Apple Announces Transition from Intel to ARM Processors

During the Apple Worldwide Developer Conference (WWDC) keynote, Apple announced that beginning this fall, they will begin transitioning their Mac computers from Intel processors to Apple designed ARM Processors.  This is easily the biggest change for apple since the release of the iPhone in 2007.

The Tale of the Processor

Talking processors can get into the weeds very quickly, so we’re going to try to keep this at a high level.  Every electronic device has some kind of processor.  On your laptop it probably has an Intel processor of some kind, you’ll hear commercials market it as “Intel 10th gen Core i5” or something similar.  Intel or AMD processors running on an x86 architecture have been present in almost every Windows computer since the mid 1980’s.  “x86” refers to the instruction set, or programming language, that the processor works on.  Mac computers also run on Intel x86 processors, but they haven’t always.  More on that in a bit.  Devices like your phone run on processors with an ARM instruction set.  If you have an Android phone, your phone likely has a processor marketed as a Snapdgragon processor, while Apple iPhones and iPads have what apple brands as an “A” series processor.  The iPhone 11, for example, has an A13 processor.  While ARM processors do the same thing as Intel processors, they do them differently, with different “instruction sets.”  The operating systems, apps, and all underlying software must be written for the instruction set they are running on.  An app written for Windows on an Intel processor won’t run on an iPhone with an ARM processor, and vice versa.

Apple and ARM

Apple has been making processors for iPhones and iPads for years, with the A series processor.  ARM processors are a generic design, and companies like Apple or Qualcomm can take those designs and tailor them for their specific needs.  Apple has been using their self designed A series processors since the A4 processor in 2010.

One of the key advantage of ARM processors over processors from Intel is efficiency.  The phones and tablets that ARM processors run on are very low power devices, as battery life is the priority.  But as the design of those ARM processors has matured, they have become more powerful, while keeping the power efficiency.  During the keynote Apple kept touting their A series processors “performance per watt” metric as being superior to Intel.  As Apple has developed the A series processors, they can, in synthetic tests, get equivalent performance to a laptop with an Intel processor, while consuming much less power.

Great, so The Switch Should be Easy

There are definite advantages to making the switch.  Apple makes the A Series processors, they’re just as good as Intel processors, and consume less power, but it isn’t always that easy.  As said earlier, apps aren’t able to run on different instruction sets natively, so this is a much more complicated process than just changing the processor in a Mac laptop.  All of the software has to be re-written to work on an ARM processor, from MacOS itself to all of the apps we use every day.  This is a big undertaking.

On the Road Again

Now, Apple has actually done this before, most recently in the mid 2000’s.  from 1994 to 2005 every Mac ran on an IBM PowerPC processor.  At the same WWDC conference in 2005 Steve Jobs announced that Apple would switch to Intel.  The reasons provided were the same as what we heard today.  Intel’s performance per watt was better, there was more battery life from Intel processors, etc.  Steve Jobs laid out a roadmap of how it would work.  Apple provided tools to developers to make their apps work on Intel processors.  Developers could make what was called a Universal Binary, which was an app that would contain the instruction sets to be able to run on both PowerPC and Intel processors.  This allowed developers to only have to write an app once, and they did not have to maintain two separate versions of apps, one for PowerPC, and one for Intel.  Apple also announced a platform called Rosetta.  This was something built into OS X that allowed apps made for PowerPC to work on Intel processors.  This worked by creating a translation layer which told the Intel processor how to handle the PowerPC instruction set.  Apps run in Rosetta were often slower and did not work quite as well, but it provided a vital compatibility layer for users who had apps that might not have been updated to work on Intel laptops.  It wasn’t a perfect system, but at least it was something.  Universal Binaries meant that users who bought a PowerPC laptop in early 2005 knew that they would be able to get support in apps for years, and users who bought an early Intel Mac in 2006 knew that most of the apps they use would run even if they weren’t updated yet.

