Qualcomm has unveiled its new AI200 and AI250 accelerators, marking its entry into the AI data center chip market and positioning itself as a direct competitor to Nvidia and AMD, CNBC[1] reported Tuesday.
The move represents a shift for Qualcomm, which has traditionally focused on mobile and wireless chips rather than large-scale computing hardware.
The company said the AI200, set to launch in 2026, and the AI250, planned for 2027, will be available in full liquid-cooled server rack systems. These setups are designed to host up to 72 chips working together, similar to Nvidia and AMD’s high-end GPU racks.
Both accelerators are based on Qualcomm’s Hexagon neural processing units (NPUs), originally developed for smartphones. The company says leveraging that technology for larger systems was a natural progression.
Durga Malladi, general manager for data center and edge, said:
We first wanted to prove ourselves in other domains, and once we built our strength over there, it was pretty easy for us to go up a notch into the data center level.
Qualcomm claims its chips will offer high memory capacity while consuming less power, aiming to attract AI labs seeking efficiency for large model training and deployment.