For two decades, Qualcomm has been the company that powers everyone's phone. Wednesday, it announced plans to become something else entirely: a credible challenger in AI data center chips, with Meta as its first customer. The numbers tell the story — $15 billion in projected data center revenue by fiscal 2029, part of a $40 billion non-handset target that would fundamentally reshape what Qualcomm actually is.
Key Takeaways
- Qualcomm raised its fiscal 2029 non-handset revenue forecast from $22 billion to $40 billion, with $15 billion from data centers
- Meta signed on as the first major customer for Qualcomm's new AI data center CPU
- Shares jumped 15% in extended trading as investors priced in reduced smartphone dependence
What Qualcomm Actually Announced
Qualcomm disclosed the Qualcomm AI data center CPU initiative and the Qualcomm-Meta partnership during its shareholder meeting Wednesday. The company raised its fiscal 2029 non-handset revenue target from $22 billion to $40 billion — an $18 billion increase — with data centers accounting for $15 billion of that figure. That's not a modest revision. That's a new business line the size of a Fortune 500 company.
The adjusted earnings target moved to over $18 per share, roughly 18% above the $15.26 analysts had forecast. Smartphones still generate two-thirds of Qualcomm's product revenue today, according to CNBC. The data center CPU represents the company's most direct attempt to change that ratio. Investors responded immediately: shares rose 15% in after-hours trading.
What Meta's Involvement Means
Here's where most coverage stops, and where the interesting question begins. Meta isn't just any customer — it's a hyperscaler with exacting technical requirements and massive AI infrastructure needs. The company doesn't experiment with unproven chip suppliers for strategic workloads unless the architecture solves a real problem.
Qualcomm has not disclosed pricing, chip specifications, deployment timelines, or contract terms. What the announcement does confirm is enterprise willingness to consider ARM-based alternatives in a market historically locked down by Intel and AMD's x86 duopoly. Meta's participation signals that openness isn't hypothetical — it's converting into actual design wins.
The 15% share price jump reflects investor interpretation of Meta as validation. If one of the world's largest AI operators is betting on Qualcomm silicon, the thinking goes, others might follow. That's the theory. Whether it holds depends entirely on what Qualcomm hasn't yet disclosed: performance benchmarks, cost structure, and the customer pipeline beyond Meta.
The Bigger Strategic Bet
What most coverage misses is the scale of the shift Qualcomm just committed to publicly. A $15 billion data center business by fiscal 2029 would represent roughly one-third of the company's projected $40 billion non-handset revenue. For context, that's roughly the size of Qualcomm's entire non-smartphone business today — $22 billion in the prior forecast — added again, mostly from a single new market.
Qualcomm faces structural risk in smartphones: slowing global sales, intensifying competition in mobile chipsets, and customer concentration in a maturing market. The revised targets reveal how urgent diversification has become. This isn't a side project. It's a rebalancing of the company's core exposure, with data centers carrying the largest single burden in the plan.
The challenge: Intel, AMD, Nvidia, and cloud providers with custom silicon programs already occupy this space. Qualcomm is betting it found an opening — likely in power efficiency, ARM architecture advantages, or AI-specific workload optimization — that incumbents haven't filled. The company has not yet explained what that opening is.
What Remains Unconfirmed
The available reports do not specify the technical architecture of Qualcomm's AI data center CPU. Does it compete directly with Nvidia's AI accelerators, Intel's Xeon processors, or AMD's EPYC chips? The positioning matters, and the company hasn't clarified it.
Qualcomm has not detailed the path from current smartphone-dominated revenue to $15 billion in data center sales in five years. No additional customers beyond Meta have been named. No pricing models announced. No clarification on whether the processors will be sold to hyperscalers directly, offered through cloud marketplaces, or licensed as intellectual property.
The $40 billion non-handset forecast includes existing automotive, IoT, and PC chip businesses. The source material does not break down how much of the $18 billion increase comes from data centers versus those other segments. That split determines whether this is a data center story or a broader diversification story that happens to include data centers.
What Data Center Buyers Should Watch
Qualcomm's next quarterly earnings call will be the first chance for management to detail the CPU roadmap, customer pipeline, and competitive positioning. Data center operators evaluating server chip strategies should watch whether the company names additional hyperscale customers. A $15 billion revenue target built on a single partnership is a different risk profile than one with multiple design wins across cloud providers and enterprises.
The fiscal 2029 targets — particularly that over $18 per share adjusted earnings projection — create measurable accountability. Meta's infrastructure disclosures over the next year, including AI training capacity plans and chip sourcing strategies, may reveal the actual scale of the Qualcomm partnership. Those updates will clarify whether this is a major platform shift or a limited trial deployment.
Competitive responses matter. Intel, AMD, and ARM-based rivals like Ampere and Amazon's Graviton team now know Qualcomm considers their market contestable. Watch for benchmark comparisons, pricing pressure, or new product launches aimed specifically at the workloads Qualcomm is targeting. The market Qualcomm just entered doesn't concede share quietly.
