During Apple's (AAPL 0.64%) Sept. 12 product launch event, the company revealed an interesting detail about its custom-designed A11 Bionic chip that's at the heart of every iPhone 8, iPhone 8 Plus, and iPhone X: transistor count. Apple told the world the A11 Bionic chip is made up of 4.3 billion transistors -- about a billion more than what the company's prior-generation A10 Fusion chip packed.

Thanks to a new 10nm chip manufacturing technology, each transistor in the A11 Bionic chip takes up less area than a transistor in the A10 Fusion did, which was built on a 16nm technology -- roughly half as much, if you believe the marketing claims from the manufacturer of Apple's A10 Fusion and A11 Bionic chips, Taiwan Semiconductor Manufacturing Company (TSM 2.84%).

Based on those claims, as well as the transistor density of another mobile chip built using the same technology that's used to build the A11 Bionic, I estimated that the A11 Bionic chip size would measure at around 78 square millimeters.

Apple's iPad Pro on the left, with an iPhone on the right.

Image source: Apple.

As it turns out, that estimate was on the low side, thanks to a teardown report from TechInsights.

A surprisingly large chip

According to TechInsights, which routinely tears down new mobile devices and digs into the chip technology inside them, the A11 Bionic chip measures in at 89.23 square millimeters -- about 15% bigger than what I'd previously estimated. It's still smaller than the A10X Fusion chip, which is built on the same TSMC 10nm technology the A11 Bionic chip is, but the delta is far smaller than I'd originally thought it'd be.

Let's take a closer look at what the implications of this should be for Apple and TSMC.

Key business implications

All else equal, larger chips are more expensive to manufacture, driven by two major factors. The first is that contract chip manufacturers like TSMC tend to charge by the silicon wafer. Since the size of a silicon wafer is fixed, the larger a chip, the fewer chips can be produced per wafer. That means the larger a chip is, the more wafers the customer will need to purchase to get the number of chips that it needs.

Second -- and again with all else equal -- smaller chips tend to yield better than larger chips; that is, a greater percentage of the chips produced prove salable. 

So the fact that the A11 Bionic is an 89-square-millimeter chip and not 78 square millimeters chip means TSMC will sell more 10nm wafers to Apple than I'd originally thought. That means more revenue for TSMC and higher per-chip costs for Apple.

I don't want to give the impression that Apple should strive to build its chips as small as possible just to maximize its margins; additional features and capabilities require more transistors, which ultimately mean bigger chip sizes. Apple should strive to endow each new generation of chips with compelling new features and capabilities.

Moreover, considering the A11 Bionic will be wrapped inside phones that cost between $699 and $1,149, a slight increase in chip cost to Apple in exchange for delivering a markedly superior user experience is a trade-off that's easy to make.