On Tuesday, Apple held its virtual event dubbed “One More Thing” where it revealed new MacBooks and provided viewers with a closer look at the M1, Apple’s first chip designed for the Mac. Johny Srouji, Apple’s SVP of hardware technologies, led a presentation focused on the chip’s configuration and capabilities.
SEE: Apple Silicon M1 Mac buying guide: 2020 MacBook Air vs. MacBook Pro vs. Mac mini (TechRepublic)
In the past, the Mac relied on multiple components including dedicated chips for the CPU processor, I/O, memory, security, and thunderbolt controller. The M1 reimagines this internal computing configuration bringing these elements into a single system on a chip (SoC).
The chip setup utilizes Apple’s unified memory architecture (UMA). As a result, all of the components of the chips are able to “access the same data without copying it between multiple pools of memory” improving efficiency and device performance, explained Srouji. With this design, the M1 also packs 16 billion transistors, the most Apple has ever placed on a chip.
The M1’s 8-core CPU consists of four high-performance cores and four high-efficiency cores. The high-performance core is the world’s fastest CPU core “when it comes to low‑power silicon” and the high-efficiency cores use one-10th of the power, according to Apple. Compared to the “latest PC laptop chip,” the M1 is able to deliver two-times faster CPU performance and match “peak PC performance” while using one-quarter of the power.
SEE: How to migrate to a new iPad, iPhone, or Mac (TechRepublic Premium)
The M1 design also includes up to eight GPU cores capable of executing “nearly 25,000 threads” concurrently. Compared to the “latest PC laptop chip,” the M1 is able to provide GPU performance that is up to two-times faster, per Apple. The M1 is also capable of matching peak PC performance with one-third of the power consumption.
SEE: Apple’s MacBook Air with M1 chip: Everything you need to know (TechRepublic)
The M1 also boasts a 16-core neural engine. Apple estimates the latest neural engine is able to execute 11 trillion operations per second. In tandem, the neural engine, GPU, and machine learning accelerators, allow the M1 to deliver machine learning performance that is up to 15-times faster. Apple said the M1 capabilities will enable voice recognition, image processing, and video analysis at “a level of performance never seen before on the Mac.”
SEE: Apple unveils a number of health and fitness features, apps, and more (TechRepublic)
Secure enclave, ISP, and more
The M1 configuration also includes an image signal processor (ISP) to enhance video quality and decrease noise, and more. A Thunderbolt controller offers up to 40Gbps transfer speeds and a dedicated Secure Enclave boosts device security. The build also includes media encode decode engines.