In an interview with Wired published today, Apple’s VP of product marketing, Phil Schiller and Anand Shimpi from the company’s Platform Architecture team give a detailed look at how they are approaching chip design and improving iPhone performance each year. With the A13 Bionic, the two have highlighted how important making the chip more energy efficient is, using software to influence its silicon and how that translates to real-world benefits for users.
As Apple detailed during its iPhone 11 event earlier this month, the A13 Bionic chip powering the new iPhone lineup features 8.5 billion transistors, a total of six cores: four efficiency-focused ones and two performance-focused cores, a quad-core GPU, an octa-core neural engine, and two machine learning accelerators that can handle one trillion operations a second.
All that makes for up to a 30% more efficient chip with up to 20% faster performance compared to the A12. While Apple’s competitors now have 8-core mobile chips compared to its 6-core A13, Apple’s tight integration between its hardware and software still give it the edge when it comes to performance.
Phil Schiller and Anand Shimpi shared with Wired how they approach chip development with a focus on efficiency.
Further explaining the chip development process, the two said Apple analyzes how specific applications are used to guide its CPU designs.
Then for the applications that don’t need the extra optimization, Shimpi said that they end up using less power. The interview also unveiled that Apple applies the same approach to its GPU and machine learning development.
Another advantage Apple has with its A13 Bionic is how it processes differently than other chips.
Schiller notes that machine learning is key for these optimizations that set its silicon apart from the rest.
With the A13, think of doing the same on-and-off approach, but on a single home basis. Fewer electrons go to waste.
The full Wired interview is definitely worth a read, check it out here.
