Apple’s two-chip strategy for the iPhone 14 makes a lot of sense

A few weeks ago, analysts started predicting that Apple would take a totally new tactic with its iPhone 14 lineup. The regular iPhone 14 (which is said to come in two sizes, standard and Max) would continue to use the A15 processor that was found in the iPhone 13 Pro this year. The iPhone 14 Pro (also in standard and Max sizes) would get the new A16 processor.

It’s a believable rumor, and a natural evolution of the differentiation Apple started in the iPhone 13 lineup this year. For the first time in memory, the “standard” iPhone models got a version of the A15 that is less capable than the Pro models: the A15 has four GPU cores in the standard iPhone 13s, while the Pro models have five GPU cores. There’s also a difference in RAM–the standard models have 4GB and the Pro models have 6GB–but that’s not really new, and Apple doesn’t even disclose the amount of RAM in iPhones.

Binning and supply chain challenges

To be clear, the iPhone 13 and iPhone 13 Pro actually have the same A15 chip. The bigger and more complex a processor is, the more defects will be present on the wafer. Manufacturers reduce this defect rate with redundant circuitry and with a process known as binning: taking chips with defects, disabling the defective core or cache, and selling it as a lesser/cheaper part. This is nothing new: CPU and GPU manufacturers have done this as regular practice for years.

Apple simply takes the A15, which has five GPU cores, and disables one of the GPU cores in the chips that have a defect in that area of the chip. This lets them get more usable processors per wafer, thus reducing costs. Apple simply puts the chips with four enabled cores in the iPhone 13, and the chip with all five enabled in the iPhone 13 Pro. This extends to the M1, which has 8 GPU cores but is available with just 7 in the most affordable MacBook Air and iMac. The M1 Pro and M1 Max have binned versions as well.

But what do you do when binning isn’t enough? When your A16 processor requires a cutting-edge manufacturing process that is in very short supply, and super expensive? When binning isn’t enough to produce over 70 million new iPhones a year? You go beyond binning, and build a new iPhone with last year’s A15 processor.

IDG

More than enough performance

Frankly, even the very best processors in Android phones today can’t compete with the A15. Apple is so far ahead that it doesn’t need to put a faster processor in its $799 model this fall just to remain competitive. Especially if, as the rumors indicate, these iPhones get the “full” Pro version of the A15 with five GPU cores and 6GB of RAM. That would still constitute an upgrade over the iPhone 13 (non-Pro models) and likely still be quite a bit faster and more efficient than any comparably-priced Android phone.

The iPhone 14 Pro, on the other hand, gets a brand new processor with performance and features that lead the industry, helping to further separate the “Pro” from the “non-Pro” models. Everyone gets an upgrade, Apple gets enough chips to actually meet demand, and we all win.

What’s in a name?

That is, everybody wins as long as this latest rumor isn’t true. It claims that Apple will rename the A15 (the 5-core GPU, 6GB RAM version) as the “A16” while naming the new chip “A16 Pro.” We saw something similar with the Apple Watch Series 7, with Apple claiming on the spec sheet that it has an S7 chip despite tests showing it’s identical to the S6.

This would be a terrible idea. It makes a lot of sense to differentiate binned versions of a chip with different names–perhaps Apple should have gone with A15 and A15 Pro for the four-core and five-core versions of that chip. But selling a chip this year under the name A15 and next year under the name A16, with no substantive changes, is pure marketing nonsense meant to confuse consumers into thinking they’re getting something new.

It seems likely to me that Apple would re-use the existing A15 chip (the full-scale, 5-core GPU version) in the iPhone 14, reserving the A16 for the Pro model. It’s a smart move in a supply constrained environment, especially when they company’s chips are so far ahead of the rest of the market. Let’s just hope the company doesn’t try to pull a fast one with the naming.