Oracle and NVIDIA Partner to Accelerate Scalable Agentic AI with New Cloud Integrations
Oracle and NVIDIA have unveiled a game-changing collaboration aimed at empowering enterprises to rapidly deploy agentic AI applications. Through a robust integration of NVIDIA’s AI Enterprise platform and Oracle Cloud Infrastructure (OCI), businesses can now access over 160 AI tools and 100+ NVIDIA NIM™ inference microservices natively via the OCI Console.
This partnership enables organizations to develop, test, and scale AI models more efficiently using Oracle’s secure and scalable infrastructure combined with NVIDIA’s cutting-edge AI and GPU technologies.
“NVIDIA and Oracle are ideal partners for this new era of reasoning,” said NVIDIA CEO Jensen Huang. “Together, we’re unlocking the potential of agentic AI to drive innovation.”
Streamlined AI Workflows with NIM Microservices
NVIDIA AI Enterprise’s integration into OCI supports rapid model deployment through no-code AI Blueprints and optimized deployment images. This includes NVIDIA NIM, a suite of inference microservices that support large models like the NVIDIA Llama Nemotron for advanced reasoning tasks.
These services are deployable across Oracle’s public cloud, dedicated regions, edge devices, and sovereign clouds—addressing growing demands for data privacy, low-latency processing, and distributed AI workloads.
Blueprints for Fast-Track AI Deployment
Oracle’s prebuilt AI Blueprints offer plug-and-play AI deployment recipes that eliminate the guesswork in infrastructure provisioning. Paired with NVIDIA’s hardware-optimized reference designs, enterprises can deploy multimodal models in minutes instead of weeks.
For instance, companies like Pipefy are already using these blueprints to streamline image and document processing, achieving faster AI integration with minimal manual setup.
Vector Search Boosted by NVIDIA Acceleration
A key highlight is the integration of NVIDIA cuVS to accelerate AI Vector Search in Oracle Database 23ai. Enterprises can now perform large-scale vector embeddings and index creation with GPU-powered speed, enabling faster insights across datasets including text, video, and images.
Media intelligence firm DeweyVision is using this to power its no-code AI media warehousing, citing faster ingestion and search performance.
OCI as a Global AI Powerhouse
Oracle will be among the first cloud providers to support the NVIDIA Blackwell Ultra GPUs, including the GB300 NVL72 system that outpaces its predecessor in AI performance by 1.5x. The collaboration also spans high-performance NVIDIA InfiniBand networks and scalable AI superclusters hosted on OCI Supercluster, featuring over 130,000 GPUs.
Enterprise Adoption and Real-World Impact
Organizations like Soley Therapeutics and SoundHound AI are leveraging the OCI-NVIDIA stack to fuel groundbreaking AI applications in healthcare and voice intelligence. By integrating NVIDIA’s AI tools with Oracle’s infrastructure, these companies are gaining faster time-to-market and enterprise-grade scalability
