Thinking Machines Lab Secures Gigawatt-Scale NVIDIA Chip Deal

Thinking Machines Lab Secures Gigawatt-Scale NVIDIA Chip Deal

Thinking Machines Lab, the AI startup founded by former OpenAI CTO Mira Murati, announced a landmark multi-year partnership with NVIDIA on March 10, 2026. Under the agreement, the company will deploy at least one gigawatt of servers powered by NVIDIA's forthcoming Vera Rubin chips — roughly the energy equivalent of powering 750,000 homes.

NVIDIA also made a significant undisclosed investment in Thinking Machines Lab as part of the deal. The two companies declined to share additional financial details.

What Thinking Machines Is Building

Thinking Machines Lab was founded with a mission to build AI systems that are "more widely understood, customizable, and generally capable." The company is targeting enterprise AI deployments and has been positioning itself as a serious competitor to frontier labs like OpenAI and Anthropic.

Compute as Strategy

The gigawatt-scale commitment stands out even against the backdrop of record AI infrastructure spending. Securing a dedicated pipeline of Vera Rubin chips — NVIDIA's next-generation data center GPU — gives Thinking Machines a substantial runway to train frontier models and serve enterprise customers. The deal signals NVIDIA's confidence in Murati's team and its willingness to back emerging competitors alongside established players.