Oracle and NVIDIA have deepened their collaboration to enhance the accessibility, power, and practicality of enterprise AI solutions. Announced at Oracle AI World, their partnership spans from cutting-edge hardware innovations to seamless software integrations designed to embed AI at the heart of organizational data ecosystems.
Unleashing Unprecedented AI Computing Power with OCI Zettascale10
At the forefront of this alliance is the introduction of the OCI Zettascale10 computing cluster. Fueled by NVIDIA GPUs, this platform is engineered to handle the most demanding AI training and inference tasks that would overwhelm conventional servers. Delivering an astonishing peak performance of 16 zettaflops, Zettascale10 leverages NVIDIA’s Spectrum-X Ethernet-a specialized networking fabric that eliminates GPU idle time by ensuring rapid data flow. This architecture enables enterprises to efficiently scale AI workloads across millions of processors.
Integrating AI Seamlessly into Business Operations
While raw computational power is impressive, the true value lies in the sophisticated software integrations that embed AI throughout business processes. Mahesh Thiagarajan, Executive VP of Oracle Cloud Infrastructure, highlights that OCI Zettascale10 not only offers multi-gigawatt capacity for complex AI workloads but also provides native access to NVIDIA AI Enterprise within OCI’s extensive cloud service portfolio. This integration empowers customers to innovate rapidly using a comprehensive AI toolkit.
Revolutionizing Data Management with Oracle AI Database 26ai
Central to Oracle’s AI strategy is the Oracle AI Database 26ai, which challenges the traditional approach of relocating data to AI models. Instead, Oracle advocates for bringing AI directly to the data, enhancing security and efficiency. This database embodies the “AI for Data” philosophy by enabling enterprises to run intelligent workflows within their data repositories.
Juan Loaiza, Executive VP of Oracle Database Technologies, explains that this architecture simplifies AI adoption by combining data and AI in a unified environment. It supports trusted AI insights across operational systems and analytic data lakes, making AI accessible and reliable for diverse datasets.
Advanced AI Capabilities Within the Database
A standout feature of Oracle AI Database 26ai is its support for agentic AI workflows that operate inside the database. These AI agents can answer complex queries by synthesizing sensitive enterprise data with external information without ever exposing private data outside secure boundaries. This is facilitated by the Unified Hybrid Vector Search, which enables AI to search across various data formats-be it relational tables, JSON documents, or geospatial data.
Security is a paramount concern, and Oracle addresses this by implementing NIST-approved quantum-resistant encryption algorithms for data both in transit and at rest. This proactive measure protects against future quantum computing threats that could compromise encrypted data.
Enhancing AI Development with NVIDIA Integration
Oracle AI Database 26ai is designed to integrate seamlessly with NVIDIA’s AI tools, including the NeMo Retriever microservices. This integration simplifies complex AI tasks such as retrieval-augmented generation (RAG), where language models consult relevant company documents to provide more accurate and context-aware responses.
Additionally, the Oracle Private AI Services Container will soon harness NVIDIA GPUs to accelerate the creation of vector embeddings-a critical step in AI-powered search-using the cuVS library. This enhancement significantly reduces data preparation time for AI applications.
Streamlining AI Adoption Across the Enterprise
Beyond databases, the Oracle-NVIDIA partnership aims to democratize AI by simplifying the entire AI lifecycle. The Oracle AI Data Platform now offers an integrated NVIDIA GPU option alongside the NVIDIA RAPIDS Accelerator for Apache Spark, enabling data scientists and engineers to accelerate machine learning workflows with minimal code changes.
All these capabilities are unified within the Oracle AI Hub, a centralized platform where organizations can build, deploy, and manage AI solutions. The hub features NVIDIA’s NIM microservices-pre-packaged AI functionalities accessible through an intuitive no-code interface.
To further reduce barriers, the full NVIDIA AI Enterprise software suite is now embedded directly within the OCI Console. This integration allows developers to launch GPU instances and activate NVIDIA tools instantly, bypassing lengthy procurement processes and accelerating project kickoff.
Driving Practical, Secure, and Scalable Enterprise AI
This collaboration addresses the real-world challenges companies face when implementing AI by uniting powerful hardware, secure data management, and comprehensive software tools into a cohesive ecosystem. Oracle and NVIDIA’s joint efforts signal the arrival of a new era where enterprise AI is not only feasible but also secure, scalable, and deeply integrated into business operations.
