This year's M9 is directly embedded in the A9. Image source: Apple.

Apple (AAPL 1.27%) first started putting its branded M-chip motion co-processors into the iPhone starting in 2013 with the iPhone 5s. The purpose of that model's M7 co-processor was to collect, process, and store data from many of the onboard sensors like the gyroscope, accelerometer, and compass, as a low-power alternative to using the primary A7 chip inside. The same is all true of the A8 and M8 found in the iPhone 6 and 6 Plus. This all translates into greater power efficiency, which in turns helps battery life.

But the Mac maker is switching it up this year, and it's possible that NXP Semiconductor (NXPI 4.18%) may suffer for it.

It's been a good two years
As it turns out, both the M7 and M8 were actually chips built by NXP. After the initial iPhone 5s teardowns, industry observers at first had difficulty finding the M7, since they were looking for a separate chip labeled with M7 branding. Chipworks eventually figured out that it was actually a custom-built Cortex-M3 microcontroller made by NXP. The M7's model number is the NXP LPC18A1. The following year, NXP similarly provided the M8 motion co-processor (model number NXP LPC18B1UK), according to the iFixit teardown.

That means that NXP has been gradually winning share in iPhones. It sourced the M7 in the iPhone 5s, while in the iPhone 6 and 6 Plus it supplied both the M8 and the NFC module and secure element that make Apple Pay possible. But this year's M9 motion co-processor is being directly integrated into the A9 chip. Does that mean Apple has cut NXP out of the motion co-processing loop?

Apple can roll its own from here on out
With the M9 being directly embedded into the A9, it doesn't appear that Apple is buying discrete M-chips from NXP in the iPhone 6s and 6s Plus. At the same time, since the NXP parts were simply ARM Cortex-M3 microcontrollers, it's unlikely that Apple would need to license any designs from NXP itself, since both companies already license designs from ARM Holdings (ARMH).

Note the M9 moving inside the A9. Image source: Apple.

Apple's chip team is advancing with breathtaking pace. It's incredible to think of how much progress Apple's chip strategy has made since it first started using A4 chips in 2010. In just five years, Apple has attained desktop-class performance in its smartphones. So it's completely conceivable that over the course of two years, Apple's chip team could integrate a Cortex-M3 microcontroller directly into the A9.

The Cortex-M3 is a relatively simple off-the-shelf design that Apple should have no problem implementing. On top of that, there are probably additional power savings by integrating the M9 directly.

All is not lost
Even if NXP has lost the motion co-processor spot, there is comfort in the fact that it is very likely still supplying the NFC and secure element modules. NXP is by far the industry leader when it comes to NFC technology, and this is unlikely an area where Apple will wade into directly. At the same time, Apple Pay is of critical importance to iDevices going forward and is likely being included in all future iPhones and Apple Watches (NXP also provides the NFC modules in Apple Watch).

Besides, NXP could very well benefit from a broader adoption of NFC that Apple could be catalyzing.