Fast forward to 2020 and Apple took that playbook and has done essentially the same thing.  The company announced the “Universal 2” program to allow developers to create apps that will run on both Intel Macs and A series Macs.  Apple claims it will take developers just “a few days” to convert their app to a universal app.  Apple also announced Rosetta 2, which will allow apps designed just for Intel macs to run on A series Macs.  Apple promises that apps running through Rosetta 2 will be much faster and there will be fewer limitations.  This, again, allows users who bought an Intel Mac today to know that they will be supported for years to come with Universal 2 Binary apps that work on both Intel and A series processors (very good for people who may have spend $30,000 on a Mac Pro in 2019).  It also gives users who buy the first A series Macs peace of mind that almost any app they want to use will at least work, if not as well as an app designed for ARM.

The Advantages

There are several advantages to Apple making this change.  First and foremost, it means that the processors that run Mac computers are designed by Apple instead of Intel.  Because of this, Apple has even more control than they already do over the entire ecosystem.  Just like on iPhones and iPads where the hardware and software are designed together, the same will happen on Macs.  This will allow Apple to get even more out of the ARM processors in future Macs.  Another benefit is that Apple isn’t beholden to Intel’s road map.  Intel has faced challenges in the past 3-4 years on their technologies, and as a result their processors have not advanced as fast as many would want, Apple included.  Apple will also be able to leverage their products all across the line.  Apple’s strategy for iPhones and iPads has been to use the previous year’s high end A series processor in lower powered/cheaper products.  The 10.2” iPad for example uses the A10 processor which originally debuted with the iPhone 7 in 2016.  The A10 is far from the best A series processor in 2020, but Apple uses it as a low cost option in the cheaper iPad.  A future Mac lineup could potentially include the expensive laptops getting a processor first like the MacBook Pro, with the MacBook Air getting the processor made the previous year. 

Apple controlling the hardware and software could also lead to interesting changes to the hardware.  Macs today are constrained by the power and heat limitations of Intel processors, and having more control over all aspects of the hardware could lead to interesting new designs for the computers themselves.  Thinner and lighter isn’t always the key, but imagine a MacBook air with 18-24 hours of battery life instead of the 6 or so they get today.  Imagine an iMac desktop that looks like a big iPad on a stand.  New hardware designs are clearly coming with this change, and Apple’s control of all aspects should lead to some fun new looks for their lineup.

The Unknown Support Period

The one unknown question is the long term support for Macs running the A series processors.  Apple has generally given their phones very good support, but that support has usually capped out around 5 years.  The iPhone 6S from 2015 was a surprise inclusion in the list of supported devices for iOS 14, giving that phone 5 years of support.  We do not, however, know how long Apple will support laptops with A series processors with updates.  Apple has so far demonstrated about 8 years of update support on Intel Macs, and I would hope that the support for A series Macs to be at least the same.  However, one complication could be the use of the older processors in those Macs.  There are not a ton of data points on how Apple will support newer products with older processors in them.  The iPad Air 2 with the Apple A8X processor.  That device was released in 2014, and it is getting iPad OS 14, giving at least 6 years of software support.  But that was a new processor at time of release.  The 5th generation iPad was released in 2017 with the, at the time, 2 year old A9 processor.  So while the device itself is only 3 years old, it’s processor is 5 years old.  That iPad will be getting iPad OS 14, so it remains supported, but we do not know how long it will get support.  Only time will tell there.

Now, a device not receiving software updates will not stop working, it will of course continue to work on whatever software it runs on, but when software updates end, devices lose access to new features, and risk not being able to access new apps and services.  The longer the update period, the better, and for now, that will remain a question mark.

I Hear You Like Apps

One of the more interesting results of a Mac running on the same processor platform as iPhones and iPads is that any Mac with the A series processor will be able to run iPhone and iPad apps natively.  This opens up literally hundreds of thousands of apps that could be run on Macs.  Now, not all of these will be ideal.  An app designed to run on a phone screen may not look great on a bit expansive Mac screen, but there is a large category of apps designed for iPads that could run very well on a Mac with an A series processor.  This opens up some very interesting possibilities.

What Does This Mean For The User?

