Apple’s iPhone 11 and iPhone 11 Pro launch was different from all of the company’s previous launches. As the iPhone’s external appearance once again starts to stagnate, Apple has shifted focus to developing microprocessors. The company’s An processor lineup stands on its own in a world that’s got Qualcomm and Samsung as its competitors. Therefore, at the 2019 iPhone 11 launch, Apple let its vice president engineering Sri Santhanam take the stage to explain to the crowd all the changes made on the A13 over the A12.
One important detail that Mr. Santhanam shared related to the chip’s power management. According to him, the A13 has hundreds of voltage domains that are catered specifically to different workloads. You can read more on the A13 in the post that we’ve linked above. Now, Wired magazine has provided us with a handful of details for the A13 in a wall of text. Take a look below for more.
Apple’s A13 SoC Is Designed Keeping In Mind Application Power Draw Requirements Based On PPW (Performance-per-Watt)
In an interview with Apple’s VP product marketing Phil Schiller and Anandtech founder Anand Shimpi, Wired magazine has managed to gain some insight into the way that the company designs and optimizes its microprocessors for the iPhone. While designing a smartphone processor is the task of hardware architects, designing applications are the job of software engineers.
At Apple, they work together to tailor applications according to performance per watt-second. PPW is defined as the number of computations an array of transistors can carry out over the watts of power delivered to them over a second.
Apple also analyzes usage data from current iPhones and uses this data to optimize its future smartphone processor designs for performance optimizations. This provides the company with a unique advantage in power efficiency. So, if an app does not require more power offered by a newer chip, it can run on old performance parameters. As Mr. Shimpi explains, “For applications that don’t need the additional performance, you can run at the performance of last year’s and just do it at a much lower power.”
Mr. Schiller provided more details on how machine learning lets the A13 optimize power efficiency. According to the vague information, we think that the new SoC will let the iPhones will use machine learning to tailor specific silicon segments according to applications’ computational and power draw needs. This makes sure that only specific chip areas are performing the tasks, while others are resting.
Thoughts? Let us know what you think in the comments section below and stay tuned. We’ll keep you updated on the latest.
The post Apple A13 Uses Machine Learning For Performance, Battery Optimization by Ramish Zafar appeared first on Wccftech.
Refference- https://wccftech.com
0 Comments