Nvidia's $100 Billion Bet on AI Falters as Rift With OpenAI Deepens
In September 2025, Nvidia made headlines by announcing a whopping $100 billion investment in OpenAI, touting the partnership as a long-term bet on the future of artificial intelligence. The deal included Nvidia building 10 gigawatts of AI data centers for OpenAI and investing billions more in the company over time. But now, it seems that the agreement may never come to fruition.
The trouble began when an Nvidia SEC filing from November revealed that the investment was still just a "letter of intent" with an opportunity to invest, rather than a finalized deal. Fast forward a few months, and reports have emerged that talks between the two companies have stalled, with Nvidia CEO Jensen Huang claiming that the initial $100 billion agreement was nonbinding and not finalized.
Huang's comments came after several anonymous sources reported that OpenAI was growing frustrated with Nvidia's chips, particularly when it comes to computing inference for some ChatGPT requests. The company is now reportedly exploring alternative chip providers, including startups like Cerebras and Groq, to take on 10% of its inference needs.
OpenAI executives have responded by praising Nvidia, with CEO Sam Altman stating that the chipmaker makes "the best AI chips in the world." However, it seems that Nvidia has been feeling the pressure itself, as increasing competition from Google has raised concerns about the company's ability to meet OpenAI's growing demands.
Google's custom AI chips, known as tensor processing units (TPUs), have been gaining traction in recent months, with some tasks deemed better than GPU chips dominated by Nvidia. The Wall Street Journal reports that Huang is worried about the competition both Google and Anthropic pose to OpenAI's market dominance, which could impact Nvidia's sales.
If investor fears are realized, and the deal doesn't go through as planned, the implications would be far-reaching. Both OpenAI and Nvidia sit at the center of a complex web of AI dealmaking, with numerous multibillion-dollar deals among a handful of companies. A collapse in one deal could have a ripple effect throughout the entire industry, potentially impacting the American economy.
The situation serves as a cautionary tale about the risks of over-reliance on single investments and the importance of diversification in the tech sector. With AI becoming increasingly important to businesses and governments alike, it's clear that this is an area where competition will only intensify.
In September 2025, Nvidia made headlines by announcing a whopping $100 billion investment in OpenAI, touting the partnership as a long-term bet on the future of artificial intelligence. The deal included Nvidia building 10 gigawatts of AI data centers for OpenAI and investing billions more in the company over time. But now, it seems that the agreement may never come to fruition.
The trouble began when an Nvidia SEC filing from November revealed that the investment was still just a "letter of intent" with an opportunity to invest, rather than a finalized deal. Fast forward a few months, and reports have emerged that talks between the two companies have stalled, with Nvidia CEO Jensen Huang claiming that the initial $100 billion agreement was nonbinding and not finalized.
Huang's comments came after several anonymous sources reported that OpenAI was growing frustrated with Nvidia's chips, particularly when it comes to computing inference for some ChatGPT requests. The company is now reportedly exploring alternative chip providers, including startups like Cerebras and Groq, to take on 10% of its inference needs.
OpenAI executives have responded by praising Nvidia, with CEO Sam Altman stating that the chipmaker makes "the best AI chips in the world." However, it seems that Nvidia has been feeling the pressure itself, as increasing competition from Google has raised concerns about the company's ability to meet OpenAI's growing demands.
Google's custom AI chips, known as tensor processing units (TPUs), have been gaining traction in recent months, with some tasks deemed better than GPU chips dominated by Nvidia. The Wall Street Journal reports that Huang is worried about the competition both Google and Anthropic pose to OpenAI's market dominance, which could impact Nvidia's sales.
If investor fears are realized, and the deal doesn't go through as planned, the implications would be far-reaching. Both OpenAI and Nvidia sit at the center of a complex web of AI dealmaking, with numerous multibillion-dollar deals among a handful of companies. A collapse in one deal could have a ripple effect throughout the entire industry, potentially impacting the American economy.
The situation serves as a cautionary tale about the risks of over-reliance on single investments and the importance of diversification in the tech sector. With AI becoming increasingly important to businesses and governments alike, it's clear that this is an area where competition will only intensify.