Nvidia puts about $5B into SSI and ties research ambition to Vera Rubin computeFUNDING
Nvidia has agreed to invest about $5 billion in Safe Superintelligence, the startup founded by former OpenAI chief scientist Ilya Sutskever. SSI will receive access to Nvidia’s next-generation Vera Rubin platform and says its available compute could increase tenfold over the next 12 months. The deal follows roughly $3 billion in prior funding and a reported $32 billion valuation, despite the company having no public product or research output. The larger market issue is circular financing: Nvidia is also reportedly discussing up to $250 billion in guarantees for OpenAI data-center leasing, with total AI infrastructure-related commitments potentially exceeding $750 billion.[1]

The AI story today is capital and control converging around frontier capability. Nvidia’s SSI investment and the U.S. frontier-model review framework point to the same structural shift: model progress is now governed by access to compute, safety review, and strategic infrastructure financing.
Web3 shows a risk-off market with active infrastructure work underneath. Prices and leverage reset sharply before the Fed meeting, but Lido’s validator migration, x402 security research, Ondo Network, Coinbase Launches, and prediction-market litigation show that protocol-level and regulatory work continues through volatility.