A Wall Street Journal report on Friday said Nvidia insiders had expressed skepticism about the transaction and Huang had privately criticized what he called a lack of discipline in OpenAI’s business approach. The Journal also reported that Huang had expressed concerns about competition OpenAI faces from Google and Anthropic. Huang called those claims “nonsense”.
Nvidia shares fell about 1.1 percent on Monday following the report. Sarah Kunst, managing director of Clio Capital, told CNBC that the back-and-forth was unusual. “One of the things I noticed about Jensen Huang is that there was no strong ‘it will be $100 billion.’ It was, ‘This will be big.’ This will be our biggest investment ever. And so I think there are some question marks there.
In September, Brian Talkington, managing partner of Requisite Capital Management, noted the circular nature of such investments to CNBC. “Nvidia invests $100 billion in OpenAI, which OpenAI takes back and gives back to Nvidia,” Talkington said. “I think it would be very rewarding for Jensen.”
Tech critic Ed Zitron has been criticizing Nvidia’s circular investment for some time, which touches dozens of tech companies, including major players and startups. They are also all Nvidia customers.
“NVIDIA seeds companies and gives them the guaranteed contracts needed to raise the debt to buy GPUs from Nvidia,” Zitron wrote on Bluesky last September, “even though these companies are dreadfully unprofitable and will eventually die from lack of any real demand.”
Chips from other places
In addition to sourcing GPUs from Nvidia, OpenAI has reportedly discussed working with startups Cerebras and Grok, which build chips designed to reduce inference latency. But in December, Nvidia struck a $20 billion licensing deal with Groke, which Reuters sources say ended OpenAI’s talks with Groke. Nvidia hired Groke founder and CEO Jonathan Ross along with other senior leaders as part of the arrangement.
In January, OpenAI instead announced a $10 billion deal with Cerebras, adding 750 megawatts of computing capacity for faster inference by 2028. Sachin Katti, who joined OpenAI from Intel in November to lead compute infrastructure, said the partnership adds “a dedicated low-latency inference solution” to OpenAI’s platform.
But OpenAI is clearly hedging its bets. Beyond the Cerebras deal, the company in October signed a deal with AMD for six gigawatt GPUs and announced plans with Broadcom to develop a custom AI chip to distance itself from Nvidia dependence. However, when those chips will be ready is currently unknown.
<a href