The Apple A13 Bionic
“The A13 Bionic is the fastest CPU (GPU and Neural Engine too) ever in a smartphone” — Apple
Apple’s comparative advantage in the smartphone industry are the brains behind their SoC, the A-series Bionic. From the A10 to the A13, each iteration shows improvement in speed and performance. The A13 is what powers the amazing features on the standard iPhone 11, and the flagship 11 Pro and 11 Pro Max. It rivals Qualcomm’s Snapdragon, Samsung’s Exynos and Huawei’s Kirin processor.
What makes the A13 so special? Let’s take a look at the A13 under the hood.
The A13 Bionic is a 64-bit ARM-based (ARMv8.3-A) SoC designed by Apple and manufactured by TSMC. Released in 2019, it is the processor used for the iPhone 11 series of smartphones. Like its predecessor, the A12, it is built on a 7 nm process. It has an integrated quad-core GPU that has 20% faster graphics performance and 40% lower power consumption than the A12 Bionic.
The SoC features a hexa-core CPU, consisting of 2 high-performance cores running at 2.65 GHz called Lightning and 4 energy efficient cores running at 1.8 GHz called Thunder. The Lightning cores contain machine learning accelerators called AMX blocks capable of 1 trillion operations per second. This is a part of the A13’s AI features that also make use of a separate 8-core neural network engine processor.
Microarchitecture Specs
Power Efficiency
Apple continues to improve when it comes to battery performance. Perhaps this is a feature that is always overlooked because consumers are more about the latest camera and iOS features. The improvements in power efficiency allow the battery to last longer, with less drain coming from the processor chips. This was accomplished with the design of the logic circuits. What the engineers did was to turn off any logic gates that are not in use to conserve power. The 4 energy efficient cores help conserve power by only turning on voltage domain at low power while allowing more processing intensive tasks to be handled by the 2 high-performance cores. There is no exact metric on how much power you will actually save on battery, but tests have shown that there was a significant increase in battery time (in minutes).
Optimized AI
As competition grows among smartphone vendors, implementing AI has become integral features on flagships. The neural network engine is what provides advanced AI capabilities that use machine learning to understand the user and provide the magic behind Face ID, camera portrait modes and Siri. The AMX blocks perform the required matrix calculations used for AI since they are faster than a CPU. There is a controller inside the SoC that is responsible for coordinating operations with the CPU, GPU and Neural Engine processor. This is what makes the A13 an optimized chip for AI (i.e. machine learning).
Getting Faster in CPU And GPU Performance
To find out how powerful the A13 Bionic is, a benchmark test result from Geekbench and 3DMark shows the data.
Let’s take a look first at the Geekbench results for single-core and multi-core processor performance.
It looks quite impressive how much leap in performance the A13 Bionic has over the previous versions (A11 on the iPhone X and A12 on the XS). Now here is a comparison with the competition.
At 6291 compute, none of the other flagships in this comparison came close. The A13 Bionic also dominated in single-core and multi-core performance.
The GPU was not exactly Apple’s strength when it comes to performance. Android-based smartphones tend to perform much better. Based on Macworld’s 3DMark Fire Strike tests, the A13 Bionic was 50–60% faster than previous A-series chips.
The results show this makes the A13 Bionic the fastest performing GPU in the market (as of this posting) as of October 2019. This seems to be due to wider memory data paths along with the GPU’s core design that gives it the advantage. It remains to be seen how much more improvements can be made with the next A-series chip.
The “Apple’s Edge”
You can forget about the criticisms of Apple’s iPhone 11 line up. Instead pay attention to the hardware that runs these phones because they are truly the innovation that can enable more compute intensive applications and speedier processing of features. Face ID unlock becomes more responsive and faster to recognize a user’s face, better machine learning allows the camera to take better photos and the efficiency of the processor helps conserve battery life. While the iPhone itself is becoming outdated and the design has not been too impressive, it is the technology within that really gives Apple the edge.
