Oracle Launches Groundbreaking Zettascale AI Supercomputer Cluster
Oracle has just introduced an innovative zettascale cloud computing cluster that showcases the immense capabilities of NVIDIA's Blackwell platform. This state-of-the-art AI supercomputer is recognized as the strongest in the cloud, specifically created to assist businesses in building, training, and running AI applications on a massive scale.
Transforming AI with Unrivaled Performance
The Oracle Cloud Infrastructure (OCI) is now ready to reshape the AI environment with the capacity to support up to 131,072 NVIDIA Blackwell GPUs. This configuration offers an astonishing performance potential of 2.4 zettaFLOPS, positioning OCI as a frontrunner in the realm of AI supercomputing. When compared to other competitors, OCI’s supercluster surpasses the Frontier supercomputer, possessing more than three times its GPU count and establishing a new benchmark for cloud-based AI computing.
Cutting-Edge Infrastructure and Features
Organizations can now order OCI Superclusters, which come in various setups, such as those powered by NVIDIA H100 or H200 Tensor Core GPUs. With the H100, users can access up to 16,384 GPUs, achieving up to 65 ExaFLOPS of performance and 13Pb/s network throughput. Additionally, the upcoming H200 GPUs promise to elevate capacities even further, with a potential of 65,536 GPUs and an impressive 260 ExaFLOPS of performance.
AI Sovereignty and Global Cooperation
As competition intensifies, both businesses and researchers are emphasizing the need for access to potent AI computing clusters. With its distributed cloud infrastructure, Oracle enables companies to deploy AI services while maintaining data sovereignty, a vital requirement for organizations adhering to rigorous regulatory standards.
Driving Innovations Across Multiple Industries
Firms like WideLabs and Zoom are effectively utilizing OCI's powerful AI infrastructure for secure cloud computing. In Brazil, WideLabs is training one of the largest Portuguese LLMs, named Amazonia IA, on the OCI platform. This AI helps generate biographical information for Alzheimer’s patients, contributing to the preservation of their memories.
Discovering New Opportunities with OCI
Zoom, a leader in collaborative technology, utilizes OCI to enhance its AI Companion service. This innovative tool aids users in drafting messages and summarizing meetings, dramatically improving productivity. Thanks to OCI's robust capabilities, Zoom can fulfill its data sovereignty needs, especially in regions like Saudi Arabia, where it’s beginning to roll out its solutions.
Shaping the Future of AI and Cloud Computing at Oracle
As organizations worldwide strive to advance AI research and applications, Oracle's latest technology offerings play a crucial role in supporting these ambitions through cutting-edge tools. Mahesh Thiagarajan, Oracle's executive vice president, highlights the importance of flexibility, stating that customers can deploy cloud services wherever they choose while ensuring the highest level of data sovereignty.
Frequently Asked Questions
What makes Oracle's new zettascale cloud computing cluster significant?
Oracle's zettascale cloud computing cluster offers unprecedented performance features that enable organizations to efficiently handle complex AI workloads.
What is the maximum number of GPUs that the OCI Supercluster can support?
The OCI Supercluster can support up to 131,072 NVIDIA Blackwell GPUs, providing substantial processing power of 2.4 zettaFLOPS.
How does OCI ensure data sovereignty for its clients?
OCI enables companies to deploy AI services in line with local regulations, which safeguards sensitive data within designated jurisdictions.
Which organizations are currently benefiting from OCI's infrastructure?
Various organizations, including WideLabs and Zoom, are making the most of OCI's capabilities to improve their AI applications in healthcare and collaborative solutions.
When are NVIDIA Blackwell GPUs expected to be available?
NVIDIA Blackwell GPUs are anticipated to launch in the first half of 2025, marking a significant step forward in AI computing technology.
