Amidst the tide of the new industrial revolution, Jensen Huang, the founder of NVIDIA, emphasized at the GTC conference in Berlin that AI computing is propelling infrastructure development into a novel era. He unveiled that the Vera Rubin platform has commenced full-scale production and introduced the DSX POD solution, which is built upon this platform. Through integrated factory verification of computing, networking, power supply, and liquid cooling systems, this solution drastically cuts down the on-site deployment time of AI factories, trimming it from the initial six-month period to a swift 14 days. The Vera Rubin platform features a modular design, slashing the assembly time of a single computing node from two hours to under five minutes, and boasts 100% liquid cooling system coverage. Its NVLink switch enables the connection of 72 GPUs to construct a unified computing system. Lenovo has initiated global shipments of the Vera Rubin NVL72 and has seamlessly integrated it into its proprietary AI Cloud Superfactory solution. As a global partner for NVIDIA's DSX POD, Lenovo aims to facilitate the worldwide rollout of the solution by early 2027, thereby broadening its product range and enhancing its delivery prowess in the AI infrastructure domain. As AI computing systems transition towards high-density racks and comprehensive AI factories, system integration, liquid cooling technology, mass production capabilities, and global delivery proficiency will emerge as pivotal factors for undertaking large-scale AI infrastructure endeavors.