October 25–28 | Las Vegas

NVIDIA at Oracle AI World 2026

Learn how you can quickly and easily harness agentic AI with Oracle Cloud Infrastructure and NVIDIA.

NVIDIA and Oracle Cloud Infrastructure

Join us at the Venetian Convention and Expo Center in Las Vegas to learn how NVIDIA and Oracle Cloud Infrastructure (OCI) are partnering to expand access to accelerated computing in the cloud. Together, we’re empowering organizations across industries to build next-generation AI applications while meeting performance, security, and sovereignty requirements.

Hear From NVIDIA Experts

Spotlight Sessions

Optimizing the Economics of AI Inference With NVIDIA Rack-Scale Architecture on OCI

Tuesday, October 27, 1:00–1:45 p.m.

The economics of AI inference are reshaping how organizations build and scale AI applications. This session unpacks how OCI, powered by NVIDIA’s latest rack-scale architecture, maximizes tokens per watt and delivers cost-competitive AI at scale. We’ll walk through what drives real-world inference economics, the key metrics to consider when evaluating AI economics, and how customers are making the most of OCI’s AI infrastructure.

Explore the Session

From Hours to Minutes: Turbocharging Oracle Database Vector Search With NVIDIA Accelerated Computing

Tuesday, October 27, 1:00–1:45 p.m.

Oracle AI Database's vector index service with GPU support brings AI capabilities directly to enterprise data. This session covers how NVIDIA cuVS GPU-accelerated vector search integrates with Oracle Database to dramatically speed up index generation and retrieval-augmented generation (RAG) pipelines, with live performance benchmarks.

Explore the Session

800,000 GPUs, One Mission: Building the Infrastructure for Frontier AI [LRN2628]

Oracle Cloud Infrastructure (OCI) is building AI infrastructure powered by NVIDIA rack-scale architecture. From OCI Superclusters supporting up to 800,000 NVIDIA GPUs and delivering up to 16 zettaFLOPS of performance to the architectural evolution of next-generation GPU and accelerator technologies, this session explores how hyperscale AI infrastructure is designed, deployed, and operated in practice. Learn how OCI and NVIDIA are advancing the infrastructure needed to train and deploy frontier AI models at massive scale.

Explore the Session

Agents Need Infrastructure Too: Building the Backbone for Agentic AI [LRN2412]

Agentic AI is reshaping how enterprises design and deploy modern applications. In this 45-minute learning session, you'll explore how to evaluate agentic AI workload profiles, including orchestration, retrieval, routing, small-model inference, data processing, and application logic to determine the right infrastructure for each stage of the workflow. Learn how to combine OCI Core Infrastructure and Cloud Native Services, when CPU-based compute is the best fit, which OCI compute shapes support different agentic AI patterns, and how to optimize performance, cost, scalability, and operational efficiency across enterprise AI deployments.

Explore the Session

Workshops

Run Vector Search With GPU Acceleration Using Oracle AI Database and NVIDIA cuVS

In this hands-on lab, you’ll configure Oracle AI Database 26ai to offload index creation to NVIDIA GPUs using NVIDIA cuVS-powered Vector Index Service. You’ll load embeddings, build a Hierarchical Navigable Small World (HNSW) vector index, compare GPU and CPU index-build performance, and execute semantic similarity searches using SQL. The lab demonstrates how organizations can accelerate the preparation of large vector datasets while keeping vectors, metadata, and business data governed within Oracle AI Database.

Search and Summarize Video With Oracle AI Database and NVIDIA VSS

In this hands-on lab, you’ll build an AI-powered video discovery workflow using the NVIDIA Video Search and Summarization (VSS) Blueprint. NVIDIA VSS processes footage to generate scene descriptions, timestamps, transcripts, and embeddings, which are stored in Oracle AI Database 26ai, with NVIDIA cuVS accelerating vector-index creation. You’ll use natural-language prompts to locate specific moments, retrieve relevant clips, and generate summaries, demonstrating how media organizations can make large video archives easier to discover and reuse.

Building Efficient AI Agents: Model Routing With NVIDIA NeMo Switchyard

In this hands-on lab, you’ll learn how to route AI agent workloads intelligently so you're not sending every step to your most expensive model. We'll walk through how NVIDIA NeMo™ Switchyard on OCI routes execution steps to NVIDIA Nemotron™ 3.5 Lightning and escalates only complex planning steps to a frontier model, with hands-on examples showing how this approach balances model capability, latency, and cost.

Find More Resources

Access Free Tools and Resources

The NVIDIA Developer Program provides comprehensive resources, tools, and a community of experts to help you build accelerated applications and solutions. Join and get a complimentary course from the NVIDIA Deep Learning Institute.

Learn More About the Developer Program

Explore NVIDIA Training and Certification

The NVIDIA Deep Learning Institute offers hands-on, self-paced courses, instructor-led workshops, and technical certifications to help you develop practical skills and validate your expertise.

Learn More About the NVIDIA Deep Learning Institute

Discover NVIDIA’s Free Program for Startups

NVIDIA Inception is a global program for AI startups that provides developer training and resources, valuable offers from NVIDIA and our partners, and preferred access to NVIDIA’s AI ecosystem.

Join NVIDIA Inception Now