A brand-new processor that Apple developed in-house will power the newest Macs. Here’s what it can do, based on what we know so far.
By Tom Brant
For the first time in nearly 15 years, Apple’s newest MacBooks and Mac mini don’t come with Intel processors. Instead, they use the brand-new Apple M1 chip, unveiled on Tuesday as a powerful replacement for the many generations of Intel CPUs that have powered Apple computers since 2006.
The M1 is the first appearance of the new paradigm that Apple has dubbed Apple Silicon. The new Macs available for pre-order now with the M1 include the MacBook Air, the 13-inch MacBook Pro, and the Mac mini. If you’re contemplating buying one right out of the gate, you’re entering uncharted waters, since the only information we have so far about how well they perform comes from Apple itself. We’re looking forward to getting these new devices into PC Labs ASAP for some extensive testing and hands-on time. But in the meantime, let’s take a look at everything we know so far about Apple Silicon in the Mac.
The M1: An Entire System on a Chip
Whether they run Windows, macOS, or Chrome, most of today’s PCs come with an array of computing components inside that each handle different processing tasks. These include one processor (the CPU) for handling essential computations, including those used for browsing the internet and opening and closing apps, and another (the GPU) for processing graphics computations and outputting a signal to your monitor or laptop screen.
On large, powerful machines like workstation desktops and gaming rigs, the CPU and GPU are completely separate items that are each connected to the motherboard. Smaller ultraportable laptops typically combine the CPU and GPU into a single component using what’s known in the industry as integrated graphics processing, with the rest of the computer’s components, including cache, memory, and storage, located in different spots on the motherboard.
As an evolution of the Apple A-series processors that have long powered the iPhone and the iPad, the M1 chip takes a different approach. Instead of a collection of separate processing parts, it’s a single system on a chip (SoC). The SoC handles all of the computations, including graphics output, which means that each software instruction can use the most efficient part of the M1.
In part because it must do everything at once, the M1 has an eyebrow-raising maximum of 16 processor cores. That sounds like a ton compared with the six cores that are in the most powerful Intel laptop CPUs. In fact, the M1 has so many cores not so it can perform tasks more quickly, but so it can perform more tasks. Four of the cores are compute cores dedicated to complex calculations that require lots of processing power. Four more are dedicated to lighter tasks that don’t require as much power, to ensure that the chip doesn’t consume more energy than it needs to. Tasks get shunted to the appropriate core set on the fly.
As many as eight additional cores are dedicated to graphics processing, similar to how Intel’s Iris integrated graphics work. On the Mac mini, the MacBook Pro, and the MacBook Air, the graphics part of the M1 is capable of powering an external monitor at 60Hz and up to a 6K resolution, such as the Apple Pro Display XDR.
Additionally, the M1 has a built-in storage controller to traffic data to and from the Mac’s solid-state drive (SSD), as well as various other processors, controllers, and sensors that handle encryption, image processing from webcams, and other secondary tasks that are required for the computer to function.
The M1 chip uses a 5-nanometer production process, similar to the latest A14 chip in the iPhone 12. Meanwhile, Intel’s latest 11th-generation CPUs use a 10-nanometer production process, and Intel doesn’t expect new chips based on 7-nanometer processes or lower until 2022 at the earliest. Intel said in a statement Tuesday that its CPUs “provide global customers the best experience in the areas they value most, as well as the most open platform for developers, both today and into the future.” But the big disparity in process technologies does speak for itself. Computer chip rival AMD’s latest desktop and mobile chips are on a 7-nanometer process.
Artificial Intelligence and the M1 Chip
Artificial intelligence (AI) and machine learning (ML) algorithms are essential to helping modern software run smoothly. Apple A-series and Intel Core processors have long had AI capabilities built in, and the M1 is no different. It has a dedicated neural engine with 16 processing cores of its own to handle AI tasks. Ever touched up a photo using an automatic filter, or scrolled through a list of pre-populated search results? Both of these tasks and many others often rely at least in part on ML to run faster.
