On March 16 (local time), Nebius, an AI cloud company, unveiled a partnership with NVIDIA, aiming to create a specialized cloud platform tailored for robotics and embodied artificial intelligence. Prior to this, on March 11, NVIDIA had already declared a substantial $2 billion investment in Nebius. This investment is earmarked to facilitate the deployment of over 5 gigawatts of NVIDIA's accelerated computing systems by the close of 2030. The two entities will embark on a profound engineering collaboration, focusing on AI computing platforms. This collaboration will encompass various domains, including the architectural design of AI factories, the optimization of inference platforms, and the deployment of infrastructure. Their overarching objective is to furnish developers and enterprises with comprehensive, full-stack AI cloud services. Nebius is set to persist in deploying multiple generations of NVIDIA's accelerated computing systems and will secure early access to forthcoming computing architectures (such as the Rubin platform and Vera CPU). This partnership is strategically designed to cater to the burgeoning global demand for intelligent computing and to construct an all-encompassing AI cloud platform that spans from chips to software.