Qualcomm Technologies, Inc. (NASDAQ: QCOM) announced a new line of data center products at its Investor Day, including the Dragonfly C1000 CPU, the Dragonfly AI300 inference accelerator, and a new High Bandwidth Compute (HBC) memory architecture, along with connectivity products and custom silicon solutions.

The company also announced a multi-year, multi-generation agreement with Meta, under which the Dragonfly C1000 CPU is planned to power Meta's next-generation server fleet. Commercial availability of the C1000 is expected in 2028.

The Dragonfly AI300, the third generation in Qualcomm's AI accelerator line following the AI200 and AI250, is expected to begin commercial sampling in 2028. The AI300 uses HBC Gen 2 technology, which Qualcomm claims delivers a 54x increase in effective memory bandwidth compared to the AI200. Commercial sampling of HBC Gen 1, paired with the AI250, is expected in mid-2027.

The C1000 CPU features a chiplet design with more than 250 cores and Qualcomm Oryon CPU cores running at frequencies above 5 GHz. Qualcomm claims the chip delivers more than 2x better performance per watt compared to existing server CPU offerings based on published specifications.

Qualcomm said more than 35 companies have expressed support for its data center portfolio, including Lenovo, Micron Technology, Samsung SDS, SK Hynix America, Supermicro, and VAST Data, among others.

Cristiano Amon, President and CEO of Qualcomm Incorporated, said in a statement that the company is "well positioned" for the shift toward AI inference workloads in the data center, citing multi-year agreements with leading customers.

The announcements were made according to a press release issued by Qualcomm Technologies.