Google’s October 1, 2026 launch of a Project Suncatcher prototype satellite carrying four Trillium TPUs aboard a SpaceX Transporter-18 mission represents the most recent catalyst, validating radiation tolerance and basic thermal management for AI chips in low Earth orbit while highlighting that full-scale systems remain distant. The test provides roughly 1 kW to the processors—far below operational needs—and does not meet the market’s 100-accelerator threshold for resolution. Starcloud’s prior H100 flights demonstrated in-orbit LLM training, yet similarly fall short on scale. SpaceX’s FCC filing for up to one million AI satellites and Blue Origin’s Project Sunrise plans underscore competitive momentum, but launch economics, vacuum cooling via radiators, laser inter-satellite links, and regulatory hurdles continue to temper 2027 expectations. Traders are monitoring Google’s planned 2027 dual-satellite follow-on and Starcloud’s next GPU missions for signs of qualifying deployments.
Polymarketデータを参照したAI生成の実験的な要約。これは取引アドバイスではなく、このマーケットの解決方法には一切関係ありません。 · 更新日View resolved

外部リンクに注意してください。
外部リンクに注意してください。
よくある質問