Core Insights - Nvidia (NVDA.US) has announced a new investment in AI startup Thinking Machines Lab, founded by former OpenAI executive Mira Murati, and will supply chips for training and running AI models [1][2] - The partnership includes a multi-year agreement where Thinking Machines Lab will utilize Nvidia's upcoming Vera Rubin AI acceleration chip, expected to provide at least 1 gigawatt of computing power, equivalent to the electricity consumption of approximately 750,000 households [1] - Nvidia's previous investment in Thinking Machines Lab occurred last year, but specific terms of the current investment have not been disclosed, described only as a "significant investment" [1] Investment Context - Nvidia, as the world's most valuable company, is actively pursuing multiple investment deals to drive AI implementation across various industries, contributing to what it terms a "new industrial revolution" [2] - Concerns have been raised regarding Nvidia's investment model, which involves taking equity stakes in its own customers [2] - Last November, Thinking Machines Lab sought new funding with a valuation target of $50 billion, which would represent a fourfold increase from its July valuation of $12 billion, following a $2 billion funding round [2] Company Background - Mira Murati, the founder of Thinking Machines Lab, previously served as the Chief Technology Officer at OpenAI and has recruited dozens of employees from OpenAI to her new venture [2] - Thinking Machines Lab launched its first product, Tinker, in October last year, aimed at helping users optimize large language models, which are foundational technologies for chatbots like ChatGPT [2] - Recently, Thinking Machines Lab has faced talent retention challenges, with several employees, including its CTO, returning to OpenAI [2]
英伟达(NVDA.US)再投AI初创Thinking Machines Lab 并供应Vera Rubin芯片