At its recent Investor Day in New York, Qualcomm presented a detailed growth plan aimed at expanding beyond smartphone chips into the broader AI infrastructure market for data centers and large-scale model providers.
Financial targets and market size
Qualcomm announced an objective to more than double its non-mobile revenue to $40 billion within three years. The company estimates the total addressable market could reach as much as $1.7 trillion by 2030, reflecting an anticipated structural shift toward AI-driven revenue streams across multiple large markets.
What Qualcomm plans to build
Qualcomm said it intends to cover the full AI stack: accelerators, networking solutions, memory architectures and the necessary software ecosystem. Key elements of the plan include:
- a CPU scaled for AI workloads (the Dragonfly C1000),
- leveraging AlphaWave IP-based connectivity solutions,
- establishing a custom ASIC design business,
- a new High Bandwidth Compute (HBC) architecture intended to speed data movement between memory and compute units, reducing a major bottleneck in AI systems while improving energy efficiency.
Production of the Dragonfly C1000 server processor is expected to begin in the second half of 2028.
Partners and software moves: Meta, Microsoft Azure and Modular
Qualcomm already has notable partners for its data-center ambitions. The first major partner is Meta, which plans to use Qualcomm’s Dragonfly C1000 server CPU in its next-generation AI systems. Microsoft Azure has also agreed to adopt Qualcomm’s HBC architecture.
Recognizing that AI competition increasingly depends on software as well as hardware, Qualcomm will acquire the AI-software startup Modular for roughly $4 billion. Modular develops developer tools to enable AI models to run across chips from different vendors, a capability that directly challenges established platforms such as Nvidia’s CUDA.
Investor reaction and risks
Investors welcomed the announcements and Qualcomm’s shares rose on the news. However, the company must translate its ambitious plans into measurable results. Diversifying away from a historically dominant single revenue source is strategically sensible but carries execution and market risks.
Why this matters
If Qualcomm successfully builds a full-stack AI infrastructure offering, it could materially change the company’s revenue mix and market position over the next decade. Moving beyond a primary focus on mobile devices toward data-center and hyperscaler customers could make Qualcomm a major player in both AI hardware and the supporting software ecosystem.
(Key figures: target $40 billion non-mobile revenue within three years; TAM up to $1.7 trillion by 2030; Dragonfly C1000 production expected H2 2028; Modular acquisition ~ $4 billion.)
