轨道数据中心远比硅谷想象的困难：物理原理和成本障碍使太空AI计算不切实际
英文摘要
Despite recent announcements from Nvidia, SpaceX, Google, and startup Starcloud about building orbital data center constellations with AI GPUs, a closer look at the physics reveals major challenges. The claimed benefit of free cooling in space is a misconception: in a vacuum, only radiative cooling works, requiring huge radiator surfaces to prevent chip overheating. Solar power requires complex sun-tracking systems, and cosmic rays degrade panels, radiators, and chips. Space maintenance is extremely difficult, necessitating redundant systems, and a rough cost comparison shows running AI GPUs in space costs at least an order of magnitude more than on Earth. Orbital data centers may have niche uses but are currently economically unviable.
中文摘要
尽管英伟达、SpaceX、谷歌和初创公司Starcloud相继宣布建设由数千颗搭载AI GPU的卫星组成的轨道数据中心星座，但从物理原理分析，太空中免费的冷却是一个误解：真空环境下只有辐射散热有效，需要巨大的散热表面来防止芯片过热。太阳能需要复杂的太阳跟踪系统，宇宙射线会降低太阳能板、辐射冷却器和芯片的性能。太空维护极其困难，必须配备冗余系统。粗略成本比较显示，太空中运行AI GPU一年的成本比地面数据中心至少高出一个数量级。轨道数据中心在特定领域可能有用，但经济上不可行。
关键要点
Nvidia, SpaceX, Google, and Starcloud have announced plans for orbital AI data centers, but the physics of space poses serious challenges.
英伟达、SpaceX、谷歌和Starcloud宣布建设轨道AI数据中心，但太空物理环境带来严峻挑战。
Free space cooling is a myth: without atmosphere, conduction and convection don’t work; only radiative cooling is possible, requiring impractical radiator sizes.
太空免费冷却是一种误解：没有大气，传导和对流均无效，只有辐射散热可行，需要不切实际的庞大散热面积。
Solar power in orbit demands complex sun-tracking and suffers efficiency loss from cosmic rays.
轨道太阳能需要复杂的太阳跟踪系统，且宇宙射线会降低太阳能板、冷却器和芯片的性能。
Space maintenance is extremely difficult, forcing redundant designs; overall cost per AI GPU-year in orbit is at least 10× higher than on the ground.
太空维护极其困难，迫使采用冗余设计；太空中每个AI GPU年的总成本比地面高出至少一个数量级。
Orbital data centers may serve niche purposes but are economically unviable for general AI workloads.
轨道数据中心可能有特定用途，但对通用AI工作负载而言在经济上不可行。