Qualcomm Dragonfly C1000: Why Meta’s First Data Center CPU Deal Matters

The Meta Deal: A Multi-Generation Commitment

Meta signed a strategic, multi-generation supply agreement to use the Dragonfly C1000 and succeeding Qualcomm CPU generations in its data center server fleet . The agreement expands Qualcomm's existing relationship with Meta, which already covers connectivity and other chip products .

Qualcomm also disclosed that it has secured "significant orders" from two hyperscale cloud service providers, with Meta as the first named deployment customer .

Revised Fiscal 2029 Financial Targets

Qualcomm set dramatically higher long-term revenue goals at the Investor Day:

• Non-handset QCT revenue target: Raised to $40 billion by fiscal 2029, nearly doubling the prior target of ~$22 billion .
• Data center revenue target: More than $15 billion in annual data center AI component sales by fiscal 2029, up from approximately $300 million in fiscal 2026 .
• Automotive target: Also aiming for ~$10 billion in automotive revenue by FY29 .
• Scaling path: Data center revenue is expected to ramp sequentially from FY26 through FY28, driven first by connectivity products, then custom silicon and CPUs .

Stock Surge and Analyst Reaction

Qualcomm shares surged ~15% in after-hours trading on June 24 and rose approximately 12% in premarket trading on June 25 following the announcements . Analysts at Bank of America and UBS raised their price targets, though some expressed differing views on execution risk .

Qualcomm's Return to the Data Center After Previous Setbacks

Qualcomm had attempted to enter the server CPU market roughly a decade ago with its Arm-based Centriq line, but shut down the effort in 2018 after failing to gain traction against Intel's Xeon. The 2026 Dragonfly roadmap marks a fully renewed push into the data center, timed to the AI infrastructure boom and built around custom Oryon cores that Qualcomm developed from the Nuvia acquisition.

The company backed this re-entry with the $3.9 billion acquisition of Modular Inc., announced at the same Investor Day, to strengthen its AI infrastructure software stack .

Strategy: Arm-Based CPUs Inside Nvidia's Ecosystem

Qualcomm's data center CPUs are designed to integrate with Nvidia's NVLink Fusion interconnect fabric, allowing the Arm-based Oryon CPUs to connect seamlessly with Nvidia's rack-scale GPU systems . This was first telegraphed in May 2025, when Qualcomm announced it would support Nvidia's NVLink Fusion, and Arm subsequently extended the Neoverse platform with NVLink support in December 2025 .

The strategy positions Qualcomm within Nvidia's ecosystem rather than as a direct competitor to Nvidia GPUs — Qualcomm CPUs act as head nodes and general-purpose cores in GPU-accelerated AI clusters. At the same time, Qualcomm is also reducing dependency on Arm and Nvidia through acquisitions like Alphawave and Ventana Micro, giving it more in-house silicon and interconnect capabilities for the long term .

Competitive Landscape: Intel, AMD, and Nvidia's Own CPUs

Qualcomm is directly challenging Intel's Xeon and AMD's Epyc lines in the general-purpose and AI server CPU market . Its claims of 2x better performance per watt and 30% more performance versus incumbents are aggressive but unverified until silicon ships in 2028 .

Nvidia itself is also developing its own Vera server CPUs, meaning Qualcomm could face competition from both Intel/AMD on one axis and Nvidia on another, even as it partners with Nvidia on interconnect . The market entry is also complicated by the emergence of custom silicon at hyperscalers (Amazon's Graviton, Google's Axion, Microsoft's Cobalt), all Arm-based, which compete for the same workload slots .

Broader Implications

Translating mobile expertise: Qualcomm's core competency — designing ultra-efficient, high-performance Arm-based SoCs at massive scale for smartphones — is directly relevant to data center CPUs. The Oryon cores powering the Dragonfly C1000 are an evolution of the same cores used in Qualcomm's Snapdragon X laptop and mobile processors .

AI infrastructure credibility: Naming Meta as a marquee customer gives Qualcomm immediate credibility in a market where trust and long qualification cycles are major barriers. Meta's massive compute footprint (Facebook, Instagram, AI training/inference) provides a real-world proving ground.

Diversification away from handsets: The $40 billion non-handset target — nearly doubling prior guidance — signals Qualcomm's ambition to transform from a mobile-chip company into a broad computing infrastructure player, with data center, automotive, PC, and IoT as co-equal pillars .

Risk: The Dragonfly C1000 won't reach production until the second half of 2028, which is several years away in an AI market that evolves extremely fast. By then, Intel, AMD, Nvidia, and hyperscaler custom chips will have advanced further, and Qualcomm must execute flawlessly on both silicon design and ecosystem adoption to hit its ambitious $15 billion data center revenue goal .