Qualcomm shares rose more than 20% on Monday after the company announced new AI chips and server systems for data centers. The move puts Qualcomm into direct competition with Nvidia and AMD in the growing AI accelerator market.
New Chip Lineup and Release Schedule
According to Yahoo Finance, Qualcomm unveiled the AI200 and AI250 chips along with rack-scale server offerings. The AI200 will launch in 2026, while the AI250 follows in 2027. A third chip and server are planned for 2028.
The AI200 refers to both an individual AI accelerator and a full server rack with a Qualcomm CPU. The AI250 will offer 10 times the memory bandwidth of the AI200. Qualcomm said it will follow this annual release cadence going forward.
Both chips use Qualcomm’s custom Hexagon NPU, or neural processing unit. The company scaled up technology from its Windows PC chips for data center use. The chips are designed for AI inference, the process of running AI models, not training new ones.
Cost and Competition Strategy
Qualcomm highlighted low power consumption and total cost of ownership as key advantages. Data center builders face high costs for constructing and operating server farms.
Customers can buy individual chips, portions of server offerings, or complete setups. Durga Malladi, senior vice president at Qualcomm, said potential customers could include Nvidia and AMD, making them both rivals and partners.
Market Context and Revenue Shift
This is not Qualcomm’s first data center attempt. In 2017, the company launched the Centriq 2400 platform with Microsoft but ended the project due to competition from Intel and AMD, plus corporate legal issues.
Qualcomm currently offers the AI 100 Ultra card for standard servers. The new AI200 and AI250 are built for dedicated AI systems.
The push helps Qualcomm reduce reliance on smartphone chips and licensing. In Q3, Qualcomm reported $10.4 billion in total revenue, with $6.3 billion from handsets. The company does not report data center revenue separately.
Qualcomm faces tough competition from cloud companies like Amazon, Google, and Microsoft, which develop their own AI chips. Nvidia and AMD already dominate the AI accelerator space.