For the average user buying a Mac laptop right now, nothing.  If you have a mac and it breaks today, You can feel ok that if you buy a Mac tomorrow, you’ll get years of support.  If you want to wait and buy one of the first ARM Macs in the fall, almost every app you want to run will at least work, if not perfectly.  No one should be afraid to buy a Mac today, or tomorrow, because of this change.

What does this mean for professionals?

For a true “professional” user, someone who uses their Mac for content creation, app development, video/photo editing, etc, it is a bit less clear.  Considering Apple released the very powerful (and very expensive) Mac Pro in 2019 with Intel processors, there should be very long term support for the Intel Platform.  But what we currently do not know is how well content creation workflows will work on early ARM Macs.  If you only use Apple made apps, it will work fine, but if some relatively obscure 3rd party tool is vital to your workflow and you’re not sure if it will get updated to support the A series processors, buying an early ARM Mac might not be the safest idea.  We also, frankly, don’t know how well the early ARM Macs will perform.  Will the first ARM MacBook pro be as fast as the latest Intel MacBook Pro?  We don’t know for sure.  I think if you can wait to buy a Mac for production purposes, it is worth waiting.

How Long Will This Take?

Apple has announced they will begin shipping A series Macs in the fall of 2020. The company has said that it intends to continue to ship Intel Macs for at least two more years, meaning that from 2020 to 2022 Apple will sell both Intel and A Series based Macs. There are also new models of Intel Macs that the company has yet to release, so Apple is not abandoning the Intel Platform in the medium term. This offers choice for users, which is something Apple does not often do. But for professionals who really do need and want the Intel Platform, you have a few years to keep buying that hardware.

Overall, this is a very exciting time for Apple.  The move from Intel to ARM is the biggest change for the PC industry since, well, Apple moved from Power PC to Intel. This will have a seismic shift in the industry that we won’t fully know the end results for for several years.  I can’t wait to see what the first ARM Macs look like, and I’m actually excited for the future of the Mac platform for the first time in several years.

The Good And The Bad of The Apple App Store

I wrote about this a bit as part of my roundup of news for June 17, but after thinking about it more, I believe this warrants more detail, and some extra content, so here we are.

The App Store

When Apple launched the App Store for the iPhone in 2008 (most people forget the first iPhone did not have any kind of app store), it changed many things in the mobile industry forever.  The App Store was, in my opinion, more important than the iPhone itself.  At the time, it was revolutionary.  Other smartphones of the day had the ability to run apps, but actually getting them was difficult.  It involved finding them on the internet, and then somehow transferring them to the phone manually. It was a cumbersome process that was too difficult for many people to attempt.  Apple changed that.  With an App Store built into the phone, users could go into the store, tap a couple buttons, and the app would automatically download to the phone.  That was it.  Everything we do on phones today revolves around apps, and this was how it began.

How Apple accomplished this was to centrally manage the system.  Apps would be stored on servers run by Apple, and users would download the apps directly from Apple.  There are several advantages of this.  Someone who wishes to create an app doesn’t have to figure out a system of getting it to people, or paying for that cost.  Since all of the apps live in one store, Apple also was able to ensure there was nothing malicious in those apps.  If a user downloads an app from the App Store, they could be assured that it would be safe.  It was a win for developers and a win for users.  How does Apple win in this scenario?  Well, developers could charge for their apps, and Apple set a policy where the company took 30% of every transaction in the store.  If a developer charged $1 for an app, when someone bought it, the developer would get 70 cents, and Apple would get 30.  Apple’s reasoning for taking 30% was explained as a way for Apple to pay for the costs of the App Store.  There are costs to actually host the applications, costs to handle the transactions themselves, and costs to have a system to make sure that apps were of a high quality and safe since Apple approves every app in the App Store  However, considering the ease of use, developers embraced this system, and the App Store exploded.  12 years later, and these policies still apply with few exceptions.

