Oracle to Invest $40 Billion in Nvidia Chips for AI Center

The American technology giant Oracle has announced plans to allocate approximately $40 billion towards acquiring advanced high-performance computer chips from Nvidia. This investment is intended to support the development of a massive new data center focused on artificial intelligence (AI).

Located in Abilene, Texas, the facility is part of a significant initiative known as 'Stargate,' which Oracle is spearheading in collaboration with leading AI developer OpenAI and Japanese investor Softbank. The project was first unveiled at the White House, highlighting its importance to the evolving landscape of AI infrastructure.

Upon completion, the data center is expected to generate a remarkable 1.2 gigawatts of power, positioning it among the largest AI data centers globally. As part of this ambitious undertaking, Oracle plans to procure around 400,000 GB200 chips from Nvidia, specially designed for training and operating extensive AI systems. The processing power derived from these chips will then be leased to OpenAI.

This data center represents a strategic move for OpenAI, as it aims to reduce its reliance on Microsoft, which has been a pivotal financial backer and partner for the AI organization led by Sam Altman. By establishing this new facility, OpenAI seeks to diversify its resources and capabilities in the rapidly evolving AI sector.

In the United States, numerous corporate initiatives are vying to enhance the country's focus on AI and build the necessary infrastructure. Alongside Oracle and OpenAI, major players like Microsoft, Google, and Amazon are actively involved in this competitive landscape. For instance, Elon Musk is reportedly expanding his 'Colossus' data center in Memphis, with plans to accommodate approximately one million Nvidia chips. Musk has characterized the next phase of Colossus as the 'first gigawatt AI training supercluster.'

Moreover, Amazon is currently in the process of constructing a data center in Northern Virginia, projected to exceed one gigawatt in capacity. These developments underscore the escalating race among tech companies to establish robust infrastructures capable of supporting advanced AI applications and research.