The whopping $100 billion investment that Nvidia said it will make in OpenAI is probably not happening, according to Nvidia CEO Jensen Huang.
Nvidia’s latest finalized investment of $30 billion, which was part of a $110 billion OpenAI funding round last month that also involved a $50 billion investment from Amazon and $30 billion from SoftBank, will likely be the last. That’s due to OpenAI’s plans to go public later this year, Huang said at the Morgan Stanley Technology, Media and Telecom conference on Wednesday, per Reuters.
OpenAI and Nvidia unveiled the $100 billion deal in September. Nvidia promised to build 10 gigawatts of AI data centers for OpenAI, investing a total of $100 billion in the company in 10 installments, with each gigawatt that comes online. The first gigawatt would come online in the second half of 2026. In turn, OpenAI would reinvest some of those dollars to lease Nvidia’s chips.
At the time, the announcement sparked an industry-wide discussion on the dangers of circular dealmaking and fears of a growing AI bubble. As a handful of top AI companies signed multibillion-dollar deals over and over again with each other, some experts raised worries that the financial dependencies could be a sign of instability. If one deal goes down or demand does not pan out as expected, the domino effect could take the whole system down.
Then, the agreement just failed to materialize, and rumors of drama between the two companies began to swirl. In a November earnings report, Nvidia only characterized the deal as “a letter of intent with an opportunity to invest.” In a January report, the Wall Street Journal claimed that talks had failed to move past early stages. The report also claimed that Huang had been privately criticizing a lack of discipline in OpenAI’s business approach and was worried it could fall behind competitors like Google and Anthropic, ultimately impacting Nvidia’s sales. The WSJ article came on the heels of a Reuters report, this time claiming that OpenAI was not pleased with the inference capabilities of Nvidia chips and had, in fact, been blaming some of the weaknesses of its AI coding assistant Codex on Nvidia hardware.
Both parties denied rumors of a fallout following the reports, and Huang even claimed that “everything’s on track.” But a few days later, The Financial Times reported that the two giants would, in fact, abandon the deal, citing anonymous sources.
In Nvidia’s earnings report last week, the company refused to give any assurance that “a transaction will be completed,” while Huang had said that they were continuing “to work with OpenAI toward a partnership agreement, and believe we are close.”
Read the full article here
