NVIDIA and Oracle Unveil AI and Data Processing Innovations at Oracle CloudWorld

In a move aimed at pushing the limits of AI and data processing, NVIDIA and Oracle have teamed up to launch the first zettascale Oracle Cloud Infrastructure (OCI) Supercluster. OCI will be powered by NVIDIA’s Blackwell platform.

Unveiled at the Oracle CloudWorld conference, this collaboration is set to empower enterprises with the computational power required to train and deploy advanced AI models, leveraging more than 100,000 of NVIDIA’s latest-generation GPUs.

OCI Superclusters: Pioneering AI Compute

OCI Superclusters offer unprecedented scalability. It allows businesses to deploy a variety of NVIDIA GPUs on-premises, in public clouds, or within sovereign clouds. Expected to be available in the first half of next year, these Blackwell-based systems can scale up to 131,072 GPUs, supported by NVIDIA ConnectX-7 NICs or Quantum-2 InfiniBand networking.

This configuration delivers a staggering 2.4 zettaflops of peak AI compute power, setting a new standard in the cloud computing landscape. Oracle also previewed NVIDIA GB200 NVL72 liquid-cooled instances designed to support generative AI applications.

With the capability to handle large-scale training and real-time inference for trillion-parameter models, these instances highlight Oracle’s commitment to pushing the boundaries of AI innovation.

Expanding AI Capabilities with New GPU Instances

Further enhancing their AI offerings, Oracle announced the availability of NVIDIA HGX H200 instances. These feature eight H200 Tensor Core GPUs connected via NVLink and NVLink Switch, capable of scaling to 65,536 GPUs.

These instances are designed for real-time inference and accelerated training workloads. In turn, providing enterprises with the tools needed to drive AI innovation at scale. Additionally, Oracle introduced NVIDIA L40S GPU-accelerated instances, aimed at midrange AI workloads, NVIDIA Omniverse, and visualization tasks.

Oracle’s Roving Edge Device v2 now supports up to three NVIDIA L4 Tensor Core GPUs. This brings scalable AI capabilities even to remote and disconnected locations. This edge solution allows enterprises to harness AI power in various environments, from data centers to the most challenging field conditions.

Real-World Applications of NVIDIA-Powered OCI Superclusters

Several companies are already leveraging NVIDIA-powered OCI Superclusters to accelerate their AI initiatives. For example, Reka, a startup developing multimodal AI models, utilizes these clusters to build enterprise agents that can interact with the world through reading, seeing, hearing, and speaking.

“Reka’s multimodal AI models, built with OCI and NVIDIA technology, empower next-generation enterprise agents that can read, see, hear and speak to make sense of our complex world,” said Dani Yogatama, co-founder and CEO of Reka.

Advancing Generative AI with Oracle Database

Oracle is also integrating NVIDIA GPUs into its Autonomous Database, allowing enterprises to accelerate data processing directly within Oracle Machine Learning notebooks. This integration supports AI workloads through various demonstrations.

This includes vector embeddings, vector graph index generation, and enhanced inference capabilities using NVIDIA NIM microservices. These advancements enable companies to bring generative AI to both structured and unstructured data within Oracle Database environments.

Global Reach and Sovereign AI

NVIDIA and Oracle are extending their AI infrastructure worldwide, addressing the data residency needs of governments and enterprises. From training large language models for Brazilian Portuguese in OCI’s data centers to enhancing financial AI platforms in Japan.

The companies are providing sovereign AI solutions that comply with local regulations and data sovereignty requirements. Notable clients include Wide Labs in Brazil, Nomura Research Institute in Japan, and Zoom in Saudi Arabia, each utilizing NVIDIA GPUs to meet regional data compliance standards.

Conclusion

The partnership between NVIDIA and Oracle represents a significant leap forward in AI and data processing capabilities. It offers enterprises the tools to innovate and scale their AI initiatives. For those interested in exploring these advancements further, NVIDIA will be showcasing its collaboration with Oracle at the Oracle CloudWorld 2024.

Originally posted on OpenDataScience.com

Read more data science articles on OpenDataScience.com, including tutorials and guides from beginner to advanced levels! Subscribe to our weekly newsletter here and receive the latest news every Thursday. You can also get data science training on-demand wherever you are with our Ai+ Training platform. Interested in attending an ODSC event? Learn more about our upcoming events here.