SpaceX 10GW in 2027 – Why It’s Real, Will Drive $300B ARR for SpaceX, and Why Microsoft Will Be the Largest Offtaker

Near-term compute scarcity, not model quality, drives what inference costs — this lays out the supply pipeline and the per-GW revenue math sitting behind API pricing.

• inference — Running a trained model to get answers — the phase where AI is actually used, as opposed to trained.

• articlexAI’s Colossus 2 – First Gigawatt Datacenter In The World, Unique RL Methodology, Capital Raise
• articleAmazon’s AI Resurgence: AWS & Anthropic’s Multi-Gigawatt Trainium Expansion
• articleMeta Compute: Everyone Wants To Be A Neocloud

• postElon shocked the world again on SpaceX's first earnings call: 6-8GW of incrementSemiAnalysis_
• postWhy would anyone pay for that much compute? Inference is unbelievably profitableSemiAnalysis_
• videoDylan Patel — The single biggest bottleneck to scaling AI computeDwarkesh Patel

Checking sign-in…

Loading comments…