Chip allocation has become Nvidia’s currency of choice in frontier AI. Its latest deployment: a multi-billion-dollar commitment to Safe Superintelligence that bundles hardware access with an equity stake in the research lab led by former OpenAI chief scientist Ilya Sutskever.
Bloomberg reported the figure at $5 billion. Under the terms, SSI gets allocation on Nvidia’s newest accelerator lineup and expects to substantially scale its model training and experimentation workload. Nvidia had participated in SSI’s earlier funding rounds before moving to deepen the relationship.
Sutskever launched the lab in 2024 following his departure from OpenAI, where his team worked on the GPT architecture and reasoning models. The startup has secured roughly $3 billion in total funding, with backers spanning Andreessen Horowitz, Alphabet, and Sequoia Capital. A February 2025 round valued the company at $32 billion.
“Sutskever’s work on AlexNet helped prove that GPU scaling and deep neural networks could work together,” Nvidia CEO Jensen Huang said. Sutskever described the increased compute capacity as allowing his team to pursue research that has reached the point where larger infrastructure is the limiting factor.
