Each year, Apple (AAPL 0.96%) releases new iPhones, and inside are new processors that offer performance and feature enhancements over their predecessors. On Sept. 12, Apple continued that tradition, announcing three new iPhones: iPhone Xs, iPhone Xs Max, and iPhone XR -- each packed with a new, Apple-designed system-on-a-chip called the A12 Bionic.
Here are three key things you need to know about Apple's latest silicon creation.
Built using 7-nanometer tech
In its press release, Apple boasted that its new trio of iPhones incorporate "the first 7-nanometer chip in a smartphone." Since Taiwan Semiconductor Manufacturing Company (TSM -1.25%) is the only company that's currently mass-producing 7-nanometer chips, it's a safe bet that the Taiwan-based contract chip manufacturing giant is responsible for cranking out Apple's latest chip design.
Apple says that the A12 Bionic chip is made up of 6.9 billion transistors -- a whopping 60% increase in transistor count from the prior-generation A11 Bionic. According to TSMC, its 7-nanometer technology has 1.6 times the logic density of its 10-nanometer technology, which was used to build the A11 Bionic. It would appear that migrating from TSMC's 10-nanometer technology to its newer 7-nanometer technology enabled Apple to dramatically boost the number of transistors that it used in crafting the A12 Bionic without leading to a significant generation-over-generation increase in chip size.
CPU and GPU improvements
The A12 Bionic, like its predecessor the A11 Bionic, incorporates six processor cores -- two so-called performance cores and four efficiency cores. According to Apple, the two performance cores are "up to 15% faster" than the ones found in the A11 Bionic and use 40% less power. The efficiency cores, per Apple, are as much as 50% more power-efficient.
Apple didn't cite a performance increase for the efficiency cores this time around (last year, Apple claimed a 70% increase in performance for the efficiency cores in the A11 Bionic relative to those in the A10 Fusion), so it looks like Apple's trying to tell us that the efficiency cores in the new A12 Bionic can deliver the same performance as the ones in the A11 Bionic but use far less energy to do so. This makes sense considering that the whole point of the efficiency cores is to process tasks that aren't particularly performance-sensitive while using as little energy as possible to maximize battery life.
Apple also revealed that the A12 Bionic has a substantially faster Apple-designed graphics processing unit (GPU), which is "up to 50% faster" than the one incorporated in the A11 Bionic. You'll recall that Apple used graphics processor designs from Imagination Technologies through the A10 Fusion, but switched to internally-designed technology beginning with the A11 Bionic.
Last year, Apple revealed that it had integrated a piece of technology called a Neural Engine into its A11 Bionic chip. In its press release announcing the iPhone X, Apple explained, "The new A11 Bionic neural engine is a dual-core design and performs up to 600 billion operations per second for real-time processing. A11 Bionic neural engine is designed for specific machine learning algorithms and enables Face ID, Animoji and other features."
Apple seems to have gone all-out with the second-generation Neural Engine. The Neural Engine in the A12 Bionic has eight cores (up from two in the A11 Bionic) and can "complete up to 5 trillion operations per second." That's more than 8.3 times the peak performance of the prior-generation Neural Engine.
Perhaps more importantly, Apple says, "For the first time the Neural Engine is open to Core ML, empowering developers to build apps that utilize this highly efficient machine learning engine." This seems like an area where Apple's control over the hardware and the software ecosystem as well as its tight integration of the two can really translate into a competitive advantage for the company's products relative to what the Android ecosystem can produce.