Safe Superintelligence Inc. (SSI), the AI research lab founded by Ilya Sutskever, has entered into a long-term strategic partnership with Nvidia. The agreement includes a substantial investment from Nvidia and grants SSI significant access to the chipmaker's next-generation Vera Rubin computing platform. This collaboration is intended to accelerate SSI's research and increase its compute capacity by an order of magnitude.
Nvidia will also collaborate with SSI on the technical development of current and future compute platforms, drawing on SSI's insights into artificial intelligence. Nvidia stated that it entered this partnership after gaining rare insight into SSI's closely guarded research, which has focused for the past two years on developing a powerful and robustly aligned artificial intelligence. Jensen Huang, founder and CEO of Nvidia, noted Sutskever's foundational contributions to modern AI, beginning with AlexNet, and expressed excitement about SSI's potential discoveries powered by the Vera Rubin platform.
Ilya Sutskever, co-founder and CEO of SSI, stated that the company has research worthy of scaling and that access to Nvidia's computing resources will enable this expansion. He expressed confidence that the partnership would advance SSI's work to the next level.
Founded in 2024, SSI's stated mission is to develop a safe superintelligence, with a singular focus on this goal and product. The company operates with a small team and is split between Palo Alto, California, and Tel Aviv, Israel. Sutskever, formerly chief scientist at OpenAI, co-founded SSI with Daniel Levy and Daniel Gross. Prior to this partnership, SSI had already secured significant investor backing, including Andreessen Horowitz, DST Global, Greenoaks, and Sequoia Capital, and reportedly reached a valuation of over $30 billion.
The partnership with Nvidia addresses a primary constraint for frontier AI companies: access to vast computing power. SSI has focused its research over the past two years on creating "safe superintelligence," diverging from the industry's prevalent emphasis on simply building larger models. The company's approach views safety and capabilities as intertwined technical problems to be solved through engineering and scientific breakthroughs.
Nvidia has established similar partnerships with other leading AI organizations, including OpenAI, while competitors like AMD have formed alliances with companies such as Anthropic.