Texas-based Oracle Corporation (NYSE:ORCL), under the leadership of Chairman and CTO Larry Ellison, is making significant strides in generative AI technology. The company’s progress is marked by its lucrative partnerships with Nvidia (NASDAQ:NVDA) and Elon Musk’s xAI, despite facing stiff competition from other tech giants such as Amazon (NASDAQ:AMZN) Web Services, Google (NASDAQ:GOOGL) Cloud, and Microsoft (NASDAQ:MSFT) Azure.

On Monday, it was revealed that Oracle has secured contracts exceeding $4 billion for capacity on its Generation 2 Cloud to train generative AI models. These models are linked with platforms like ChatGPT and Bard. This development has contributed to a remarkable doubling of Oracle’s earnings for the fourth quarter of 2023.

Oracle attributes its superior speeds and cost benefits to its remote data memory access (RDMA) network and OCI Compute instances. These instances are powered by Nvidia H100 Tensor Core, Nvidia L40S GPUs, and Ampere AmpereOne CPUs. This GPU interconnection allows Oracle’s cloud to operate at a faster pace than its competitors, leading to improved speed and pricing.

As a result of these advancements, top-tier companies such as Nvidia, Cohere, and xAI are using Oracle Cloud for their generative AI models. Ellison anticipates that one of the Nvidia superclusters currently under construction will soon become the world’s largest scientific computer.