A New Phase in the Oracle–NVIDIA Alliance
At Oracle AI World, Oracle and NVIDIA unveiled an expanded partnership that aims to make enterprise AI faster, more accessible, and deeply integrated across business operations. The collaboration blends NVIDIA’s industry-leading GPUs with Oracle’s cloud infrastructure, bringing advanced AI computing and intelligent database technology into a unified ecosystem.
Ian Buck, Vice President of Hyperscale and High-Performance Computing at NVIDIA, described the partnership as “marking new frontiers in accelerated computing—streamlining database AI pipelines, powering enterprise use cases, and making inference easier to deploy and scale on OCI.”
At the center of the announcement is OCI Zettascale10, Oracle’s new high-performance computing cluster designed for intensive AI workloads such as training large language models and complex inference tasks.
Zettascale10: Raw Power Meets Intelligent Design
The OCI Zettascale10 cluster, accelerated by NVIDIA’s latest GPUs, boasts a staggering 16 zettaflops of peak AI compute performance. Its design leverages NVIDIA Spectrum-X Ethernet, a high-speed networking fabric that keeps GPUs supplied with data efficiently—eliminating common processing bottlenecks and allowing workloads to scale to millions of processors.
Mahesh Thiagarajan, Executive Vice President of Oracle Cloud Infrastructure, said the system “delivers multi-gigawatt capacity for the most challenging AI workloads,” while the native integration of NVIDIA AI Enterprise on OCI gives customers direct access to one of the most comprehensive AI toolsets available in the cloud.
This combination, he noted, supports “a long tail of customer innovation” by linking cutting-edge compute power with over 200 existing Oracle cloud services.
Oracle AI Database 26ai: Bringing AI to the Data
A cornerstone of Oracle’s AI push is its newly launched Oracle AI Database 26ai, which flips traditional AI data workflows on their head. Instead of moving sensitive data to external AI models, Oracle brings the AI to the data—keeping it secure and compliant within enterprise systems.
Juan Loaiza, Executive Vice President of Oracle Database Technologies, said: “By architecting AI and data together, Oracle AI Database makes ‘AI for Data’ simple to learn and simple to use.”
The database allows agentic AI workflows—intelligent agents that operate directly inside enterprise databases. These agents can combine private corporate data with public sources to generate insights, all while keeping sensitive information safely on-site. A key enabler is the Unified Hybrid Vector Search, which lets AI retrieve context across multiple data types, from relational tables to JSON documents and spatial maps.
Oracle has also reinforced the database with quantum-resistant encryption, adopting NIST-approved algorithms to defend against future quantum computing threats—an increasingly relevant security measure in enterprise environments.
Holger Mueller, VP and Principal Analyst at Constellation Research, called the system “the single place where business data lives—current, consistent, and secure,” adding that it “goes beyond vector search by embedding agentic AI and no-code tools directly into the database.”
Deep Integration with NVIDIA AI Tools
The new database works hand-in-hand with NVIDIA NeMo Retriever, a suite of microservices that help enterprises build advanced AI workflows such as retrieval-augmented generation (RAG)—a method that allows language models to fetch relevant facts from company documents in real time.
Meanwhile, the Oracle Private AI Services Container—a secure environment for running private AI models—will soon gain GPU acceleration. Using NVIDIA’s cuVS library, it will handle vector embedding creation much faster, cutting down the time required to prepare data for AI applications.
Simplifying the Enterprise AI Pipeline
Oracle and NVIDIA are also working to democratize AI adoption. The new Oracle AI Data Platform integrates NVIDIA RAPIDS Accelerator for Apache Spark, enabling data scientists to accelerate analytics and machine learning workloads using GPUs—often without changing their existing code.
All of these capabilities come together in the Oracle AI Hub, a central workspace for building, deploying, and managing AI projects. From here, developers can access NVIDIA’s NIM microservices—prepackaged AI skills—through a no-code interface, significantly reducing the barrier to entry for AI experimentation.
Developers can also launch GPU-powered instances with NVIDIA AI Enterprise tools directly from the OCI console, streamlining the entire process from setup to deployment.
A Unified Path to Scalable, Secure AI
The strengthened Oracle–NVIDIA alliance signals a new phase in the enterprise AI race—one focused not just on raw computational power, but on integration, security, and ease of use. By combining Oracle’s data expertise with NVIDIA’s AI acceleration, the companies are positioning themselves as key enablers of the next generation of intelligent enterprise infrastructure.
For businesses looking to harness AI at scale—without compromising on data governance or performance—the partnership represents a significant step toward making secure, practical, and high-impact AI a reality.
