Google’s October 1 launch of its Project Suncatcher prototype—carrying just four Trillium TPUs aboard a SpaceX rideshare—underscores why traders assign only a 2% chance of an orbital AI data center meeting the 100-accelerator threshold by end-2026. Earlier single-chip tests from Starcloud (Nvidia H100, late 2025) and smaller Chinese or Axiom nodes similarly fall short of data-center scale. Larger ambitions, including SpaceX’s FCC filing for up to one million compute satellites and Google’s planned 2027 follow-ons, face steep barriers in launch economics, vacuum cooling, radiation hardening, and inter-satellite optical links. The 25% implied probability for a 2027 resolution reflects credible hardware progress and regulatory pilots, yet traders recognize that moving from refrigerator-sized prototypes to constellations delivering meaningful AI inference or training remains a multi-year engineering and capital challenge.
如果任何轨道数据中心在指定日期 23:59 ET 前成功发射,本市场将判定为“是”。否则,此市场将判定为“否”。 “轨道数据中心”是指任何搭载计算基础设施、发射至地球轨道,旨在提供数据中心、云计算或人工智能计算服务,且包含至少 100 个数据中心级 AI 加速器、GPU、TPU 或实质上同等的计算处理器的航天器、卫星或同等技术(例如 NVIDIA H100 GPU、Google TPU,或同等或后续芯片)。 “成功发射”是指任何成功将符合条件的轨道数据中心送入地球轨道的发射。 本市场结算的主要依据将是可信报道形成的共识。
机器翻译。如有争议,以原文为准。