to mobile as well as desktop devices, and lighter, slimmer and faster are important too. So is optimizing battery life.
Intel and x86 will be with us for many years to come in the performance computing area no matter what happens in mobile doohickey world.
Then why is Intel focusing on drones, cloud services, robotics, etc. and continues to shift away from innovating in the PC chip market? And just because Intel is there isn’t a justifiable reason for Apple to pay more for any of their products when they can make better, cheaper custom designed chips that will enable everything to run better.
And if this ARM thing does actually happen on the Mac, it would either mean ARM and Intel Macs will continue to co-exist (possibly even within the same product) or Apple would essentially have to abandon the pro market (and likely desktops) entirely.
This is BS. They’ll develop a chip or chips that will make all Macs run better and play nice with other Apple goods and services. Macs have survived because they always evolved along with the way people use them, and people will be using differently as time goes on. Personally, I’m planning to buy the next generation MacBook Pro as my million year old model’s trackpad is about to explode, and I’d love for it to have Face ID and 3D Touch, but that’s impossible with Intel Inside.
And I hope eventually there will will be one OS that can do everything.
Marilyn
The desktop line is long in the tooth, but that’s a different subject, really. It’s easily explained by lack of priority largely driven by lack of demand.
The MacBook lines on the other hand are not being ignored by Apple but are certainly being held back by Intel’s failures in their chip lines. They’ve had plenty of problems that Arm has not run into with the new sub-14-nanometer chips. In fact, IIRC they haven’t been able to make any yet.
That translates into obvious problems for Apple in their second largest hardware segment.
There is zero doubt that Apple is exploring this.
I think people have an outdated view on ARM processors. Cray is putting them into supercomputers now https://www.nextplatform.com/2017/11/13/cray-arms-highest-end-supercomputer-thunderx2/
You totally spaced the technical differences between ARM and Intel chips. I guess you weren’t around when Microsoft pitched the Windows RT tablet with a slimmed down version of Windows that could run on an ARM processor—because regular Windows could not. But app developers took a pass on the RT and it died an ignominious death. Microsoft’s subsequent Windows Surface tablets and the Surface Studio desktop run on Intel chips.
So do Macs. Yes, to my surprise Intel failed to rise to the ARM challenge. But their high-end chips are fine in Mac and Windows PCs. Maybe “Intel Inside” has passed into history as an add campaign, but Intel inside is still the predominant chip in just about every PC on the market. Walk into any Best Buy and every PC there has an Intel sticker on it. Apple, of course, doesn’t do stickers.
Also, apparently Apple itself put the kibosh on the Mac/ARM rumors at the recent WWDC. I haven’t watched the video so I only have that information second hand. Apparently you didn’t watch it either.
Sorry, you may love Josh but his pitch was dead before the pixels reached the Net. So was your defense. But thanks for the response anyway.
And you ignored that Apple has done this exact thing successfully in the past.
Twice.
You totally spaced the technical differences between ARM and Intel chips. I guess you weren’t around when Microsoft pitched the Windows RT tablet with a slimmed down version of Windows that could run on an ARM processor—because regular Windows could not. But app developers took a pass on the RT and it died an ignominious death. Microsoft’s subsequent Windows Surface tablets and the Surface Studio desktop run on Intel chips.
Nobody bought the RT because the slimmed down version of Windows 8 it ran sucked even worse than the full version of Windows, which sucked and turned out to be a disaster in and of itself. MS claimed it could run full versions of many MS apps, but the few initial buyers quickly went nuts because it couldn’t. It also wasn’t an attractive enough platform for app developers and didn’t come with fun stuff like Garage Band, etc. Plus, it wasn’t like MS was had much success with Windows Phone. There was no way it could compete with iPad and Android tablets which were already selling like hotcakes. The bad reception in the press, and the horrible PR caused, as well as non negligible sales out of the gate, caused developers to take a pass on RT because nobody wanted to buy it when iPad and Android tablets were going strong. Even worse was the fact that the few PC clone manufactures that made RT tablets dropped them almost immediately. Years later, Microsoft has been doing well with Surface Pro and the vast majority consumers don’t know what’s inside it, which is now ARM.
The fact that RT failed had nothing to do with ARM. And if it was true, it was ARM technology from around 2010. Apple is working on totally new ARM chips right now, so it’s not a valid comparison at all. HP had been bragging about Snapdragon ARMs in their laptops and other PC manufacturers are as well.
