Mobile chip giant Qualcomm (QCOM 0.73%) has been running into some notable headwinds lately on multiple fronts. One of the more notable challenges to its chip business is that its top two customers, Apple (AAPL 0.64%) and Samsung, continue to slowly but surely transition toward in-house designs. Qualcomm relies on the pair for nearly half of total revenues (including licensing and royalty revenue).

Samsung made a high-profile switch from Snapdragons to its own Exynos processors in the Galaxy S6, which put a dent in sales this fiscal year. Earlier this year, Samsung mobile chief J.K. Shin said it could potentially use Qualcomm chips in upcoming Galaxy phones, but, alas, the trio of devices that Samsung just launched earlier this month are all powered by Exynos processors. There had been rumors that the Snapdragon 808 would win a slot or two, but the Galaxy Note 5, Galaxy S6 Edge, and Galaxy S6 Edge+ all have an Exynos 7420 inside. There's always next year's Galaxy S7.

As far as Apple goes, Qualcomm has long supplied the cellular baseband inside to complement Apple's A-chip processors. Not only is Apple likely building its own cellular baseband, a move that's arguably long overdue, but Intel (INTC -0.38%)has been actively vying for that spot for years, and each year Intel's chances seem to improve as its modem performance improves. There's been a lot of debate over whether Intel has won some of this year's iPhone modem business or if it might win a slot next year. Even if Apple's own designs are years away from implementation and Qualcomm is able to defend the slot, Apple will naturally use Intel's competition as leverage and likely negotiate volume pricing concessions from Qualcomm, putting pressure on sales and margins.

As if these headwinds weren't challenging enough, another important chip customer might be looking to do likewise.

Xiaomi wants more control over its chip road map
China's Xiaomi has been successfully winning share in recent years thanks to its aggressive pricing of high-end hardware. That includes using Snapdragons in flagship devices like the Mi Note Pro (Snapdragon 810) or mid-range phones like the Mi 4i (Snapdragon 615). Meanwhile, Xiaomi puts more affordable MediaTek processors in its lower-end phones. But recent reports suggest that Xiaomi may try to shift to in-house chips, too, in the foreseeable future.

Xiaomi planted the seeds of this possibility earlier this year when it partnered with Leadcore to develop the processor that ended up in the low-end Redmi 2A. That was a faster solution than building an in-house team from scratch, while at the same time being able to direct the development of the chipset. While the Leadcore LC1860 processor powering the Redmi 2A is fairly modest in performance and wasn't necessarily anything to brag about, it's easy to see the appeal in developing and implementing a semi-custom chip to power future flagships.

Not only would Xiaomi have more say in its chip road map, but it could brand or co-brand the processors as a point of differentiation. Android phones, especially those made by low-cost Chinese OEMs, are incredibly vulnerable to commoditization, and differentiating based on a core component like the processor could go a long way.

These types of vertical integration strategies often take many years to play out, and Xiaomi is supposedly going to start with its low end and mid-range in 2016, but eventually it's conceivable that it will continue moving up as its ambitions grow. I wouldn't ever expect Xiaomi to be able to rival Qualcomm in terms of sheer technical prowess and performance, since Qualcomm's custom chips are incredibly sophisticated, but it's possible that Xiaomi can eventually develop a processor that's sufficient for its purposes while realizing the benefits of vertical integration.

Qualcomm's chip business has been having a hard time recently, with sales falling 22% last quarter. Chip sales can certainly rebound, but it might not be easy.