Most Investors Overpay for AI Hype. Nvidia Is the Exception.

I get cautious when an artificial intelligence (AI) stock asks investors to pay today for profits that may not show up for years.

Nvidia (NASDAQ: NVDA) has been the dominant player in AI hardware for some time, but its stock still deserves a closer look even after gaining more than 360% over the past three years (as of Sept. 2, 2026). Is Nvidia stock now priced for distant hopes, or is it one of the tech sector's rare exceptions?

Image source: Getty Images.

Nvidia is already making huge profits from AI

Nvidia's results for its fiscal 2027 second quarter (which ended July 26) show how much the company is already benefiting from AI infrastructure spending. The company's revenue rose 106% year over year to $96.2 billion. The performance was mainly driven by the data center segment, where revenue increased 117% year over year to $89 billion. The company also generated $63.7 billion in generally accepted accounting principles (GAAP) operating income during the quarter. This pace of growth is particularly impressive considering Nvidia's massive size.

Nvidia is also converting a large portion of its revenues into cash. The company's free cash flow reached $69.9 billion, while the company returned roughly $46 billion to shareholders through share repurchases and dividends in the first half of its fiscal 2027. This shows that Nvidia has significant funds to invest in research and development, secure future supply, and return capital to shareholders.

However, strong cash generation does not automatically make Nvidia's valuation attractive. The stock trades at roughly 25 times analysts' consensus estimate for fiscal 2027 earnings of $9.26 per share. But the multiple falls to around 15 times based on the analysts' fiscal 2028 earnings estimate of $15.59 per share. Hence, I believe that the attractiveness of Nvidia's valuation depends significantly on whether the company can deliver the growth Wall Street currently expects.

Nvidia is expanding beyond GPUs

Nvidia's big upcoming growth driver is the Vera Rubin system, which is already in production. Management expects the new platform to account for about 20% of data center segment revenue in the third quarter. Nvidia has also received purchase orders for Vera Rubin from every major hyperscaler, AI cloud provider, and equipment manufacturer.

Nvidia estimates that its revenue opportunity for every gigawatt (a measure of power capacity) of AI infrastructure has increased from around $18 billion with its Hopper GPUs to $25 billion with its Blackwell systems and $40 billion with the Vera Rubin platform. Vera Rubin includes not only Rubin GPUs, but also Nvidia's Vera CPUs, NVLink high-speed interconnect technology, and InfiniBand or Ethernet networking. By selling more parts of the overall computing system, Nvidia can generate more revenue per data center.

Demand also remains strong. Nvidia gave preliminarily guidance saying it expects revenue to grow by about 70% in fiscal 2028, despite its production capacity being supply-constrained. Management said customer forecasts indicate that demand could support around 100% revenue growth for fiscal 2028.

Nvidia still faces significant risks

But I will also give due attention to the company's risks. Nvidia has warned that shortages of land, power, and other data center infrastructure could delay customer deployments of its hardware.

Nvidia's growth will also depend heavily on whether its customers can earn attractive returns from their massive AI infrastructure investments. There are encouraging signs. Microsoft's Azure revenue grew 43% year over year in the fourth quarter of its fiscal 2026 (which ended June 30), while Amazon Web Services' net sales increased 37% year over year in the second quarter. Alphabet's Google Cloud revenue also surged 82% year over year in the second quarter, driven by strong demand for AI infrastructure and AI solutions.

While these growth rates suggest that AI infrastructure spending is already helping cloud companies, they do not yet prove that the huge amounts being invested in new AI capacity will generate sufficiently high returns over time to justify the outlays.

Nvidia is also making increasingly large financial commitments to support future demand. As of the end of its latest quarter, the company had $366 billion in future spending commitments, including $279 billion related to supply and capacity. Nvidia has also guaranteed up to $108.5 billion, including a $105 billion guarantee related to a data center project for OpenAI. These commitments increase the company's downside exposure if AI infrastructure demand slows or if its customers face difficulties funding their projects.

Margins are another factor to watch. Nvidia reported a 75% gross margin in the second quarter. But management expects margins of 71% to 72% in the fourth quarter and 72% to 73% in fiscal 2028, due to pressure from higher memory prices.

Overall, my view is that Nvidia stock is not cheap based on a risk-reward analysis, but its share price is much better supported by the company's current earnings and cash flow than most AI stocks.

Manali Pradhan, CFA has no position in any of the stocks mentioned. The Motley Fool has positions in and recommends Alphabet, Amazon, Microsoft, and Nvidia. The Motley Fool has a disclosure policy.