Qualcomm Technologies (NASDAQ: QCOM) has entered a multi-generation agreement to supply data centre CPUs to Meta.

Under the collaboration, Qualcomm’s Dragonfly C1000 data centre CPU is expected to support Meta’s next-generation server fleet as the company expands large-scale compute infrastructure.

The processors are scheduled to enter production in the second half of 2028 and will support future data centre capacity expansions.

Qualcomm said its platform combines compute, connectivity and system-level optimisation, with a focus on improving performance per watt and reducing operating costs at scale. 

“We designed our data centre CPU to deliver leading performance per core and a breakthrough in power efficiency for large-scale data centre deployments, and this multi-generation agreement with Meta is a significant validation of that approach,” said Cristiano Amon, president and CEO, Qualcomm Incorporated.

Meta founder and CEO Mark Zuckerberg said the company is continuing to work with Qualcomm Technologies as it designs next-generation CPUs for Meta. “Along with our other compute investments, we’re quickly building the infrastructure we need to deliver personal superintelligence to everyone in the world.”

New products

The Meta agreement follows Qualcomm’s launch last week of a broader data centre roadmap, including the Dragonfly C1000 CPU, High Bandwidth Compute, the Dragonfly AI300 inference accelerator, connectivity products, and custom silicon solutions.

The C1000 is Qualcomm’s planned data centre CPU for agentic, general-purpose, and AI head node workloads. The chip will use a 250-plus core chiplet design built on custom Oryon CPU cores, with commercial availability expected in 2028.

Qualcomm is also pushing High Bandwidth Compute as part of its AI inference strategy. The technology is designed to bring compute closer to memory and reduce data movement bottlenecks, a key issue for large language models and agentic AI workloads.

The Dragonfly AI300 joins the previously announced Dragonfly AI200 and AI250 accelerators. Qualcomm said AI250 sampling is expected in mid-2027, while AI300 sampling is expected in 2028.

Qualcomm said its data centre roadmap will follow an annual cadence, with a focus on AI inference performance, energy efficiency, and total cost of ownership.