The new platform will include CPUs, AI accelerators and ASICs as Qualcomm targets billions of dollars in data center revenue from fiscal 2027.
has officially launched Dragonfly, a new brand dedicated to its growing portfolio of data center products, as the company seeks a larger role in the rapidly expanding AI infrastructure market.
The announcement was made by Qualcomm CEO Cristiano Amon during his keynote at Computex 2026 in Taipei. While detailed specifications were not disclosed, Qualcomm said it is already working with major hyperscale cloud providers to deploy Dragonfly solutions in large-scale data center environments. A more comprehensive product roadmap is expected later this month.
Dragonfly will encompass three key product categories: data center CPUs, AI inference accelerators and custom application-specific integrated circuits (ASICs).
Qualcomm’s AI accelerator lineup includes the AI200 and AI250 chips, introduced last year. The AI200 features 768GB of LPDDR memory, while the AI250 uses a near-memory computing architecture designed to significantly boost memory bandwidth and improve energy efficiency. The company claims the accelerators can reduce power consumption by 35% to 70% compared with conventional GPU-based systems while delivering strong AI inference performance.
The AI200 is expected to launch this year, followed by the AI250 in 2027.
Qualcomm has already secured a supply agreement with AI company Humain and plans to support the development of a 200-megawatt-scale AI data center using its accelerator technology. The company now expects revenue from the project to begin in fiscal 2027, earlier than previously anticipated.
The chipmaker is also preparing to launch its first data center ASIC later this year. Reports indicate the initial customer may be ByteDance, with the collaboration expected to extend across multiple product generations.
Through Dragonfly, Qualcomm is broadening its computing portfolio beyond smartphones, PCs, wearables and automotive platforms, positioning itself as a competitor in the AI data center market. The company expects the business to generate billions of dollars in annual revenue beginning in fiscal 2027.
