NEW YORK, September 13 — Nvidia’s own pitch to investors rests on a single, striking ratio: every dollar the company puts into research and development returns a hundred. That claim, which the American semiconductor giant has used to justify its heavy spending on artificial intelligence chips and software, is now being tested by a market that has turned skittish on the stock.
Shares of the company founded in 1993 by Jensen Huang, Chris Malachowsky and Curtis Priem have declined in recent trading sessions, even as Nvidia remains the dominant supplier of graphics processing units used to train and run large AI models. The pullback comes amid broader concerns about whether the enormous capital expenditure on AI infrastructure by cloud providers and technology giants can be sustained.
A trillion-dollar run, then a wobble
Nvidia went public in January 1999 at a split-adjusted price of roughly $0.04 per share. Its market capitalisation crossed $1 trillion for the first time in May 2023, and later surpassed $3 trillion in June 2024, briefly making it the most valuable public company in the world. The stock rose more than 200% in 2023 and continued to climb in early 2024 before pulling back.
The financials behind that run were extraordinary. Data centre revenue, which includes Nvidia’s AI chips, reached $18.4 billion in the quarter ending January 2024, up more than 400% from the same period a year earlier.
For the quarter ending April 2024, data centre revenue was $22.6 billion, up 427% year-on-year, while total revenue for that quarter came in at $26.0 billion, a 262% increase from the prior year. In its fiscal year 2024, which ended January 28, 2024, Nvidia reported total revenue of $60.9 billion, up 126% from the prior fiscal year, with net income of $29.8 billion. Gross margin in the January 2024 quarter was 76.7%.
The China question and the competition
Investors are weighing several risks. Competition is one, from Advanced Micro Devices and Intel, and from custom chips designed by major customers including Google, Amazon and Microsoft. Export restrictions are another.
The US government imposed controls on advanced semiconductors sold to China in October 2022 and tightened those rules in October 2023. Nvidia designed China-specific chips, such as the A800 and H800, to comply with the earlier rules, but the October 2023 update restricted those as well.
The company then developed the H20, a lower-performance chip for the Chinese market. China historically accounted for a significant portion of Nvidia’s data centre revenue. Nvidia’s stock is also sensitive to interest rate expectations and broader tech sector valuations.
The recent decline may reflect profit-taking after the massive run-up. Its forward price-to-earnings ratio has been above 30, high relative to historical averages for semiconductor companies.
The $1-to-$100 claim is not independently audited and is based on internal metrics the company has not fully detailed. Investors are left to judge whether it is a reliable indicator of future returns or a marketing statement. Huang has repeatedly said Nvidia’s investment in AI research pays off exponentially because it enables new markets and products.
Nvidia has not issued a new statement explaining the recent stock decline. Its next earnings report is expected in August 2024. As of the time of writing, the stock price and market capitalisation are changing, and no current figure should be treated as fixed.
