Today, Oracle announced the opening of the first zettascale cloud computing clusters in history. GPUs made by NVIDIA Blackwell power this cluster. It provides peak performance of 2.4 zettaFLOPS and offers up to 131,072 GPUs. 

With the help of NVIDIA GPUs and Oracle's strong AI infrastructure, companies can handle massive AI workloads with more flexibility and autonomy.  Oracle Cloud Infrastructure's EVP, Mahesh Thiagarajan, stated that the company supports customers running some of the most demanding AI workloads in the cloud and has one of the most comprehensive AI infrastructure capabilities. 

Through the integration of Oracle's cloud infrastructure with NVIDIA's cutting-edge GPU architecture, they are providing scalable AI computing solutions that are essential for researchers and enterprises worldwide. NVIDIA's vice president of hyperscale and high-performance computing, Ian Buck, stated that the company's full-stack AI computing platform on Oracle's widely distributed cloud will enable AI compute capabilities on a never-before-seen scale. 

Oracle's latest innovation promotes cutting-edge AI research and development while maintaining regional data sovereignty, an essential component for sectors such as healthcare and collaboration platforms like Zoom and WideLabs, among others. 

Six times more NVIDIA B200 GPUs are available from Oracle Cloud Infrastructure (OCI) than from other cloud hyperscalers like AWS, Azure, and Google Cloud. 

For instance, OCI offers more than three times as many GPUs as AWS UltraClusters, allowing for hitherto unheard-of levels of computational power.

Oracle also claimed that it is the first to offer a zettascale AI supercomputer, with competitors only able to achieve exascale performance—2.4 zettaFLOPs, to be exact.