Qualcomm has introduced its first-ever CPU designed for Data Centers, the Dragonfly C1000, which leverages the Oryon architecture.
Qualcomm Enters The Agentic AI CPU Race With Dragonfly C1000 Chip, Oryon-Based With Over 5 GHz Clocks, Over 250 Cores, & Aims To Achieve Single-Core Leadership
One of the biggest announcements by Qualcomm today was its first release of a CPU for the data center segment, called the Dragonfly C1000. This is a chip purpose-built for Agentic AI & General-Purpose workloads, delivering best-in-class power efficiency and TCO.
As per Qualcomm, the Dragonfly C1000 is based on a custom-designed Oryon core architecture that is optimized for core performance and delivers frequencies beyond 5 GHz, offering superior performance in agentic workloads deployed at scale. Qualcomm also claims that the chip is going to offer leadership single-threaded performance.
In terms of specifications, the Qualcomm Dragonfly C1000 CPU will offer a multi-chiplet design, allowing it to align with leading-edge advanced packaging technologies for performance and IO scaling. The primary CPU chiplet will feature over 250 cores, offering throughput and scale while delivering the aforementioned "exceptional per-core" performance.
Performance-wise, the chips are said to be over two times better in terms of perf/watt when compared to existing server CPUs. The company doesn't disclose any direct comparisons or benchmarks yet. Each CPU will offer over 2 TB/s of PCIe Gen7 connectivity, CXL connectivity & support for next-generation accelerators such as Qualcomm's very own AI series products.
CPU portfolio includes: agentic CPU designed for high-throughput agentic orchestration and low latency interactive AI use cases; general-purpose CPU designed for optimal performance-per-TCO for first-party workload and performance-per-vCPU for third-party usage elasticity; AI head node CPU designed to maximize XPU utilization of XPU for generative AI compute through low overhead host processing through high-speed CPU
Just like its AI accelerator portfolio, the Dragonfly C1000 CPUs will come with an optional HBC attach to boost their memory capacity and bandwidth capabilities. From a security standpoint, the C1000 chips will integrate advanced reliability, availability, and serviceability (RAS) features, including ECC memory correction, fault isolation, and error recovery.
The first platforms based on these chips will feature both air and liquid cooling support in OCP ORv2 compliant racks and servers. Qualcomm has set the commercial availability of C1000 for 2028.
The Battle of The Data Center CPU Throne
Now, with this announcement, Qualcomm is entering a very contested space. The data center CPU segment has seen several new ARM additions while x86 opponents continue to build upon their successes.
Both AMD & Intel have new data center chips coming out through 2028, with the Intel camp readying its Diamond Rapids and Coral Rapids Xeon family. AMD is setting up its Zen 6 Venice and the Inference-optimized Verano CPUs for the Agentic AI segments.
NVIDIA is already pushing its Vera CPUs as the crown holder for 2026, and future ARM-based CPUs are already in plans with Rosa set up for 2029, and ARM itself has its AGI CPUs backed by a multi-year roadmap. CPUs are the new boss of the AI segment, and Qualcomm's entry makes the Agentic AI race even more interesting.
So those gains that Qualcomm is claiming today might become less spectacular by the time the chip launches and faces the true competition that is mentioned above. And its 250+ core count figure might sound grand now, but AMD is already mass producing its 256-core Venice chips at the moment, which will be rolling out soon. There's a lot to prove for the Dragonfly time.
But First Agreements Already Underway
Despite all of that, Qualcomm has managed to sign a multi-generation agreement with META on its data center CPUs under the Dragonfly series.
"We're excited to continue partnering with Qualcomm Technologies as they design the next generation of CPUs for Meta," said Mark Zuckerberg, Founder and CEO, Meta. "Along with our other compute investments, we're quickly building the infrastructure we need to deliver personal superintelligence to everyone in the world.”
Qualcomm is all set to enter the $200B CPU server TAM by the time Dragonfly C1000 launches, bringing a multi-generational portfolio that is expected to win major hyperscaler adoption.
About the author: A Software Engineer by training and a PC enthusiast by passion, Hassan Mujtaba serves as Wccftech's Senior Editor for hardware section. With years of experience in the industry, he specializes in deep-dive technical analysis of next-generation CPU and GPU architectures, motherboards, and cooling solutions. His work involves not only breaking news on upcoming technologies but also extensive hands-on reviews and benchmarking.
Follow Wccftech on Google to get more of our news coverage in your feeds.
