search slide
search slide
pages bottom

10 years on, Apple's risky move to Intel Macs is one of its biggest successes

Nine years ago today, Apple launched the very first iPhone. But the foundation for its success had actually been laid a year earlier. Ten years ago, in the summer of 2006, Apple had just successfully transformed the Mac, completing a transition from using PowerPC chips to Intel chips and essentially transplanting the the "brains" of its entire line of personal computers.

The transition, which was first announced at Apple's Worldwide Developers Conference in 2005, was a big deal. Apple was moving from IBM-built PowerPC processors in its Mac computers to x86 processors made by Intel.

The shift was necessitated by the simple fact that IBM’s consumer processors — especially for laptops — simply weren’t competitive with the work Intel was doing. PowerPC processors had the potential to be very fast, but they ran hot and required a lot of power, which is bad for laptops. Moreover, IBM was late in delivering chips, and Steve Jobs felt Apple needed better chips sooner to be competitive.

And to be competitive, Apple needed to adopt the same x86 processors that Windows-based PCs had used for two decades.

But switching hardware architectures this way wouldn’t be easy. Apps are compiled not just for an operating system, but for a hardware platform. Changing the processor means that not only does the underlying operating system (in 2005, this was Mac OS X) need to be compiled to run on the new processor types, but every app built for the old processor architecture would need to be recompiled, too.

Apple set in its sights a goal of starting the hardware transition (that is, offering Intel-based Mac computers) by June 2006 and finishing by the end of 2007. It turned out, by August 2006, the hardware transition was done, with Apple managing to overhaul its entire consumer Mac line into Intel machines.

The fact that the transition succeeded — and ahead of schedule — was a huge moment for Apple.

Moving from one hardware platform to another isn’t easy. In fact, Apple is the only company that has successfully shifted from one platform to another. Notably, the Intel transition was the second time such a shift occurred.

In the mid 1990s, Apple shifted from Motorola 68000 processors to the PowerPC platform. This transition — though also successful — was smaller in scope than the transition to Intel.

John Gruber of Daring Fireball, admitted he didn’t think Apple would move to x86 processors, primarily because he was concerned with how Apple could preserve compatibility with existing Mac software. Gruber didn’t think that was possible.

It turns out, it was.

Apple developed a piece of technology, aptly called Rosetta, which would act as an translator to convert code built for the PowerPC to run on Intel-based Macs.

Built into Mac OS X Tiger, Rosetta would work in the background, translating the code for a PowerPC app to work on an Intel Mac. The end result was that for many applications, developers wouldn’t need to modify their code at all. It would just work on the new systems.

Of course, not every PowerPC app would work on Intel. Some apps had codebases just too complex to be automatically translated and emulated for Intel. Notably, Adobe’s Creative Suite 2 was such app. Adobe didn’t release a version of Creative Suite that worked on the Intel Mac until 2007.

The bigger challenge was for developers who were trying to build new apps. How could they build apps that would work on an older PowerPC Mac and and Intel Mac.

This was a huge concern because, even after fully transitioning to new Intel-based Macs, millions of older PowerPCs were still in the wild. Cutting off those customers from new software was simply not a viable option for many developers. For companies like Adobe and Microsoft with huge swaths of customers on the older platform (and who had publicly committed to supporting Intel Macs), it was very important that Apple offer them a way to serve both sets of users.

And this is where Apple’s transition strategy became particularly brilliant with the advent of universal binaries. A universal binary would mean that an app would be packaged with code that could run on either platform. So if you inserted a CD (remember, this was 2006) or downloaded an app from the web, it would work on your Mac, regardless of whether it was coded for Intel or not.

Apple's universal binary tools wound up being a huge success. I was writing about Mac software after the Intel transition, and it was amazing to see how many developers — large and small — quickly migrated to the universal binary format. And even after Apple officially dropped support for the PowerPC with Mac OS X 10.6 (in 2009), there were still some developers who continued to support that platform via universal binaries for years to come.

Even though Apple primarily moved from PowerPC to Intel x86 for performance reasons, one of the huge advantages of the move was that for the first time, Mac users could run Windows natively on Mac hardware.

In April 2006, Apple introduced Boot Camp, a way to natively run Windows XP on your Mac.

Boot Camp worked by partitioning a user's hard drive so it could run Windows alongside Mac OS X. At startup, you could choose what operating system you wanted to run.

For the first time, users who enjoyed using a Mac at home — but also had a few Windows-only apps they needed to use — could use one machine.

And the rise of Boot Camp led to virtualization programs such as Parallels Desktop for Mac (first released in 2006) and VMWare Fusion (first released in 2007), which would allow users to run Mac and Windows apps side by side, without a noticeable impact on performance.

