PANews reported on January 31 that, according to Global Times, sources revealed that Nvidia originally planned to invest up to $100 billion (approximately RMB 696.456 billion) in OpenAI to train and run the latest artificial intelligence models. However, due to internal doubts, this highly anticipated deal has stalled.

Last September, Nvidia and OpenAI announced their collaboration at Nvidia's headquarters in Santa Clara, California. At the time, they signed a memorandum of understanding, with Nvidia committing to building at least 10GW of computing power for OpenAI and investing up to $100 billion to help cover related costs, while OpenAI agreed to lease Nvidia chips. OpenAI anticipated a swift resolution at the time, but discussions remain in the early stages. Currently, both parties are reassessing the prospects of their collaboration, and the latest proposal may involve a dollar equity investment in OpenAI's current funding round.

Sources familiar with the matter further revealed that Nvidia CEO Jensen Huang has privately emphasized in recent months that the initial $100 billion agreement is not binding and is far from finalized. He has also criticized OpenAI for its lack of discipline in commercialization and expressed concern about the intense competition it faces from rivals such as Anthropic.

In response, an OpenAI spokesperson stated, "The details of the collaboration are still underway. NVIDIA technology has supported our breakthroughs from the beginning, drives the current system, and will remain core as we scale up in the future."