A Wall Avenue Journal report on Friday stated Nvidia insiders had expressed doubts in regards to the transaction and that Huang had privately criticized what he described as an absence of self-discipline in OpenAI’s enterprise strategy. The Journal additionally reported that Huang had expressed concern in regards to the competitors OpenAI faces from Google and Anthropic. Huang known as these claims “nonsense.”
Nvidia shares fell about 1.1 % on Monday following the stories. Sarah Kunst, managing director at Cleo Capital, told CNBC that the back-and-forth was uncommon. “One of many issues I did discover about Jensen Huang is that there wasn’t a robust ‘It is going to be $100 billion.’ It was, ‘It is going to be massive. It is going to be our largest funding ever.’ And so I do assume there are some query marks there.”
In September, Bryn Talkington, managing associate at Requisite Capital Administration, noted the round nature of such investments to CNBC. “Nvidia invests $100 billion in OpenAI, which then OpenAI turns again and provides it again to Nvidia,” Talkington stated. “I really feel like that is going to be very virtuous for Jensen.”
Tech critic Ed Zitron has been vital of Nvidia’s round investments for a while, which touch dozens of tech corporations, together with main gamers and startups. They’re additionally all Nvidia prospects.
“NVIDIA seeds corporations and provides them the assured contracts needed to lift debt to purchase GPUs from NVIDIA,” Zitron wrote on Bluesky final September, “although these corporations are horribly unprofitable and can ultimately die from an absence of any actual demand.”
Chips from different locations
Exterior of sourcing GPUs from Nvidia, OpenAI has reportedly mentioned working with startups Cerebras and Groq, each of which construct chips designed to cut back inference latency. However in December, Nvidia struck a $20 billion licensing cope with Groq, which Reuters sources say ended OpenAI’s talks with Groq. Nvidia hired Groq’s founder and CEO Jonathan Ross, together with different senior leaders, as a part of the association.
In January, OpenAI announced a $10 billion cope with Cerebras as an alternative, including 750 megawatts of computing capability for quicker inference by way of 2028. Sachin Katti, who joined OpenAI from Intel in November to guide compute infrastructure, stated the partnership provides “a devoted low-latency inference resolution” to OpenAI’s platform.
However OpenAI has clearly been hedging its bets. Past the Cerebras deal, the corporate struck an agreement with AMD in October for six gigawatts of GPUs and announced plans with Broadcom to develop a customized AI chip to wean itself off of Nvidia dependence. When these chips might be prepared, nevertheless, is at the moment unknown.
