英伟达发布Cosmos Reason：AI机器人进入推理时代

Core Insights - Nvidia has launched a new generation of global AI models and infrastructure for robot developers, notably the "Cosmos Reason," a reasoning visual language model with 7 billion parameters aimed at physical AI applications and robotics [1] - The models are designed to create synthetic text, image, and video datasets for training robots and AI agents, enhancing their reasoning capabilities [1] Group 1: Product Features - Cosmos Reason enables robots and AI agents to infer actions based on physical understanding and memory, allowing them to plan and execute tasks effectively [1] - The model can automate data organization and labeling, significantly improving the efficiency of processing diverse training datasets [2] - In video analysis, AI agents built on Nvidia's Blueprint can extract valuable insights from large volumes of recorded or real-time video, suitable for urban traffic networks and factories [2] Group 2: Technological Advancements - Nvidia introduced a new neural reconstruction library and rendering technology that allows developers to simulate the real world in 3D using sensor data [2] - The updates to the Omniverse software development kit and the launch of the RTX Pro Blackwell Server provide a unified architecture to support robotic development workloads [2] - DGX Cloud is introduced as a cloud-based management platform for robotics, indicating Nvidia's commitment to enhancing robotic workflows [2] Group 3: Strategic Direction - The releases indicate that Nvidia is accelerating its entry into the robotics sector, seeking new application scenarios for AI GPUs following AI data centers [3]