OpenAI and Oracle to Build Massive AI Data Center with Tens of Thousands of NVIDIA Chips

OpenAI and Oracle have unveiled an ambitious plan to construct a state-of-the-art AI data center in Abilene, Texas, which will house hundreds of thousands of NVIDIA's high-performance AI chips. This initiative is part of their $100 billion Stargate infrastructure project, marking a significant step toward expanding AI computing capabilities.

Massive AI Chip Deployment in Stages

According to sources, the new facility will feature 64,000 NVIDIA GB200 AI chips by the end of 2026. The rollout will occur in multiple phases, with an initial batch of 16,000 chips expected to be deployed by mid-2025. These chips will be distributed across multiple data center halls, ensuring scalability and efficient computing power allocation.

The Significance of the Stargate Project

The Stargate infrastructure project is a groundbreaking investment designed to accelerate AI model training and inference. As demand for AI-driven applications, large-scale computing, and cloud services grows, this AI supercomputing hub will play a critical role in supporting next-generation generative AI models.

Oracle’s Cloud Infrastructure and OpenAI’s AI Ambitions

Oracle’s cloud infrastructure will serve as the backbone of the new data center, enhancing AI model deployment efficiency. This partnership strengthens OpenAI’s ability to train and optimize future versions of its AI models, including the anticipated GPT-5 and beyond.

With NVIDIA's cutting-edge GB200 AI chips, this facility is poised to become one of the world's most powerful AI data centers, reinforcing the U.S.'s leadership in AI innovation and high-performance computing.