- NVIDIA has struck a strategic investment and compute deal with Sutskever’s Safe Superintelligence.
- The deal gives SSI early access to NVIDIA’s Vera Rubin chips, the same structure critics call circular financing.
- It follows a Bloomberg report on a possible $600B NVIDIA-OpenAI financing package, reviving scrutiny of NVIDIA’s playbook.
Safe Superintelligence, also known as SSI, the AI lab founded by OpenAI co-founder Ilya Sutskever and Daniel Levy in 2024, has signed a strategic partnership with NVIDIA that pairs an equity investment with long-term access to NVIDIA’s next-generation Vera Rubin computing platform.
Neither company disclosed the size of the investment, but people familiar with the matter told Bloomberg it is worth around $5B, which would make it one of NVIDIA’s largest single bets on a frontier AI lab.
SSI said the deal will increase its compute capacity by an order of magnitude, and that NVIDIA sought the partnership after gaining rare access to SSI’s closely guarded research programme.
“Ilya has pioneered fundamental breakthroughs at the foundation of modern AI. We’re excited to see what new breakthroughs SSI will discover powered by our Vera Rubin platform,” said Jensen Huang, founder and chief executive of NVIDIA.
A familiar shape
Strip away the framing and the deal is familiar. NVIDIA invests in an AI lab, the lab gets priority access to NVIDIA’s newest chips, and NVIDIA’s own hardware roadmap benefits from the research relationship in return.
That structure has followed NVIDIA all year: the chipmaker has already committed more than $40B to AI equity investments in the first four months of 2026 alone, per Tech Funding News’ own tracking, on top of roughly $53B across 170 deals since 2020, according to PitchBook data.
That tally includes a $30B stake in OpenAI and a $10B investment in Anthropic. CoreWeave remains the clearest example of the pattern in practice: NVIDIA holds equity in the company and has a $6.3B agreement to buy back its unused compute capacity through 2032.
A test case with no revenue
SSI is a useful stress test for that argument precisely because it has no product and no revenue to point to.
The lab has previously raised $3B across two rounds, most recently $2B at a $32B valuation with backing from Alphabet, NVIDIA, Andreessen Horowitz, Lightspeed Venture Partners, Greenoaks, and DST Global, as TFN reported at the time.
SSI describes itself as the world’s first straight-shot lab, built around a single goal of safe superintelligence rather than a near-term commercial offering. That makes the compute-for-equity exchange harder to justify as ordinary vendor financing and easier to read as NVIDIA extends its balance sheet to the frontier of AI research on the strength of Sutskever’s reputation alone.
The $600B backdrop
The timing sharpens the question. Bloomberg reported on July 27, 2026 that NVIDIA is negotiating a financing package with OpenAI that could exceed $600B, and credit default swap premiums on major AI infrastructure names have widened as investors price in the risk that some of this spending is financially engineered rather than organic.
NVIDIA generated close to $200B in annual revenue last year and holds a market capitalisation above $5T, giving it far more room to absorb a write-down than any single portfolio company could survive a funding gap.
Whether that changes if the AI investment cycle cools is the open question hanging over every deal NVIDIA signs. SSI, with a research mission and no commercial product to fall back on, is now part of that story too.
