英伟达(NVDA.US)黄仁勋：对OpenAI千亿美元投资“并非承诺” 将逐轮考虑任何融资

Group 1 - Nvidia's CEO Jensen Huang stated that the proposed investment of up to $100 billion in OpenAI is "never a commitment" and will be considered "step by step" [1] - The investment aims to support the construction of data centers with a power capacity of at least 10 gigawatts, equivalent to New York City's peak power demand, using Nvidia's advanced chips for AI model training and deployment [1] - Reports indicate that the investment plan has stalled due to internal doubts at Nvidia, with Huang privately emphasizing that the $100 billion agreement is non-binding and expressing concerns about OpenAI's business discipline and competitive landscape [1] Group 2 - Huang described the potential investment in OpenAI as "huge" and possibly "the largest investment in our history," but clarified that Nvidia's contribution to the current funding round would not approach $100 billion [2] - Nvidia and OpenAI are reportedly reassessing their partnership, considering equity investment in OpenAI's new funding round rather than a strategic partnership tied to orders [2] - Nvidia's recent investment plans include an additional $2 billion investment in cloud provider CoreWeave, another key customer [2] Group 3 - OpenAI is reportedly conducting a new funding round with a maximum target of $100 billion, which could raise its valuation to $830 billion [3] - Amazon is also in talks for a potential investment of up to $50 billion, which could make it the largest investor in this funding round [3] - SoftBank is negotiating for an additional investment of up to $30 billion, with previous investments exceeding $40 billion through its Vision Fund and joint investments [3] Group 4 - OpenAI's CFO Sara Frier announced that the company's recurring revenue is projected to exceed $20 billion in 2025, a significant increase from $6 billion in 2024, with computing capacity rising from 0.6 gigawatts in 2024 to 1.9 gigawatts in 2025 [4] - OpenAI is on track to release its first hardware device in the second half of the year, marking its entry into the AI hardware market [4]