Qualcomm: Data Center Revenue To Reach $15 Billion In 3 Years

Qualcomm is expected to gain data-center share through FY2029 via connectivity products, AI accelerators, custom ASICs and Arm-based server CPUs, supported by hyperscaler wins. Acquisitions including Nuvia, Alphawave and Modular, alongside partnerships with Hugging Face and Confidential Core AI, are positioned to strengthen its technology and software ecosystem. Despite intense Nvidia competition, the outlook supports management's ambitious $15B FY2029 data-center revenue target.

Analysis

The investable issue is not whether QCOM can assemble credible IP, but whether it can convert design activity into recurring hyperscaler revenue before its mobile earnings cycle re-rates lower. A $15B data-center business would require a material departure from QCOM’s historically handset-centric customer concentration and a sustained gross-margin profile against NVDA, AVGO, AMD and MRVL. Custom silicon can create large revenue pools, but it generally carries lower margins, lumpy deployment schedules and greater customer bargaining power than merchant accelerator sales; the market should not award NVDA-like multiples to prospective ASIC revenue without disclosed backlog, wafer commitments and named production deployments.

Near term (days to 3 months), QCOM is likely sensitive to incremental evidence that its AI/server programs have moved beyond evaluations: hyperscaler capex commentary, disclosed design wins, HBM supply access, and foundry capacity commitments matter more than roadmap language. The second-order beneficiary of credible QCOM traction is ARM, since higher-volume Arm server adoption expands royalty and licensing optionality; the more direct losers would be smaller merchant networking/ASIC vendors such as MRVL where custom-program competition can pressure design-win economics. NVDA’s risk is limited initially: its moat is the integrated software, networking and installed-base stack, not simply compute architecture.

Consensus may be underestimating the time and capital required to establish a data-center software ecosystem, particularly around compiler maturity, framework support, cluster management and customer migration costs. The bull case is underpriced only if QCOM demonstrates at least one scaled deployment with repeatable economics; absent that, the target functions primarily as a long-dated valuation narrative while handset inventory, China demand and Android premium-tier share remain the nearer earnings drivers. Falsify a constructive view if the next two earnings cycles lack quantified data-center bookings/backlog, if segment-margin guidance implies dilution, or if hyperscalers continue concentrating incremental accelerator spend with NVDA/AVGO.

Key Decisions for Investors

• Maintain QCOM as a watch-list long rather than add on roadmap commentary; initiate only after a named hyperscaler production win or quantified data-center backlog, targeting a 6-12 month holding period. Size for a 10-15% downside if the next earnings call provides no commercialization metric or mobile guidance weakens.
• Use a 6-12 month pair trade only after verification: long QCOM / short MRVL, where a confirmed QCOM custom-ASIC win would pressure MRVL’s design-win scarcity premium. Exit if MRVL discloses offsetting hyperscaler programs or QCOM’s disclosed program economics indicate materially lower-than-expected gross margin.
• For diversified AI exposure, retain NVDA over QCOM until QCOM shows deployed cluster-scale software performance; NVDA’s nearer-term earnings risk is valuation and capex digestion, while QCOM’s is execution. A QCOM catch-up trade requires evidence, not a stated FY2029 objective.
• Monitor ARM as a second-order beneficiary of credible Arm server-volume adoption. Add only if QCOM or another hyperscaler program discloses production volumes, since architecture enthusiasm without shipments is unlikely to change ARM’s royalty trajectory within the next 1-3 quarters.