The Apple Tax

That is about 400 words on how the App Store came to be, but context is important.  And we aren’t even done with the context.  Apple has taken the two core pieces of the App Store, the 30% cut, and approving every app, and used it to their advantage.  On the 30% cut, this applies not only to apps purchased up front, but also applies to purchases made in an app.  We’ve all seen the thousands of free to play games in the store, where in app purchases rule the world.  Every time you spend a couple dollars on those gems, or extra lives, or whatever “gotcha” feature these games have, Apple also takes 30% of that transaction.  It was a logical expansion of the system, and does make sense.  And the financial windfall for both developers and Apple has been massive.  Apple claims that $519 Billion US worth of commerce was done on the App Store in 2019.  That is a lot of revenue for both Apple, and the developers making money.  This should be a win/win scenario, with little to complain about.  However, Apple applies this 30% to every transaction, even subscriptions, and that is where the troubles really begin.

Let me use Spotify as an easy example.  The single user Spotify Premium plan costs $10/month.  You can go to Spotify’s website, sign up for an account, and then use Spotify across all of your devices.  In this scenario, Spotify keeps the entire subscription fee of $10/month.  If Spotify offered the option to subscribe to Premium from within their iPhone app, they are still bound by the App Store policy of a 30% cut.  This would mean that for that $10/month cost, Apple would take about $3.33.  But that doesn’t happen once, that happens with every transaction, so every single month, Apple would take $3.33 ofo the $10 cost.  Spotify does not want to lose 30% of their fee every month, so the company has decided to not allow users to sign up for a Spotify subscription within the app on the App Store. You can download and install, and the app will tell you all about the features of Spotify Premium, however the app tells you to go to the Spotify website to sign up.  This is obviously not an ideal user experience, as it would be easier for users to just be able to sign up within the app, but Apple’s App Store policy makes that difficult.  Adding to the difficulty, Apple also has a policy that apps cannot link at all to an external source to make a transaction.  This means that Spotify is not allowed to add a link in their app to take users to a page to sign up for the Premium Plan.  Spotify can tell users to go to the website, but can’t tell users how to get there.  Other major services face the same restrictions.  This is why you can’t sign up for Netflix, Disney+, Crave, etc.  Or buy books on the Kindle app, or audio books in the Audible app.  There are many other examples, but those are some of the more high profile.  Some apps do allow users to buy the subscription from within the app, and some just take the 30% loss on revenue, while others increase the price to subscribe to the service from within the App Store to make up the difference.  It is a less than ideal scenario because of Apple’s draconian application of this policy.  And to top it all off, there are even more complications to that, which I’ll get to later.

The Apple Gatekeeper

The other tent pole feature of the App Store is that Apple approve every app.  This has the advantage of ensuring that apps are safe to download, and won’t include malware, or any content that will cause performance or other issues with the phone.  However, approval is a double edged sword.  Apple also requires all apps to adhere to all of their App Store policies to be approved.  If a developer builds an app in a way Apple doesn’t like, or a feature Apple doesn’t like, it doesn’t get into the app store.  So, back to the Spotify example, if they tried to submit a version of the Spotify app that included a link to subscribe to Spotify Premium, Apple would not approve it, and it wold not be allowed in the App Store.  This forces developers to play nice with Apple.

Trouble in Paradise

1100 words of context later, the current situation is a culmination of long term results of these policies.  Earlier this week it was found that email service hey.com had an app update rejected.  Hey.com sells a premium email subscription service, the app is unusable without that subscription.  Hey.com did not allow users to sign up for the service within the app so as to not lose 30% of the revenue, nor did they provide any link to a place to sign up within the app.  Apple did not provide a reason for a rejection of the app update, and Basecamp, the company that makes the hey.com service, went to tech news outlets.  Those outlets, which have more of an ability to talk to Apple, asked, and Apple gave a statement which indicated that any app that has a subscription service offering *must* include the ability within the app to subscribe, giving apple a 30% cut.  But, as stated earlier, there are many apps in the store that have subscription services that do not face this restriction from Apple.

