Developer Explores Why Apple’s M1 Chip Is So Fast

Apple’s M1 chip is the fastest chip Apple has ever released on a Mac based on single-core processor benchmarks, and it beats many high-end Intel Macs in multi-core performance. Developer Erik Engheim recently shared an in-depth dive into the ‌M1‌ chip, exploring why Apple’s new processor is so much faster than the Intel chips it replaces.

First and foremost, the ‌M1‌ is not a simple processor. As Apple explained, this is a System-on-a-Chip, which is a series of chips that are all bundled together in a single silicone case. The ‌M1‌ houses an 8-core processor, an 8-core GPU (7-core in some MacBook Air models), unified memory, an SSD controller, an image signal processor, Secure Enclave, and much more.

Intel and AMD also ship multiple microprocessors in a single package, but as Engheim describes, Apple has a head start because rather than focusing on general-purpose processor cores like its competitors, Apple is focusing on specialty chips. who manage specialized tasks.

In addition to the processor (with high performance and high efficiency cores) and GPU, the ‌M1‌ has a neural engine for machine learning tasks such as speech recognition and camera processing, video decoder / encoder built-in for energy efficient conversion of video files, Secure Enclave to handle encryption, digital signal processor to handle mathematically intensive functions like decompression of music files and image processing unit which speeds up tasks performed by image processing applications.

Notably, there is also a new unified memory architecture that allows the CPU, GPU, and other cores to exchange information with each other, and with unified memory, the CPU and GPU can access memory simultaneously rather than copy data between one zone and another. Accessing the same memory pool without copying speeds up information exchange for faster overall performance.

All of these specific purpose chips speed up specific tasks, leading to the improvements that people see.

This is part of the reason why many people working on images and video edits with Mac ‌M1‌ see such speed improvements. Many of their tasks can be performed directly on specialized hardware. This is what allows a cheap Mac Mini ‌M1‌ to encode a large video file, without breaking a sweat, while an expensive iMac blows up all of its fans and still can’t keep up.

Specialized chips have been in use for years, but Apple is “taking a more radical turn in this direction,” as Engheim describes it. Other Arm chipmakers like AMD take a similar approach, but Intel and AMD rely on selling general purpose processors and for licensing reasons PC makers like Dell and HP are unlikely to be able to design a complete in-house SoC like Apple is capable of doing this. .

Apple is able to integrate hardware and software in ways that most other companies simply cannot duplicate, which still gives the iPhone and iPad an edge over other smartphones and tablets.

Of course, Intel and AMD can just start selling full SoCs. But what should they contain? PC manufacturers may have different ideas on what to contain. You might have a conflict between Intel, AMD, Microsoft, and PC makers over what kind of specialized chips to include, as they will require software support.

In addition to the benefits of an in-house designed system on a chip, Apple also uses Firestorm processor cores in the ‌M1‌ which are “truly fast” and able to execute more instructions in parallel via out-of-order execution. RISC architecture and some specific optimizations implemented by Apple, of which Engheim has an in-depth explanation.

Engheim believes that Intel and AMD are in a difficult situation due to the limitations of the CISC instruction set and their business models that make it difficult to create end-to-end chip solutions for PC manufacturers.

Engheim’s full article is worth reading for those interested in how the ‌M1‌ works and the technology Apple has adopted to take a giant leap in computing performance.