Qualcomm Says AI Data Centers Will Become Its Next Big Growth Engine

Qualcomm targets more than $15 billion in AI data-center revenue by fiscal 2029 and raised its non-handset revenue goal to $40 billion, nearly doubling the prior target. The company also disclosed named hyperscaler partnerships with Meta and Microsoft, plus a planned $3.9 billion acquisition of Modular to strengthen its AI software stack. The update broadens Qualcomm's growth story beyond smartphones into AI infrastructure, edge AI, and automotive, though execution risk remains in a crowded data-center market.

Analysis

The market is likely underpricing how much of this is a software-and-ecosystem transition, not just a chip-design story. If Qualcomm can couple custom silicon with a portable software stack, its real optionality is in lowering customer switching costs at the exact moment hyperscalers are trying to diversify away from a single-vendor accelerator model. That makes the strategic threat to incumbents less about raw performance today and more about where the next procurement dollars are allocated over the 24-36 month capex planning cycle.

The second-order read-through is mixed for the competitive set. Nvidia is the obvious headline loser on sentiment, but the more immediate pressure point is custom silicon and networking vendors that depend on hyperscaler buildouts remaining concentrated in a few architectures; if buyers actively seek power-efficiency alternatives, the mix shifts toward CPUs, interconnect, and system-level optimization rather than pure accelerator spend. Broadcom and Marvell are therefore exposed to a potential reallocation of wallet share if Qualcomm’s pitch gains credibility, while Alphabet/Google benefit indirectly if Qualcomm’s edge and on-device work tightens Android/AI integration and reinforces non-Apple device differentiation.

The bull case is not a straight-line revenue ramp; the key catalyst is validation, not announcements. Meta and Microsoft are the sort of customers that can de-risk the narrative, but meaningful upside only arrives if Qualcomm proves it can move from pilot to repeatable deployment by 2H28-2029, which is a long time in AI infrastructure. The main tail risk is that the data-center opportunity becomes a “strategy headline” without enough design wins or margin durability, at which point the stock rerates back to handset-cycle multiples rather than AI-platform multiples.

Key Decisions for Investors

• Go long QCOM on pullbacks over the next 1-2 weeks, but size as a catalyst trade rather than a full secular allocation; target a 10-15% upside rerating if management converts the AI narrative into multiple incremental design-win updates by the next two quarters.
• Pair trade: long QCOM / short NVDA for 3-6 months as a relative-value bet on power-efficiency and customer diversification; risk is that Nvidia re-accelerates on product cadence, so keep the short leg smaller than the long leg.
• Pair trade: long QCOM / short AVGO over 6-12 months if hyperscalers start favoring more CPU/interconnect-heavy architectures; this works best if Qualcomm’s software stack begins to appear in customer roadmaps and custom-silicon budgets shift away from the current winners.
• Buy QCOM call spreads with 6-9 month tenor to capture rerating from execution milestones while limiting downside if the data-center roadmap remains aspirational; ideal strike structure is 10-20% out of the money.
• Avoid chasing META or MSFT on this headline; the incremental impact is more about optionality preservation than immediate earnings uplift, so the better trade is the supplier with a call option on new wallet share.