So do Macs. Yes, to my surprise Intel failed to rise to the ARM challenge. But their high-end chips are fine in Mac and Windows PCs. Maybe “Intel Inside” has passed into history as an add campaign, but Intel inside is still the predominant chip in just about every PC on the market. Walk into any Best Buy and every PC there has an Intel sticker on it. Apple, of course, doesn’t do stickers.
Macs are most definitely not run of the mill PCs, and they command premium prices because of the technically advanced and unique hardware and software they offer as well. They are also get points for being distinctly designed, lightweight and fast. Having an Intel sticker would send a message that “this overpriced box is just a boring Windows wannabe that costs a lot more money for the same stuff.”
Also, apparently Apple itself put the kibosh on the Mac/ARM rumors at the recent WWDC. I haven’t watched the video so I only have that information second hand. Apparently you didn’t watch it either.
I did read the article and watch the video. But you clearly didn’t even bother to read what I quoted in my post. Please at least read the third sentence.
"Tim Cook made it quite clear at the WWDC that iOS and Mac OS will not completely merge.
At this point, Apple is beginning to share more features between devices:
Apple definitively confirms iOS and MacOS will not merge. How Apple sees iOS apps bringing new life to Macs
And despite Apple’s denials, people can’t help spotting hints that Macs and iPhones may converge one day."
https://www.cnet.com/news/how-apple-sees-ios-apps-bringing-new-life-to-macs/
Marilyn
While that is entirely true…
This is complete marketing BS.
None of the electronics on a Mac board are anything out of the ordinary. It’s straightforward to build a hackintosh that blows the doors off of anything with an Apple on its side. The fact that more people don’t do it is because it takes some knowledge and skill, it’s against the macOS EULA, it’s never as well integrated as Apple does, and Apple will not support it long-term (drivers, updates, etc.) like they do their own kit. But none of the electronics are anything else than industry standard components that every other PC maker either buys (or in principle could buy) and uses themselves. The one exception might be the rare case when Apple gets a specific component first and it takes PC manufacturers a few weeks to months to be able to buy the same component from eg. Intel. But in general, there is nothing magically special about what is on a Mac board. The magic comes from the surrounding packaging, the integration, and the software.
Not necessarily. There are different quality levels of components that a computer maker can buy…while Apple obviously wants the cheapest parts they can get…they don’t necessarily buy bottom feeder quality of components. For instance…back in the day when they put optical drives in it was…at least intuitively…a higher quality drive as failure rates for Apple DVD drives were lower than what you would get from Joe’s Backyard Computer Company. If you pay 10 bucks for a DVD drive you’re likely going to get a better quality mechanism than if you pay 5 bucks. Just as Rolls Royces are better mechanically than a Yugo…and $30 wine usually tastes better than Two Buck Chuck…in computers as in many other things one gets what one pays for.
Apple is a premium brand with accordingly high profit margins…but as others have noted you get better software, don’t have driver or DLL conflicts, and the computer as a whole is more of a joy to use…I used to be a Windows sysadmin and while their server products are fine there were many many more problems with desktops. Windows was…and still is but that’s the drawback of needing to support every possible combination of hardware pieces and parts…the IT guy’s Full Employment Act.
I’ve rarely been able to not eventually solve any problem with any of the Macs I’ve owned, supported, or assisted with…the same can’t be said for Windows. Frequently the solution the user ends up with is that the problem went away on reboot or he has a workaround because either it’s not a reproducible problem, or it’s intermittent or something similar. One of my worker bees back in the day had a saying…”A frequently rebooted Windows machine is a happy Windows machine.”…and there’s a whole bunch of truth behind that statement.
Apple commands premium prices…because they can…and because people are willing to pay more for software that just works, hardware that’s more elegantly designed, and for fit and finish that’s just better. I can’t tell you how many times I’ve snagged a finger on the inside of a case and had to wipe blood off of something inside…but it’s never happened with an Apple built product.
When the tech was ready. Which it’s not. As well, the barriers were not as substantial. And then, Apple did not deny they were working on the changes. As they have now. The fantasm of synergy here is apparently too appealing, even when the evidence is lacking. They used to call this vaporware, except that it’s not Apple pushing it. It’s daydreaming tech pundits—who don’t know enough to understand what they don’t know. Ignorance is bliss, or so they say. Suppositions and maybes are not evidence. And that’s all Josh’s article, and your defense of his article, have going for them. Apple could be doing this, and they might be doing that. Empty. Air has more substance.