So, more digging was done.  Apple’s App Store policies list several exemptions to their policy about offering subscriptions in app, the most notable is a category of “reader” style app.  A “reader” style app is described as an app that provides content consumption services like video streaming, or music, or books.  Apps in those categories are not required to offer an ability to subscribe to the service in app.  However according to the policy, other types of apps require the ability to subscribe within the app, giving apple that 30% cut, if they offer the ability to subscribe to the service from anywhere else in the world.

Let me break this out into its own paragraph.  If a company or developer offers a way to subscribe to a service via any method at all elsewhere, Apple also requires that company or developer to provide a way to subscribe within the app, giving Apple 30% of that subscription.

Now, whether you agree with that policy or not (and for the record, I definitely do not), that is Apple’s policy, and one would assume it is applied evenly.  But that is not even true.  Another premium email application, Newton Mail, offers a subscription service, and does not offer a way to subscribe within the app.  That service has existed for years, and has not run into these issues from Apple.  This shows that Apple is not applying their own rules evenly or fairly, and are arbitrarily applying the rules.  Newton mail is a service that has existed for a few years, while hey.com is a new service.  I am speculating that Apple grandfathers some older applications in, which is part of the reason they created the entire “reader” category for exemptions. Many of the apps in that category are from larger, previously established companies.  Think of the uproar Apple would face if it attempted to tell Netflix that it had to offer a way to subscribe within the app and take 30% of that.  That would make news headlines all over the world, and is PR Apple does not want, so exceptions were created to allow apps like that to continue as is, while punishing newer, sometimes more innovative services.

Not All Apps are Equal

Apple makes that 30% cut the cornerstone of the App Store.  It has been part of every App Store transaction since 2008.  Apple’s refusal to budge became a big issue.  Now, in recent years Apple has budged just a little bit.  In 2016 Apple changed the policies on subscriptions where if a user remains continuously subscribed to a service for over 1 year, the revenue split changes.  After one year Apple will only take 15% of the revenue instead of 30, giving the Developer 85%.  This was a concession, but still leaves the vast majority of transactions at 30%, including every one time purchase.

On April 1 of all days, a bombshell dropped.  Amazon’s Prime Video app suddenly started allowing purchases of individual TV shows or Movies within the app, and those purchases could be done using Amazon’s infrastructure, not Apple’s.  This indicates that those transactions are not subject to Apple’s 30% cut.  It was believed to be the first time Apple had ever allowed this.  Apple later put out a statement that said, in part, that Amazon was allowed to do this through a “well established” program for Premium video services.  However, the only examples provided for Premium services in this program were two services almost no one had ever heard of before.  Amazon became the first large provider, and as far as I know, still the only large provider, to get this type of exemption.

The Apple Advantage

Almost 1900 words in and we can see just how confusing and convoluted this situation is.  Apple has an iron grip on the the App Store, and does not exercise that grip evenly across the ecosystem.  I haven’t even talked about Apple’s own services.  Apple has music, cloud storage, video services and many more that it offers.  Apple Music, for example, has the same $10/month price that Spotify has, but Apple can provide that subscription in app and get all of that revenue, while if Spotify provides that subscription for the same price Apple takes their 30%.  That gives Apple an inherent advantage in that they get the revenue no matter what, while punishing 3rd party services.  Apple’s App Store polices create an unfair playing field, and there is no way to get an app onto an iPhone or iPad without going through Apple.  It has always been this way, but now the situation is coming to a head.

Storm’s Coming

On the day that the issues that hey.com has been experiencing came to light, the European Union announced two separate antitrust investigations into Apple and the App Store policies.  Several high profile developers have also become more vocal with their opposition about the system as it exists now.  It is an unfair playing field, and it seems that a tipping point may finally be here.  It is becoming more and more clear that the current system, in place largely unchanged for 12 years, is not working.  It is unfair to smaller developers, newer apps and services, and significantly disadvantages anyone who dares to build a service that competes with something Apple provides.  I do not expect Apple to go exquisitely into the night here, but my hope is that after all the legal fights, and the higher level of backlash for large companies, that the end result is a fairer system for all, one where anyone can thrive, instead of the current ways that Apple has taken advantage of for far too long.