⚠ The Modem ProblemLow threat
Apple (AAPL) — threat to the moat
The cellular modem Apple struggled for years to build shows the limits of doing everything in-house.
Apple's strategy of capturing supplier margin by bringing components in-house has been a triumph almost everywhere — except the cellular modem, where a long, costly, and humbling effort has repeatedly fallen short. For years Apple has wanted to design its own modem and stop paying Qualcomm's markup on a critical part of every iPhone1, and for years the sheer difficulty of matching Qualcomm's decades of wireless patents and know-how has forced it, grudgingly, to keep buying. The modem is the exception that proves how hard the in-house game really is.
The danger this illustrates is broader than one chip. Vertical integration looks inevitable in hindsight for the parts Apple has conquered, but each new component is a fresh and expensive bet that may not pay off, and the modem stands as a reminder that some suppliers hold moats of their own that Apple cannot simply spend its way through. Every in-house effort that stumbles is money spent and margin not captured, while the supplier keeps its cut.
To its credit, Apple has at last begun shipping its own modems in some products2 and will, in time, likely reduce its dependence; even where it still buys, its scale wins favorable terms, and the components it has successfully brought in-house vastly outnumber the handful it has not. The direction of travel remains toward more capture, not less: Qualcomm has told its investors that its Apple-related revenue will fall by roughly half between the September and December 2026 quarters3.
A low, narrow worry. This is not a threat to the business so much as a boundary on the strategy — a reminder that supplier-margin capture is a hard-won advantage rather than an automatic one, and that a few well-defended suppliers can hold their ground and keep their margin even against a customer as formidable and determined as Apple.
- ReportedApple bought Intel's smartphone-modem business in 2019 (~$1B) and took until 2025 to ship its own C1 modem, in the iPhone 16e.Apple — C1, Apple's first in-house cellular modem (debuted in iPhone 16e); follows Apple's 2019 purchase of Intel's smartphone-modem business (~$1B) — C1 shipped Feb 2025; Intel modem deal 2019 · publ. 2025 · source ↗
- ReportedQualcomm remained the supplier of iPhone modems throughout that period.Qualcomm — Snapdragon (Oryon CPU cores from the 2021 Nuvia acquisition, a firm founded by former Apple chip engineers) — Nuvia acquired 2021; Oryon-based Snapdragon from 2023 onward · publ. 2021–2026 · source ↗
- ReportedQualcomm guided Apple-related revenue down roughly half between the September and December 2026 quarters.Coverage of Qualcomm's third-quarter fiscal 2026 results, July 2026 — revenue of $9.9 billion and non-GAAP earnings per share of $2.21, meeting the high end of revenue guidance and short of consensus on earnings. QCT handset revenue of $5.09 billion declined 20% year over year on industry-wide memory supply constraints, higher input costs affecting device pricing and demand, and inventory adjustments at major OEMs; automotive revenue of $1.59 billion rose 61% and IoT of $1.83 billion rose 9%. Management said a sharper-than-expected reduction in Apple modem-related sales contributed to a lighter fourth-quarter outlook, with Apple-related revenues expected to fall by approximately 50% between the September and December quarters, and doubled the fiscal 2029 non-handset revenue target to $40 billion. Fourth-quarter guidance was $9.7-10.5 billion of revenue and $2.05-2.25 of non-GAAP diluted EPS, both below consensus near $10.02 billion and $2.36. — Q3 FY2026 · publ. 2026-07-30 · source ↗