M1 inside: how Apple broke free from the Intel chip

November 16, 2019 by No Comments

“As manoeuvrable as a steamship.” Apple founder Steve Jobs was not very complimentary when he spoke about Intel in his 2011 biography. The largest chip manufacturer in the world has been supplying the processor for Apple computers since 2005, but is being pushed aside. As of this month, Apple will switch to computers with a self-designed chip, originating from the iPhone. It is more economical and faster than what Intel can deliver, is the promise.

For an average user who does not demand the utmost of a computer with some surfing, e-mail, Netflix and Microsoft Excel, the chip does not seem very interesting – you can’t see it and if all goes well you can’t hear it either. But the efficiency of a processor does determine how long your laptop can last with one battery charge and how fast the fan starts to blow with some extra calculations.

Apple promises with its own chip, the M1, a giant step in operating time and computing power for ordinary computers. Such as a laptop with which you can work non-stop for seventeen hours – useful when we are allowed to travel again.

Not Jobs’ idea

The basis of the M1 is the chip that is in the iPhone and iPad. The first impulse of this stems from the time of Steve Jobs. His sneer at Intel, at the time, was about whether the iPad should get an Intel processor. Jobs opted for his own Apple chip. That was not his own idea; he was convinced by Tony Fadell, the creator of the iPod.

Apple acquired the chip company PA Semi in 2008 for $278 million. Their technology formed the basis for the A4, Apple’s first proprietary chip. Under the leadership of chip designer Johny Srouji from Israel, Apple continued to improve the technology. Right up to the current A14 processor, a chip that is more powerful than what can be found in MacBooks. This is partly due to the fruitful collaboration with Samsung and the Taiwanese TSMC, the companies that have produced Apple’s chips in the past ten years.

A proprietary processor fits in with Apple’s aim to keep control over all parts of the computer system and software. The other reason to put Intel aside is that MacBooks lagged behind the competition in computing power. Apple – the fourth largest computer manufacturer worldwide after Lenovo, Dell and HP – noticed this in sales, which come 10 percent from laptops and desktops.

Apple’s revenues from smartphones are also falling: consumers buy a phone less often. That decline can no longer be compensated with even more expensive iPhones (the basic model of the iPhone 12 costs 909 euros).

But if existing iOS users can be persuaded to also buy the matching computer – which can also run the same iOS apps – then that will generate additional revenue. In addition, Apple can continue to work on the development of a cross-fertilization between a tablet and a laptop. In that area, Windows now offers a lot more choice.

One housing

The M1 is a so-called system on a chip (SoC). That means that parts that would normally be scattered around the motherboard are now packed closely together, in one housing. That saves power.

Building such a new engine in a laptop takes preparation time. Smartphone chips are based on a more economical design technique from the British ARM. However, it works with different instructions than a ‘normal’ Intel processor.

The transition from Intel Inside to Apple Silicon requires adjustments in the software, both in the operating system and in applications from Adobe or Microsoft, for example. Apple will allow two years for the transition. Until the software is adapted, users can use a detour – ’emulation’ – to run programs that normally only work with Intel chips.

Microsoft has created a version of Windows that runs on ARM chips. For many Windows applications, this is still too great a threshold. But if Apple does manage to make the leap to the much more fuel-efficient ARM laptops, it’s likely that more PC makers will consider that option.

The implications for Intel

The smartphone processor is catching up with the desktop and laptop. This is at the expense of Intel (turnover 71 billion dollars), which sees its technological backlog growing. Or rather, shrink. Because Intel’s competitors TSMC and Samsung produce much finer chips, accurate to 5 nanometers (a nanometer is a millionth of a millimeter). Intel gets stuck at 10 nanometers and that too is not easy.

The backlog is partly due to Intel’s difficulty deploying ASML’s latest EUV lithography machines, while TSMC and Samsung are succeeding. It also has to do with Intel’s lack of agility. Intel’s competitor in the PC market, AMD, is now making a remarkable comeback. AMD does not have its own factories and outsources production. With success: AMD’s market share in PCs is increasing and with it the stock market value.

Last summer, Intel announced it was considering outsourcing production to other chip manufacturers. Investors saw this as a sign of weakness: Intel, one of the founders of Silicon Valley, has to consult competitors. Bob Swan, Intel’s chief executive, was speechless when he told Bloomberg how his company was dismissed by analysts as a “stunning failure.”

The gap with Taiwan and South Korea cannot be made up in the next five years. The US has two options: whether to invest more in the semiconductor industry with government support. Or accept that America has missed the boat when it comes to chip production.

There is one consolation: When it comes to chip design, hardware and software industries, US tech companies are more profitable than ever. Apple, with a market capitalization of more than $2,000 billion, is the best example of this.