Apple M1: anatomy of a revolution

This year 2020 that is now ending has been terrible in many sections, but paradoxically it has been a spectacular year for the world of technology. We have witnessed twelve frantic months in which there have been small and large revolutions, but the one raised by Apple with its M1 chips has undoubtedly been the greatest of all of them .

The technological leap that Apple has taken in its laptops and desktops is so surprising that we wanted to analyze it in depth: here is our anatomy of a revolution . One in which we have reviewed the path taken by this company and the keys to a transition simply prodigious due to its magnitude and execution. And it is said by someone who has very, very little as a fanboy .

At Apple they know a bit about transitions . They carry a few behind their backs, and all of them have helped lead us to this radical leap in the design of their desktops and laptops.

These transitions are especially complex because this change of processors and architecture also entails the change of the so-called Instruction Set Architecture (ISA).

The software that is developed for these chips makes use of the instruction set of those processors, and that is the reason why programs belonging to one family of processors do not work well in another .

At Apple they know a bit about transitions: they carry a few behind their backs.
All this makes these hardware transitions very complex, especially on platforms with a large user base and a large software catalog: suddenly all or a large part of this software may no longer be available to users , who must assess whether They are worth the leap to a new architecture or not.

This is where layers of emulation come into play that try to alleviate the problem and “translate” instructions from one ISA to another in real time, although this usually means that the emulated software runs slower than on the original processor.

Apple has made several notable transitions throughout its history, and it doesn’t hurt to review how there have been various jumps and decisions that have brought us to where we are now.

The appearance in 1984 of the first Macintosh with Motorola 68000 family processors was a milestone for Apple, but a few years later the company seemed to be clear that those chips were lagging behind.

That led to Apple’s first attempt to develop its own processor . In 1989, the so-called “Scorpius Architectural Specification” was published, a document that explained the general concepts of multicore CPU architectures, and although that effort – known as “Project Aquarius” and led by Jean Louis Gassée – never saw the Daylight laid the foundation for what we have today with Apple’s M1 chips.

Rather than go down that path, Apple teamed up with IBM and Motorola to make the leap to PowerPC processors . That would allow it to compete with the increasingly powerful PCs based on microphones from Intel, but it required a component that would allow users to work with applications of the previous architecture.

As explained in ComputerHistory , Apple had to create an emulator so that 68k programs could run on PowerPCs . The effort to create it began in the mid-1990s, and by the time Apple announced its first PowerPC-based Macs, the emulator was ready and part of all the equipment the firm sold in the following years.

That emulator was part of the ROM of PowerPC-based Macintoshs and allowed running “old” Macintosh applications without (too many) problems .

There were some exceptions that did not work in the emulator – like the mythical RAM Doubler – but in general m68k applications simply ran slightly slower than when they were compiled for the PowerPC. If all this sounds familiar to you, it is because it is exactly what is happening more than 25 years later with the M1 and Rosetta 2 when executing code from the Mac based on Intel microphones.

Just as important as the hardware transitions was also the largest software transition in Apple’s history , which in March 2001 introduced the first stable version of Mac OS X 10.0, codenamed “Cheetah.”

The gestation of that operating system came from afar: when Steve Jobs had to leave Apple, he began an ambitious project with NeXT and its NeXTSTEP operating system , which was based on two key components: the Mach kernel from Carnegie Mellon University and the BSD operating system. , an implementation derived from early UNIX.

Those efforts made perfect sense when Jobs returned to Apple in 1997. By then the company was going through a critical moment and its Mac OS operating system paled before Windows 95 , which already offered preemptive multitasking (compared to the mac OS cooperative) and that it was also getting dated in terms of its user interface. Copland, the project that tried to offer an alternative, was a failure , and Apple looked for another way to revamp its operating system.

When Jobs returned to Apple in 1997, the company was going through a critical moment and its Mac OS operating system paled before Windows 95.

The return of Jobs ended up offering the solution. In February 1997 Apple bought NeXT for $ 427 million and with that acquisition it made both Jobs and his leadership and the OpenStep operating system, which became the basis of Mac OS X.

That was a huge software leap that forced the creation of several layers of compatibility to ensure that Mac OS users could continue working with their old applications (Carbon) and enjoy the new options of Mac OS X (Cocoa) .

The appearance of Mac OS X was a real bombshell for an Apple that with this operating system was able to introduce not only highly demanded features in areas such as multitasking or multi-user support, but also an interface that shone with its own light thanks to Aqua and components like its famous Dock.

Ten years after the jump to the PowerPC, Apple found itself with the same problem that it had had then: its machines were beginning to lag behind , and the fault was the slow progress of processors over which they had no control .

top 10 booters