Thanks to the iPod, Mac usage was on the rise. But by moving to Intel and gaining support for Windows via Boot Camp or a virtualization program, millions of people who wanted a Mac — but couldn’t commit to giving up Windows — could finally have both.

Nowhere was this more evident than with the MacBook: the 13.3-inch Intel notebook Apple first released in May 2006.

The MacBook wasn’t the first Intel-based Mac (an Intel iMac as well as 15- and 17-inch MacBook Pros were released in early 2006), but it was the most important.

A successor to the iBook, the MacBook was the first modern Mac aimed at the masses. Plus, it was hot.

Available in black or white, the design was a departure from the iBook of old. It was sleeker, had a wide screen, and it had attitude.

Moreover, with a starting price of $1,099 and a Core Duo processor, the original MacBook was actually competitively priced when stacked against Windows machines with similar specs.

Moreover, the MacBook was arguably better than a standard Windows machine because it could run both Mac and Windows apps.

I watched my own college campus transform from Windows laptops to a sea of MacBooks between 2006 and 2007. For under $1,500, students could get a powerful laptop that would do everything they wanted — and could still run Windows when absolutely necessary.

The MacBook launch also coincided with the ‘Get a Mac’ campaign. For four years, “I’m a Mac and I’m a PC” became an advertising staple and helped sell millions of Macs.

I’d even argue that the success of the MacBook — and Intel Macs in general — helped excite the general public about what would be Apple’s most important product: the iPhone.

When the iPhone was announced in January 2007, the Intel transition had only been complete for a few months. And although many people had become re-acquainted with Apple thanks to the iPod, the resurgence of Mac popularity only helped make the product that much more desirable.

The Mac, as a platform, had a huge renaissance after the Intel transition, too. There were already tons of indie developers building apps for the Mac, but the success of the Intel Mac out of the gate helped create a market for fantastic first-party Mac apps.

Games notwithstanding, plenty of companies now build apps for the Mac first (Evernote, 1Password, Sketch and Pixelmator, to name a few) and put Windows second (if at all). Moreover, the move to web-based apps made the distinction between operating systems less important than ever.

Ten years ago, if you walked into an office, a university or the average person’s home, you were probably met by a PC.

Today, that’s just not the case. Look at any Starbucks, school or startup office and you’re greeted by Macs.

The MacBook Air, which basically became the successor for the first Intel MacBook, has inarguably defined what a modern laptop looks like. The new MacBook is already starting to have a similar impact.

Windows still far outsells the Mac — Chromebooks now outsell Macs too — but the Mac has made a remarkable comeback. And for Apple, Mac users tend to be loyal to the whole ecosystem. They buy iPad tablets and iPhone devices. They buy the Apple Watch.

And none of that would have happened without the transition to Intel x86. The hardware wouldn’t have been able to keep up – but more importantly, having that ability to run Windows — just in case — helped give consumers and business users the reassurance that worst case, they could still run Microsoft stuff, too.

When I look back at how well the Intel transition was handled, it’s also important to think about how it could have gone wrong.

When Windows 8 was released, Microsoft attempted to bring parts of its operating system to the ARM processor using a new operating system, Windows RT. Windows RT was a disaster.

Microsoft didn’t create a Rosetta-like layer to bring existing x86 Windows apps to Windows RT. Instead, it required developers to build in code that would run both places (similar to universal binaries). The problem was, there was never a big enough investment from Microsoft (or more importantly, its hardware partners) for developers to bother switching their coding practices.

In fact, Microsoft is still trying to transition developers to build Universal Windows Apps. This time, the goal isn’t to have people build apps for the ARM-based Windows RT platform, but for ARM-based phones running Windows 10.

And to see Microsoft struggle to get its developers to adopt Universal Windows Apps (along with the total collapse of the Windows RT platform) simply underscores just how amazing it is that Mac OS X transitioned quickly and with few battle scars.

For a few years, there has been speculation that Apple would transition its desktop operating system to ARM. The release of the Intel Core M-based MacBook dispelled many of those rumors, but the truth is that iOS and macOS have the same kernel and now share much of the same code.

A transition for most developers from x86 to ARM would probably require less work than the transition from PowerPC to Intel — if only because of the changes made in software development and tooling.

I’m not convinced we’ll see an ARM-based Mac anytime soon, but I wouldn’t be surprised if it happened at some point in the future. Moreover, if Apple were to transition processing platforms again, I’m even more convinced it would succeed.

Have something to add to this story? Share it in the comments.

Leave a Reply

Captcha image