The iPhone 14 Pro saw the debut of a new Apple system-on-a-chip, the A16 Bionic. So what kind of improvements over the A15 does it bring, and where else can we expect to see the A16 implemented?
A16 Arrived With the iPhone 14 Pro
Apple’s flagship mobile chip for 2022 was the A16 Bionic, which powers the high-end iPhone 14 Pro and Pro Max models. The iPhone 14 instead uses a slightly improved version of the A15 Bionic from 2021’s iPhone 13 (with an additional GPU core). This could be a result of a global semiconductor shortage and pressure on manufacturing processes due to the COVID-19 pandemic.
It’s the first time Apple has ever used different generations of system-on-chip for its numbered and “Pro” iPhone models. You might think this indicates that there’s not a lot of difference between them, but you’d be wrong. On the surface, you can see the chip has an identical number of CPU and GPU cores (6 and 5, respectively) as the A15 Bionic used in the iPhone 14.
But Apple claims that the A16 has a faster CPU thanks to its nearly 16 billion transistors (up from 15 billion on the previous model), though Apple provides no statistics to back this up. Transistors are to microprocessors what neurons are to the human brain, so the more of them you have available the better.
Another area that has been improved is the memory bandwidth of the GPU. Even with the same number of GPU cores as the current A15, improved memory bandwidth will translate into an improvement in performance on the new chip.
Early Geekbench 5 scores point to a modest 10% faster performance in single-core operations, with little difference in multi-core scores. Even then, synthetic benchmarks designed to test raw performance in a set of controlled tests don’t always reflect on the real-world performance of the chip.
But we can assume a few things with some degree of confidence. The A16 Bionic will be a better overall chip than what came before it, especially in GPU-intensive tasks thanks to the additional memory bandwidth. It’s also likely to make its appearance in 2023’s iPhone 15, and possibly a few devices in between. (iPad refresh, anyone?)
A16 and the Quad-Pixel Sensor
Apple has harnessed the power of the A16 in one clear area for the iPhone 14 Pro, and that’s the new quad-pixel sensor. This is Apple’s 48-megapixel main camera sensor, which captures four times as many pixels as the previous generation iPhone 13 Pro.
An upgraded image signal processor (ISP) improves photo and video capture across the board, with “four trillion operations per photo” quoted by Apple in its event. Apple says that the Deep Fusion technology introduced with the iPhone 11 can occur sooner on uncompressed images, hence the quality improvement.
The 16-core Neural Engine uses Apple’s machine learning algorithms to make AI-powered enhancements to photos, further boosting image quality.
In terms of other improvements to graphical components, Apple points to the A16’s new display engine as being pivotal for making the always-on-display function. The iPhone 14 Pro display also features 1 Hz refresh rates, 2000 nit peak brightness in full sunlight, and antialiasing to smooth out jagged lines for a sharper image.
Apple’s First 4nm Chip
Perhaps the biggest achievement on Apple’s part is the fact that the A16 Bionic is the first chip that the company has brought to market that uses the 4 nanometer (nm) process. This is in line with industry aspirations to manufacture smaller and smaller components, which has several benefits.
Smaller chips are seen as being more power efficient, which could explain Apple’s claims that the A16 will use 20% less power than the A15 that came before it. A more efficient chip means better battery life and less power “wastage” in the form of heat.
A smaller process also requires less silicon, which means that yields can be higher. This reduces the costs associated with production, though don’t hold your breath for a price cut any time soon thanks to the research and development costs associated with silicon fabrication (and the small matter of a global semiconductor shortage).
As we said in our 2019 explainer about Intel’s 10nm process:
“The way each semiconductor foundry measures can vary from one to another, so it’s best to take them more as marketing terms used to segment products rather than exact measurements of power or size.”
The move to 4nm is significant, but don’t read too much into this change from a performance standpoint.
A16 and M2: Apples and Oranges
It’s hard to talk about the A16 without mentioning Apple’s desktop-class chips, particularly the newest M2 which appears in the refreshed MacBook Air and 13-inch MacBook Pro models.
The M1 and M2 use the 5nm processor, as opposed to the 4nm process used by the A16 Bionic. Apple’s M2 chip packs in 20 billion transistors, with an 8-core CPU, up to 10-core GPU, and dedicated ProRes encode and decode.
These chips may share the same ARM-based architecture, but they’re designed with different tasks in mind. Mobile processors (like the A16) are designed to be powered by a much smaller battery, so they must be more power efficient than their notebook or desktop counterparts.
These differences are further exemplified by the fact that the mobile processor has a GPU that’s half the size of the M2 and features 2 CPU cores dedicated to performance and 4 cores dedicated to efficiency (as opposed to 4 performance cores and 4 efficiency cores on the M2).
Power efficiency has been one of the big upshots of switching to an ARM-based architecture on the Mac desktop, but the move also made it possible for Apple to deliver big performance gains. This saw the Mac Studio (with its M1 Ultra) claim the title of Apple’s “most powerful computer” despite the Intel Xeon-powered Mac Pro.
iPhone Pro Only, For Now
The iPhone Pro is Apple’s high-end iPhone that receives all the newest and most exciting features first. In 2021 this included a new ProMotion display, and before that a LiDAR scanner, while in 2022 it’s the A16 chip and the evolution of the notch known as the Dynamic Island.