The integration of NVIDIA AI Enterprise into the Oracle Cloud Infrastructure console, alongside new superclusters with up to 131,072 Blackwell GPUs, enables businesses and developers to accelerate the training and deployment of advanced AI, ushering in a new era of innovation and productivity.
Oracle and NVIDIA have announced the expansion of their strategic collaboration to provide businesses and developers with access to the infrastructure and software needed to accelerate innovation in artificial intelligence. Through the native integration of the NVIDIA AI Enterprise platform directly into the Oracle Cloud Infrastructure (OCI) console, users can now deploy over 160 AI tools for training and inference with maximum flexibility and performance.
Direct and Simplified Access to NVIDIA AI on Oracle Cloud
Unlike other platforms, where NVIDIA AI Enterprise is typically available as an additional service or in a marketplace, Oracle has integrated it directly into the OCI console. This allows customers to acquire the service using their Oracle universal credits and benefit from centralized billing and support, reducing the complexity and time required for launching AI projects.
With this integration, businesses and developers can easily access optimized microservices like NVIDIA NIM to deploy generative AI models, combining these capabilities with OCI’s native services to build applications, manage data, and meet security and regulatory requirements, even in distributed cloud environments, sovereign regions, and dedicated settings.
AI Superclusters: 131,072 Blackwell GPUs and NVIDIA DGX Cloud Lepton
As part of its strategy to democratize access to large-scale AI, Oracle offers its customers NVIDIA GB200 NVL72 systems on the OCI Supercluster, reaching up to 131,072 Blackwell GPUs. This infrastructure is capable of supporting both the training of cutting-edge language models and the execution of advanced AI agents, providing unprecedented performance, scalability, and energy efficiency.
Oracle also becomes one of the first major cloud providers to integrate the new NVIDIA DGX Cloud Lepton platform, connecting developers worldwide with a global marketplace for GPU resources, facilitating collaborative projects, scientific research, and strategic deployments in specific regions to address digital sovereignty needs.
Real Cases: AI Applied in European Businesses
The partnership is already showing tangible results in clients across various sectors. Almawave, a leading Italian company in artificial intelligence, uses OCI infrastructure and NVIDIA Hopper GPUs to train and infer multilingual generative models called Velvet, adapted to European languages and applied in sectors such as health, finance, and public services. “Oracle and NVIDIA are key partners in our commitment to building high-performance, transparent AI, fully integrated into the European context,” says Valeria Sandei, CEO of Almawave.
In Denmark, the health tech company Cerebriu is utilizing OCI and NVIDIA Hopper GPUs to accelerate the clinical analysis of brain MRIs through deep learning models trained on thousands of images. “OCI and NVIDIA provide us with the infrastructure and performance necessary to advance our product strategy, exploring new use cases for AI efficiently and at scale,” highlights Marko Bauer, a machine learning researcher at Cerebriu.
More Accessible, Flexible, and Regulated AI
The collaboration between Oracle and NVIDIA not only aims for performance but also facilitates regulatory compliance in sensitive sectors, deploying AI in public, sovereign, dedicated, or edge clouds, depending on each organization’s needs.
Companies can combine OCI’s AI, storage, and computing services with NVIDIA’s tools and libraries, creating a robust foundation for artificial intelligence initiatives worldwide.
Conclusion
In this way, Oracle and NVIDIA solidify a strategic alliance that positions both companies at the forefront of infrastructure and software for enterprise artificial intelligence. From the direct integration of NVIDIA AI Enterprise into OCI to the availability of Blackwell GPU superclusters, the goal is clear: to eliminate technical and access barriers, accelerating the development of more powerful, efficient AI models that are aligned with each client’s regulatory and business requirements.