Overview
- Amazon Web Services will provide OpenAI access to hundreds of thousands of Nvidia GB200 and GB300 GPU systems on EC2 UltraServers over seven years, with capacity targeted by the end of 2026 and options to expand thereafter.
- OpenAI gained flexibility to add AWS after a renewed Microsoft partnership loosened prior exclusivity and veto rights over cloud providers, according to recent partner disclosures.
- Microsoft’s latest quarterly report shows a $3.1 billion hit to its net income from its OpenAI stake, implying an OpenAI loss in the low double‑digit billions for the quarter and noting $11.6 billion of Microsoft’s $13 billion commitment has been funded.
- Reported multi‑year infrastructure commitments include roughly $300 billion with Oracle and about $250 billion with Microsoft, with several contracts staggered and contingent on available data‑center capacity.
- Microsoft CEO Satya Nadella and OpenAI CEO Sam Altman say power and ready data‑center shells, not GPU supply, are the main bottlenecks, with chips already on hand that cannot be deployed until sufficient electricity and facilities come online.