AI

Qualcomm targets $15B data center chip revenue with Meta deal

The mobile chipmaker unveiled its Dragonfly portfolio and doubled non-handset revenue guidance to $40 billion by fiscal 2029.

Omega Editorial· June 25, 2026· 3 min read

Qualcomm announced a major push into artificial intelligence data centers on Wednesday, unveiling its Dragonfly chip portfolio and securing Meta Platforms as a committed customer for processors expected to ship in 2028.

The mobile chipmaker disclosed the strategy at its 2026 Investor Day, where it doubled its fiscal 2029 non-handset revenue target to $40 billion and set a specific data center revenue goal exceeding $15 billion within that timeframe. Qualcomm shares jumped 11% in pre-market trading Thursday following the announcement.

New chip designs land major customers

The Dragonfly portfolio includes two distinct architectures aimed at different data center workloads. The C1000 central processing units will become available in 2028, when Meta has committed to deploying them in its infrastructure, according to details first reported by Barron's. A separate design called High Bandwidth Compute is scheduled for mid-2027 and has already been selected by Microsoft's Azure cloud division.

Qualcomm also announced it would acquire AI software startup Modular for $3.92 billion in an all-stock transaction. Modular develops software that enables AI models to operate across different hardware architectures—CPUs, GPUs, neural processing units, and custom chips—without requiring developers to rewrite code for each processor type. The acquisition aims to strengthen Qualcomm's software capabilities for data center AI inference and deployment.

Why it matters

Qualcomm's data center ambitions represent a fundamental business model shift for a company that generated 72% of its chip revenue from handsets in fiscal 2025. By fiscal 2029, the company projects smartphones will account for roughly one-third of chip revenues as it expands into higher-margin enterprise markets. The move positions Qualcomm to compete in the rapidly growing AI infrastructure market currently dominated by Nvidia, though the company is entering with established hyperscale customers already committed to future products. Success in data centers could reduce Qualcomm's vulnerability to smartphone market cycles and Apple's ongoing efforts to design its own mobile processors.

Diversification beyond mobile

President and CEO Cristiano Amon framed the announcements as "defining Qualcomm's next chapter" through accelerated diversification and evolution into a platform company. Beyond data centers, Qualcomm set fiscal 2029 targets of $10 billion in automotive revenues and more than $14 billion in Internet of Things revenues. The company also targets more than $18 in non-GAAP earnings per share by that year.

Qualcomm identified a combined addressable market of approximately $1.7 trillion across its target segments by 2030, reflecting opportunities in automotive computing, industrial IoT, and cloud infrastructure that extend well beyond the smartphone processors that built the company's current market position.

These details were first reported by Quartz.

#qualcomm#data center chips#ai infrastructure#meta platforms#semiconductor industry#enterprise computing

This is an original analysis by the Omega editorial team. Source reporting: AI Watch.

Want systems like this working for your business?

Book a Call

More in AI

AI· 3 min read

Cheap Chinese AI Models May Boost Industry Demand, Analysts Say

Price competition from open-weight models is driving down costs but could massively expand AI adoption across enterprises.

Via AI Watch · Aug 9, 2026
AI· 3 min read

Runware Packs 1MW AI Data Center Into Shipping Container

Startup's modular Pods deploy in a day with liquid cooling and no water consumption, targeting 1GW by 2027 as grid delays stall traditional builds.

Via AI Watch · Aug 8, 2026
AI· 3 min read

Generative AI in Biology Risks Creating Nonexistent Discoveries

When synthetic data replaces experimental measurements, fabricated molecular patterns could be mistaken for genuine biological effects.

Via AI Watch · Aug 8, 2026