Announcing Arm AGI CPU: The silicon foundation for the agentic AI cloud era - Arm Newsroom
British chip designer Arm has announced the Arm AGI CPU, a new production-ready silicon designed to power the next generation of AI infrastructure and coordinate agentic AI systems at global scale.
For the first time in its 35-year history, Arm is delivering its own silicon products, extending the Arm Neoverse platform beyond IP and Arm Compute Subsystems (CSS). This move reflects both the rapid evolution of AI infrastructure and growing demand for production-ready Arm platforms that can be deployed at scale.
In the era of agentic AI, systems operate continuously at global scale, with software agents coordinating tasks, interacting with multiple models and making decisions in real time. This places new demands on the CPU, which becomes the pacing element responsible for keeping distributed AI systems operating efficiently.
Rack-scale agentic efficiency
The Arm AGI CPU is designed to deliver high per-task performance at sustained load across thousands of cores within modern data center power and cooling limits. Every element, from operating frequency to memory and I/O architecture, supports massively parallel agentic workloads in dense rack deployments.
Arm's reference server configuration is a 1OU, 2-node design with two chips featuring 272 cores per blade. A standard air-cooled 36kW rack can house 30 blades delivering 8,160 cores total. Arm has also partnered with Supermicro on a liquid-cooled 200kW design capable of housing 336 Arm AGI CPUs for over 45,000 cores.
In this configuration, the Arm AGI CPU delivers more than 2x the performance per rack compared to the latest x86 systems, achieved through class-leading memory bandwidth, efficient single-threaded Arm Neoverse V3 CPU cores, and more usable threads that do more work.
Commercial momentum
The Arm AGI CPU is already seeing strong commercial momentum with partners at the forefront of scaling agentic AI infrastructure. Meta serves as the lead partner, co-developing the CPU to optimize gigawatt-scale infrastructure for its apps and work alongside Meta's custom MTIA accelerators.
Other launch partners include Cerebras, Cloudflare, F5, OpenAI, Positron, Rebellions, SAP, and SK Telecom. Commercial systems are now available for order from ASRockRack, Lenovo and Supermicro.
Delivering AI experiences at global scale demands a robust and adaptable portfolio of custom silicon solutions, purpose-built to accelerate AI workloads and optimize performance across Meta's platforms. We worked alongside Arm to develop the Arm AGI CPU to deploy an efficient compute platform that significantly improves our data center performance density and supports a multi-generation roadmap for our evolving AI systems.
New data center chapter
The launch represents a new chapter in Arm's data center journey and continues its leadership in computing innovation. The Arm AGI CPU is the first offering of Arm's new data center silicon product line and is available to order now, with follow-on products committed targeting best-in-class performance, scale and efficiency.
To accelerate adoption, Arm is introducing the Arm AGI CPU 1OU Dual Node Reference Server, an Open Compute Project (OCP) DC-MHS standard form factor server. Arm plans to contribute this reference server design and supporting firmware to the OCP community.
More than 50 leading companies across hyperscale, cloud, silicon, memory, networking, software, system design and manufacturing are supporting the expansion of the Arm compute platform into silicon. With the AGI CPU, Arm is both defining the architecture of the AI-native data center and building it.
Related
