Key Highlights
- Nvidia has provided an undisclosed financial investment in Thinking Machines Lab, the artificial intelligence company led by ex-OpenAI CTO Mira Murati
- A strategic multi-year collaboration between the two firms was revealed on Tuesday
- Thinking Machines plans to utilize no less than one gigawatt worth of Nvidia’s Vera Rubin infrastructure
- Rollout is scheduled for the beginning of next year
- Following a $2B capital raise in July 2025, Thinking Machines achieved a $10B valuation
Nvidia has taken a financial stake in Thinking Machines Lab while simultaneously establishing a multi-year strategic collaboration with the artificial intelligence venture. Both companies made the announcement public on Tuesday morning.
Neither organization revealed the precise amount of Nvidia’s financial commitment.
Under the terms of the partnership, Thinking Machines intends to implement a minimum of one gigawatt worth of Nvidia’s forthcoming Vera Rubin infrastructure. This hardware will power the company’s advanced model training operations and platform services.
According to Thinking Machines, the Vera Rubin infrastructure rollout is slated to commence in early 2026.
Thinking Machines Lab was established by Mira Murati, the former Chief Technology Officer at OpenAI. The company’s mission centers on developing scalable, customizable artificial intelligence solutions tailored for corporate clients, academic researchers, and the broader scientific sector.
In a prepared statement, Jensen Huang, founder and CEO of Nvidia, remarked: “Thinking Machines has assembled an exceptional team dedicated to pushing the boundaries of AI. We are excited to collaborate with Thinking Machines as they pursue their ambitious vision for AI’s future.”
Murati offered her perspective: “NVIDIA’s infrastructure serves as the bedrock upon which our entire industry stands. This collaboration dramatically expands our ability to create AI systems that individuals can customize and truly own.”
Inside the $10 Billion AI Startup
This marks not Nvidia’s initial engagement with Thinking Machines. Previously, in July 2025, the startup secured $2B in financing through a funding round that established its valuation at $10B.
Andreessen Horowitz (a16z) spearheaded that investment round, with additional capital provided by Nvidia, AMD, ServiceNow, and Cisco Systems.
Tuesday’s disclosure represents a formalization and expansion of that relationship, elevating Nvidia’s role from mere investor to committed long-term strategic collaborator.
The alliance also encompasses collaborative efforts to engineer training and deployment systems optimized explicitly for Nvidia’s hardware architecture.
Vera Rubin Takes Center Stage
The Vera Rubin platform represents Nvidia’s upcoming generation of GPU infrastructure, and this agreement positions a gigawatt-scale implementation as its centerpiece.
A gigawatt-level power allocation represents a massive infrastructure investment — reflective of the extraordinary scale required for cutting-edge AI model development today.
According to Thinking Machines, the Vera Rubin hardware will serve as the backbone for its efforts to create AI systems that end-users can directly customize and engage with.
The collaboration is positioned as an initiative to broaden access to advanced AI capabilities and open-source models throughout enterprise and academic environments.
Nvidia maintains investment positions across numerous AI startups and has been strengthening strategic relationships industry-wide through semiconductor supply contracts and ownership positions.
The Thinking Machines agreement represents another significant addition to that portfolio, distinguished by its gigawatt-scale hardware deployment commitment.





