Intel and Oracle plan multiyear Xeon expansion on OCI for agentic AI

What the two companies announced

Intel and Oracle said at Oracle AI World in Las Vegas that they plan a multiyear collaboration covering current and future Xeon platforms on OCI. According to the companies, the effort is intended to meet growing demand for enterprise applications, agentic AI and inference.

The companies say they intend to coordinate roadmaps and processor capacity planning to support OCI's expansion, building on what they describe as a decades-long relationship. The announcement is framed as a statement of intent: the release is full of forward-looking language such as "plan" and "intend", and it warns that the expected benefits may not be realized.

Why CPUs matter for AI agents

Intel argues that as more enterprises deploy AI agents at scale, CPUs are becoming increasingly critical for executing applications, managing data movement, orchestrating agentic actions and supporting inference across existing business systems.

Kevork Kechichian, executive vice president and general manager of Intel Data Center Group, said that "Xeon brings together performance and hardware-based confidential computing, giving organizations a trusted, cost-effective foundation for AI at scale."

Pradeep Vincent, executive vice president for compute and AI infrastructure at Oracle Cloud Infrastructure, said that "Enterprise AI depends on the applications, data, and tools that let an agent complete useful work." He added that the collaboration is intended to help Oracle grow the compute capacity behind those workloads.

What Diamond Rapids and OCI X14 are expected to bring

The planned Diamond Rapids-based OCI X14 offerings are expected to bring advances in both AI acceleration and general-purpose computing. The companies point to two features:

• Intel Advanced Matrix Extensions (AMX): enhanced, with FP8, FP16 and BF16 support for optimized inference software.
• Intel Advanced Performance Extensions (APX): intended to help optimized applications execute with less instruction overhead.

Together, these are expected to support embedding generation, reranking and smaller-model inference alongside the application services and tools that AI agents use.

What it means for OCI customers

According to the companies, the approach lets customers extend existing enterprise systems with AI while keeping familiar software environments and operational tooling. Oracle says customers are expected to gain a path to build and scale AI using the enterprise software environments they already rely on.

OCI's current Xeon-based offerings pair Intel compute with built-in AI acceleration, flexible CPU and memory configurations, and Oracle Acceleron networking. Organizations can tailor application services, data processing and agent execution to their needs and scale each independently of the accelerators serving larger models. The companies say this lets customers better match resources to each stage of an AI workflow and manage overall application costs.

The release gives no dates, pricing or performance figures for X14. Oracle's disclaimer states that the development, release and timing of any features described remain at Oracle's sole discretion and should not be relied upon in purchasing decisions.