The M1 chip’s neural engine relies on instructions from software to work properly. Many third-party software developers integrate AI and ML algorithms into their apps, including many parts of the Adobe Creative Suite. Apple says that the new M1-powered MacBook Air can handle ML workloads up to nine times faster than the previous Intel-powered MacBook Air.
Are All M1 Chips Created Equal?
No matter whether a computer uses an SoC or a collection of separate computing components, it’s still subject to certain unbreakable properties of computing and physics. The more cores a processor dedicates to a certain task, and the faster each of those cores run, the faster the task will complete. This process also generates a lot of heat, which is why most computers have fans, heat sinks, and other equipment.
While Apple hasn’t released many details yet, we do know that there will be slightly different versions of the M1 for each of the different Macs it powers. This is mainly due to the fact that each device handles heat differently. The new MacBook Air has no cooling fan, and we do know that the M1 chip in the base configuration of the Air will feature a slightly less-powerful graphics processor, with seven cores instead of eight.
Meanwhile, the larger MacBook Pro and the Mac mini will both come with cooling fans, which enables their M1 chips to have full-powered GPUs. It’s possible that the Mac mini and MacBook Pro M1 versions will also have higher clock speeds than the MacBook Air’s chip, but Apple has not shared this information.
How Fast Is the M1 Chip?
According to Apple, the M1 chip is faster than the Intel processors in the Macs it replaces. In some cases, the company claims, it’s much faster. General performance of the M1 MacBook Air is 3.5 times faster than before, while graphics performance is five times faster, Apple says. On the M1 MacBook Pro, AI computation is 11 times faster than before, and Apple says that it can build code in the Xcode app up to 2.8 times faster.
Apple also claims that M1-powered Macs are faster than their Windows counterparts. The company estimates that the new M1 in the MacBook Air will outperform 98 percent of PC laptops sold in the past year. These claims are based on performance tests that Apple performed in-house, though the company isn’t providing more details about them except to say that they are “industry-standard” benchmarks.
Benchmarks are easily manipulated, however, so it’s too soon to say definitely whether the M1 is as capable as Apple says it is. We’re eager to try out the M1 using our own objective benchmarking process. If Apple’s claims ring true, the M1 is indeed poised to surpass the current crop of Intel-powered Macs.
Apple also claims that the M1 will offer excellent battery life in the MacBook Air and the MacBook Pro. In practice, previous models of the Air and Pro with Intel chips already offer excellent battery life, typically between 10 and 20 hours of light tasks like watching videos.
Will My Apps Run on an M1 Mac?
A chip is only as good as the software that runs on it, and many macOS apps are designed to run on Intel processors, not the M1. Developers have had a bit of a head start, as Apple offered them a development kit this summer to help them translate their code to the M1. But the fact remains that while the macOS operating system itself will run natively on the M1, many third-party apps will not, at least initially.
The lack of native M1 support doesn’t mean third-party apps won’t run at all, however. For apps that have yet to make the transition, Apple says the programs will still be able to run on the new Macs through the company’s Rosetta 2 software, which can act as an emulator. But don’t be surprised if the programs drag. “The translation process takes time, so users might perceive that translated apps launch or run more slowly at times,” Apple wrote in the developer documentation for Rosetta 2.
Meanwhile, Apple confirmed on Tuesday that starting with macOS Big Sur, all of its Mac software runs natively on both Intel and M1 systems. As an added benefit, iPhone and iPad apps can now run directly on the M1-powered Macs. Big Sur will come standard on all new M1 systems, and it will be available as an update for older Macs later this week.
What If I Want an Intel Mac Instead?
While Apple silicon is clearly the future for the Mac ecosystem, Intel-powered Macs aren’t going away overnight. The new MacBook Air and Mac mini use the M1 exclusively, but Apple continues to offer the 13-inch MacBook Pro with Intel Core i5 or Core i7 CPUs as options. And the other Macs in Apple’s lineup-including the iMac, the Mac Pro, the iMac Pro, and 16-inch MacBook Pro -were not updated on Tuesday and are still exclusively powered by Intel CPUs and (in some cases) optional AMD GPUs.
Originally published at https://www.pcmag.com.