Qualcomm CEO: AI Firms Want Local 100B Parameter Mobile Platforms by 2028
Qualcomm CEO Cristiano Amon revealed that AI companies are requesting mobile platforms capable of running 100B parameter models locally by 2028. Additionally, Amon highlighted AI agents as the catalyst for the next smartphone upgrade super cycle. Running 100B parameter models locally on mobile devices would drastically enhance offline AI capabilities, privacy, and multi-agent coordination without relying on cloud infrastructure. This forward-looking hardware requirement signals a major shift in how edge computing and mobile silicon will evolve over the next few years. Beyond mobile platforms, Amon views smart glasses as his top choice for wearable AI devices and anticipates early 6G deployments around 2029 to 2030. He also noted that Qualcomm's expertise in low-power technology serves as a strong foundation for entering the data center market.
## BACKGROUND
Edge AI refers to running artificial intelligence models directly on local hardware devices like smartphones and PCs rather than remote cloud servers, offering lower latency and better privacy. Parameter scale, measured in billions (e.g., 8B, 100B), generally correlates with a model's complexity and capability, with 100B-scale models traditionally requiring immense computational power and high-end server hardware.