Nvidia has struck a multiyear strategic partnership with Thinking Machines Lab, the AI company founded by Mira Murati, the former chief technology officer of OpenAI, to deploy a substantial amount of next-generation computing power for training advanced AI models.
The scale of the deal is measured in gigawatts, a unit of electrical power that has become a shorthand in the AI industry for the enormous energy demands of running large clusters of AI chips: one gigawatt is roughly equivalent to the output of a large power station and enough to power hundreds of thousands of homes.
Under the agreement, Thinking Machines will deploy at least one gigawatt of Nvidia's forthcoming Vera Rubin systems, the company's next generation of AI chips, with targeted deployment beginning early next year.
The partnership also includes work to design training and serving systems specifically tuned for Nvidia hardware, and Nvidia has made a financial investment in Thinking Machines Lab to support its long-term growth.
The collaboration is aimed at making advanced AI more accessible to enterprises, research institutions and the scientific community, with a focus on customisable models that users can adapt to their own needs.
Jensen Huang, founder and chief executive of Nvidia, said: "AI is the most powerful knowledge discovery instrument in human history."
Related reading
- NVIDIA invests $4bn in Lumentum and Coherent to scale optics manufacturing
- NVIDIA posts record full year revenue, Jensen Huang says "agentic AI inflection point has arrived"
- Nvidia's Jetson chip is bringing AI out of the cloud and into the machines around us
Murati said: "This partnership accelerates our capacity to build AI that people can shape and make their own, as it shapes human potential in turn."
Thinking Machines Lab is one of several well-funded AI startups competing to train frontier models, the most capable and resource-intensive class of AI systems, at a time when demand for computing infrastructure is outpacing supply.
The recap
- Multiyear strategic partnership between NVIDIA and Thinking Machines Lab
- Agreement covers at least 1 gigawatt of Vera Rubin systems
- Deployment on Vera Rubin targeted for early next year
