Scaling AI Infrastructure with Open Systems and Arm-Based Silicon

Why It Matters

With agent-based AI poised to multiply inference requests and stress existing server architectures, ARM’s energy-efficient, standardized CPU and reference designs aim to give cloud providers and regional operators a scalable, interoperable path to deploy massive agent workloads. This could reshape data-center density economics and speed time-to-deployment for hyperscalers and neo-clouds.

Summary

ARM unveiled a purpose-built data-center CPU optimized for agentic AI, leveraging 136 Neoverse V3 cores and a balanced I/O and memory design to prioritize low-latency inference and power efficiency. The company is delivering OCP-compliant reference servers including a oneU design and both air- and liquid-cooled rack configurations—promising up to ~8,000 high-performance ARM cores in a 36 kW air-cooled rack and liquid-cooled designs scaling into the tens of thousands of cores. The project was developed in close collaboration with Meta and partners, and ARM announced a deployment partnership with European NeoCloud Verta alongside NVIDIA GB300 racks to enable heterogeneous, interoperable AI deployments. ARM is also pushing OCP standards, chiplet enablement, and server standardization to accelerate ecosystem adoption and faster silicon-software co-development.

Original Description

Presenter: Eddie Ramirez , VP of Marketing- Cloud AI Business Unit- Arm

Eddie Ramirez, VP of Marketing- Cloud AI Business Unit- Arm

AI is accelerating the shift toward modular- open data center architectures built for scale- efficiency- and rapid innovation. In this keynote- Arm will highlight new Arm-basedsilicon and system designs that address the growing demands of AI and cloud workloads through system-level co-design with premier hyperscale partners. Arm will share how its contributions to the Open Compute Project are helping shape the foundational building blocks of open infrastructure- spanning silicon- chiplets- systemarchitecture- firmware- and security. Through active leadership in initiatives such as the Foundation Chiplet System Architecture (FCSA)- Arm is working with the ecosystem toenable interoperable- modular platforms that reduce fragmentation and accelerate adoption across data center environments.The keynote will also spotlight the continued momentum of Arm Total Design and ATD- aligned chiplet ecosystems- demonstrating how open collaboration across the supplychain is accelerating time to market. Together with the OCP community- Arm will outline a clear path toward AI infrastructure that is more scalable- sustainable- and built in theopen.