The planned offering would add Arm compute for cloud applications and agentic AI workloads on OCI.
At Oracle AI World, Oracle outlined plans to bring Arm AGI CPU to Oracle Cloud Infrastructure (OCI) in fall 2027. The planned offering will provide Arm compute for cloud applications and agentic AI services.
The announcement follows COMPUTEX, where OCI was added to the Arm AGI CPU ecosystem and the collaboration was extended with a roadmap for bringing that compute to Oracle customers.
Arm’s role across OCI
Arm is already part of OCI’s compute portfolio. Ampere-powered Arm offerings support workloads across enterprise applications, containers, web services and data processing. NVIDIA’s Arm-based Vera CPUs, now generally available on OCI, add compute for agentic AI and enterprise workloads. The planned addition of Arm AGI CPU would expand this portfolio further.
Compute for cloud and agentic AI
Cloud workloads and AI create different CPU demands. Agents rely on applications, databases and services to complete tasks. They also add CPU demand when they execute code, retrieve data, invoke tools and coordinate work. Infrastructure needs to support these workloads at scale.
Arm AGI CPU combines Arm Neoverse cores, high memory bandwidth and a design intended for sustained performance under load. These capabilities are intended to support cloud applications and agentic systems within data center power and cooling limits.
The expansion also builds on OCI’s software foundation. OCI’s Arm investment includes Oracle Linux, Java, MySQL and cloud-native tools, giving developers a familiar environment as compute options expand.
The addition of Arm AGI CPU would provide another compute option across the Arm ecosystem. With Oracle, Arm is extending that option across applications enterprises use today and the AI services they are building for the future.
For more information, visit arm.com.
