Thinking Machines Lab inks massive compute deal with Nvidia
Summary
Mira Murati's two-year-old AI research lab, Thinking Machines Lab, has formed a multi-year strategic partnership with Nvidia. The deal, announced Tuesday, involves Thinking Machines Lab deploying at least one gigawatt of Nvidia's Vera Rubin systems starting in 2027. Nvidia, an existing investor, is also making an additional strategic investment in the lab, which has raised over $2 billion since its February 2025 founding and is valued at more than $12 billion. Thinking Machines Lab, known for its work on reproducible AI models, released its first product, an API called Tinker, in October. The partnership also includes a commitment to develop training and serving systems optimized for Nvidia architecture.
Key takeaway
For AI/ML Directors evaluating future compute infrastructure, this partnership underscores the critical need for massive, dedicated resources. Your strategy should account for multi-year commitments with leading hardware providers like Nvidia to secure the necessary scale for advanced AI model development. Expect significant capital expenditure in AI infrastructure, potentially reaching trillions industry-wide, necessitating early planning and strategic vendor relationships.
Key insights
AI research labs are securing massive compute deals to build reproducible AI models.
Principles
- AI compute demand drives multi-billion dollar infrastructure deals.
- Strategic partnerships with chipmakers are crucial for AI development.
In practice
- Consider Nvidia Vera Rubin systems for large-scale AI infrastructure.
- Explore API-based AI products like Tinker for model integration.
Topics
- AI Research Labs
- NVIDIA Vera Rubin
- AI Compute Infrastructure
- Strategic Investments
- AI Model Reproducibility
Best for: VP of Engineering/Data, Director of AI/ML, Entrepreneur, Investor, CTO, AI Product Manager
Related on AIssential
Editorial summary, takeaway, and curation by AIssential. Original article published by TechCrunch.