Qualcomm's Bold Trade: Exiting Apple to Catch the Data Center Wave

Memory shock and an accelerated Apple exit are short-term pain for a company betting ~$15B on data center silicon — the question is execution.

QCOM · Earnings Call · 2026-07-29

The Pivot, In Numbers

QUALCOMM's July 29 report was a study in strategic whiplash. The headline — $9.9B revenue at the high end of guidance, EPS of $2.21 — papers over two structural moves happening at once: an accelerated departure from Apple's modem business and a doubling-down on the data center. The intent is clearest in the June Investor Day targets: data center revenue targeted at $15B by fiscal 2029, lifting the non-handset ambition to $40B — nearly double what was pledged just a year earlier. The closed acquisition of Modular Inc. adds a full-stack AI software layer to the silicon story. The timing, management insists, is deliberate, as inference disaggregates and moves toward the edge. “This is the ideal and logical time for Qualcomm to enter the market as agentic workloads are reshaping the economics of AI.” — Cristiano Amon, President and Chief Executive Officer · 2026-07-29 Two custom silicon wins go into production in the December quarter with wafers already started, and the HBC Gen 1 tape-out — a high-bandwidth compute solution co-packaging memory and compute — is complete. The AI accelerator road map is the long pole, with management describing "active conversations with nearly every leading data center player" ahead of silicon demonstrations in coming quarters. This pivot has been hardening across calls. Back in November, data center revenue was only promised to become "material" in fiscal 2027, and solely off the back of the Humane engagement — a timeline the company has now pulled forward and broadened to two hyperscalers.

The Memory Squeeze

The short-term story is uglier. The whole industry is running near full utilization, with data center demand for HBM starving handsets of DRAM, driving what management calls "unprecedented memory prices." Qualcomm's response is double-digit price increases passed through as contracts reset. “the scale of the increase that we're looking at is double digit and consistent with some of the actions from other players” — Akash Palkhiwala, Chief Financial Officer · 2026-07-29 But the near-term damage is real: QCT gross margins run slightly below the historical 48–50% band as input cost inflation hits across wafers, packaging, and test — and the data center mix will drag further once custom-chip revenue layers in.

the data center revenue coming in since the first revenue is mostly from custom chip engagements... that will be a drag of 1.5% to 2% on the weighted average gross margin for QCT.
Akash Palkhiwala, Chief Financial Officer · 2026-07-29

The handset revenues story mirrors the whole macro bind: China OEMs hit bottom this quarter after draining channel inventory, with double-digit sequential growth guided for Q4 even as the overall market shrinks in the low teens.

Trading the Apple Stream for a Bigger One

The Apple step-down is not new — the 20% share assumption had been firm for over a year. What changed is the acceleration. In April, the CFO was unmoved. “No change, Tim, to our assumption there. We've set... 20% share of the phones that will launch in fall this year and no product relationship beyond that.” — Akash Palkhiwala, CFO · 2026-04-29 Three months later, supply constraints turned that planning assumption into a self-fulfilling exit: “our supply constraints were a part of it and then where discussions ended up is that we'd have a share less than -- materially less than 20%” — Akash Palkhiwala, Chief Financial Officer · 2026-07-29 — crystallized in guidance:

as a result of our supply constraints, we now expect an acceleration in the step-down of Apple product revenues starting in the fourth fiscal quarter as our share for upcoming iPhone launch is expected to be materially lower than our prior estimate of 20%.
Akash Palkhiwala, Chief Financial Officer · 2026-07-29

Management frames it as an even trade: fiscal 2027 non-handset growth of >60% should "replace the entire Apple product revenue within the year," the BMW agreement lifts automotive's annualized exit run-rate to ~$7B, and the Apple product stream is being ripped out on a schedule the company now controls. The market is pricing the risk. The stock sits 36% below its May 29 peak of $251, unwinding a 67% five-week melt-up triggered by the June Investor Day, and the whole AI data center complex has been selling off. The de-rating shows in the fundamentals: Price-to-Revenue at 3.1x, while gross margins slipped to 54% and net cash flipped to a $749M net-debt position after a $2.8B buyback quarter.

Why It Matters

This is the rare call where a company deliberately accelerates the loss of its most profitable near-term stream to fund a bet on a market where it has yet to prove it can win. The early proof points — two hyperscaler POs, HBC tape-out, Modular closed — are encouraging but unproven. Whether the >$1.50 EPS hit from handset weakness plus the Apple exit is a bridge too far, or the cost of admission to agentic compute's next decade, is the open question. Qualcomm is betting the trade works, and the stock is pricing in considerable doubt.