Arguably, Nvidia is the global economy’s most important company. Its accelerated computing graphics processing unit chips and compute unified device architecture software are at the center of the artificial intelligence revolution. Nvidia supplies about 94% of the semiconductors that are embedded in the AI servers of the world’s data centers.
Nvidia has achieved an effective monopoly position for accelerated computing GPUs with superior semiconductor design and complementary software products. And Nvidia is highly profitable. The company enjoys high profit margins, which generate significant free cash flow, which Nvidia then reinvests in its business, particularly in research and development. Put simply, Nvidia is ahead of the competition and running fast to stay ahead.
Recommended Stories
- The hidden flow of federal guidance to Wisconsin state agencies
- NAS's fast-track climate review is a political end-run
- The GAO’s deep state double standard
On Monday, Jensen Huang, a cofounder and CEO of Nvidia, along with Sam Altman, one of the creators of OpenAI, which developed the popular AI platform ChatGPT, announced that Nvidia would invest up to $100 billion in OpenAI. The money will be used to build new AI data centers and other AI infrastructure. The transaction demonstrates that the AI revolution is real and accelerating. Financial markets cheered the deal. The Nvidia money is intended to build data centers with an electricity capacity of 10 gigawatts or more. To illustrate the scale of the investment, 10 gigawatts is the amount of electricity consumed each day by the residents and businesses of New York City during working hours.
The first $10 billion will be disbursed when the deal is signed, and Nvidia will receive equity in OpenAI in return. In addition, OpenAI will use some of the money from Nvidia to buy Nvidia’s GPU chips. The deal is a form of vendor financing. Nvidia gives cash to OpenAI, which OpenAI then uses to buy GPUs. This part of the transaction raises questions, but it is accepted practice in the AI business.
For example, Nvidia has invested in the AI company CoreWeave, which has used the Nvidia money to buy GPUs from Nvidia. CoreWeave leases GPUs to customers who run AI services. CoreWeave is a very successful company. In addition to the issue of vendor financing, critics of the transaction question whether AI companies will ever achieve an acceptable return on capital.
Today, capital expenditure dramatically exceeds profit in the AI business. It is fair to say that Nvidia is effectively capturing almost all of the profit. However, informed AI experts remain confident that acceptable returns will be generated and that the AI revolution will deliver the promise of a surge in productivity for the U.S. economy. Accelerated productivity growth for the U.S. economy is worth trillions of dollars.
Importantly, AI acolytes note that demand for AI inference in 2026 will grow 10 times from current year levels. Moreover, global spending on AI will approach $2 trillion in 2026. It is a safe bet that the world’s leading AI companies are confident in future profits. The more difficult question is whether the U.S. energy industry can provide the electricity the investment demands. As noted, the data centers that OpenAI will build will consume about 10 gigawatts of electricity. America’s largest nuclear power plant, Plant Vogtle in Georgia, produces about 4.5 gigawatts of electricity.
JUDGE SHUTS DOWN WOULD-BE TRUMP ASSASSIN RYAN ROUTH’S RAMBLING OPENING STATEMENT
The OpenAI deal requires more than two Vogtles — a tall order. Still, there is hope; shuttered nuclear power plants are being reopened. Other power plants, including coal plants, will stay in service longer. And the natural gas sector will build new plants. The combination of gas, nuclear, coal, and renewables will likely provide the required electricity. Very importantly, President Donald Trump is fully committed to AI and accelerating the increase in electricity supply in the United States.
The Nvidia/OpenAI transaction is important. It will help ensure U.S. AI superiority over China. Equally important, the deal lays the foundation for a sustainable surge in productivity growth in the U.S. economy. Faster productivity growth raises the standard of living for all Americans and provides the resources necessary to defend the country against adversaries.
The writer owns shares in Nvidia.
James Rogan is a former U.S. foreign service officer who has worked in finance and law for 30 years. He writes a daily note on the markets, politics, and society. He can be followed on X @JamesY54939 and reached at [email protected].