Qualcomm Technologies announced a new data center roadmap and product portfolio designed for the agentic AI era. The company introduced new data center solutions, including the Qualcomm Dragonfly C1000 CPU, Qualcomm High Bandwidth Compute technology, the Qualcomm Dragonfly AI300 inference accelerator, connectivity products, and custom silicon solutions.
Qualcomm said the new platforms are engineered to maximize performance per watt and token throughput while reducing total cost of ownership.
The announcement expands Qualcomm’s role in full-stack data center infrastructure optimized for AI. The portfolio spans data-center-class CPUs, AI inference accelerators, high-performance connectivity, and custom silicon.
The Qualcomm Dragonfly AI300 joins the previously announced Qualcomm Dragonfly AI200 and AI250 in the company’s multi-generation AI accelerator roadmap, which is expected to move on an annual cadence.
Qualcomm said agentic AI is increasing demand for inference capacity in the data center, making tokens-per-watt an increasingly important measure for reducing infrastructure costs.
The Dragonfly C1000 CPU is a purpose-built data center CPU designed for agentic, general-purpose, and AI head node workloads. It uses custom-designed Qualcomm Oryon CPU cores optimized for core performance and frequencies above 5 GHz.
The C1000 features a chiplet design with more than 250 cores and is designed to deliver more than two times better performance per watt compared with existing server CPU competitive offerings, based on Qualcomm’s estimates. Commercial availability is expected in 2028.
Qualcomm High Bandwidth Compute is a near-memory computing architecture that bonds compute with accelerated memory bandwidth in a 3D-stacked silicon solution. Qualcomm said HBC is designed to address AI’s data movement bottleneck and improve processing efficiency at lower total cost of ownership.
With HBC Gen 1, Qualcomm said the AI250 is designed to enable 133 TB/s per card, an 18x increase in effective memory bandwidth compared to AI200 with LPDDR5X. AI300 with HBC Gen 2 is designed to enable a 54x increase over AI200. Commercial sampling of HBC Gen 1 with AI250 is expected in mid-2027.
The Dragonfly AI300 is a third-generation rack-level AI inference platform that supports air and direct-liquid cooling. It integrates Qualcomm HBC Gen 2 technology for compute acceleration, integrated memory, and increased effective memory bandwidth.
Qualcomm said AI300 is designed for high-throughput, low-latency performance for large language model, large multimodal model, and agentic AI inference workloads. Commercial sampling is expected in 2028.
The roadmap also includes custom silicon for next-generation AI and cloud data center infrastructure, with end-to-end co-design across silicon, systems, and software.
Qualcomm’s connectivity portfolio spans die-to-die, copper, optical, and campus-reach interconnects. The company said the portfolio supports high-bandwidth 800G and 1.6T connectivity across optical, AOC, and AEC applications, including campus-reach deployments up to 20 kilometers.
Alongside the Dragonfly portfolio, Qualcomm announced a multi-year, multi-generation agreement with Meta. Qualcomm’s Dragonfly C1000 data center CPU is planned to power Meta’s next-generation server fleet.
Qualcomm also said more than 35 global technology and AI ecosystem leaders are supporting its data center vision and commercial solutions.
KEY QUOTES:
“Agentic AI is driving a significant increase in demand for AI inference in the data center. As these become the dominant workloads, infrastructure has to deliver much higher performance at lower power and cost. That plays directly to Qualcomm’s strengths, and we’re well positioned for this shift. With Qualcomm Dragonfly, we’re bringing our high-performance, low-power computing into the data center, with multi-year, multi-generation agreements with leading customers.”
Cristiano Amon, President and CEO of Qualcomm
“What enterprises need now goes far beyond individual components. Orchestrating multiple types of compute across distributed, always-on infrastructure is critical. With Qualcomm Dragonfly, we’re bringing together compute, AI, memory, and connectivity into a unified, rack-scale platform designed for increasingly complex, agent-driven workloads while addressing key bottlenecks in memory bandwidth and power consumption. This builds on what Qualcomm Technologies has been delivering for decades: high-performance, low-power compute at scale, now applied to the data center in a way that very few companies can match.”
Tony Pialis, EVP and GM, Data Center, Qualcomm Technologies
