I'm LongbridgeAI, I can summarize articles.Nvidia's stock faces bearish skepticism despite record profits and a $5 trillion market cap. Critics cite industry cyclicality, competitor chip development, and rapid depreciation risks. However, data shows H100 GPU rental prices rose 22% monthly, defying depreciation expectations. This durability supports Nvidia's multi-tiered pricing model and leadership in AI chips, suggesting the company is undervalued by Wall Street.
Nvidia's (NVDA -2.01%) accomplishments speak for themselves, but despite the stock having grown to a market cap of more than $5 trillion and it becoming the most profitable company in the world, there's still a lot of skepticism facing it.
Even as revenue nearly doubled in its most recent quarter, the stock trades at a price-to-earnings ratio of just 29, roughly in line with the S&P 500, indicating that investors expect its long-term earnings growth to generally resemble the broad-market index, even though it more than doubled net income in its latest quarter and expects strong growth to continue at least through 2027.
There are a number of reasons why Nvidia doesn't get the premium you might expect for a company growing this fast. First, the semiconductor industry is historically cyclical, and investors are expecting the momentum in the AI boom to eventually fade. At that point, Nvidia's revenue and earnings growth could turn negative as it has in past cycles. Second, competitors, including Nvidia's hyperscaler customers, are building their own chips to substitute for Nvidia components. While they're unlikely to replace them entirely, it could signal that Nvidia's competitive advantage is likely to erode over time. Finally, some investors think that depreciation in Nvidia's chips is an outsize risk facing the company and the broader AI boom. If its chips lose their value quickly, that is likely to hurt their selling price and the broader sustainability of AI, as eventually, Nvidia's customers will need to sell enough services to pay for its chips.
This theory, advanced by Michael Burry of "The Big Short" fame, has been used to criticize hyperscalers and neocloud companies, as well as Nvidia.
However, there's some evidence that Nvidia chips are retaining their value much better than the skeptics would expect.
Image source: Nvidia.
Jensen Huang weighs in
The comments and chart below, taken from X, show the market average for live cloud GPU rental costs based on the Ornn H100 SXM Index.
NVIDIA compute is fungible, durable and highly rentable. It is a productive, revenue-generating asset. https://t.co/cvmjaNoiK8
— Jensen Huang (@JensenHuang) September 8, 2026
The H100 is a three-year-old training chip. Its rental price is up 22 percent on the month, to $3.28 an hour.
— Ornn (@OrnnExchange) September 7, 2026
Every depreciation schedule assumes a chip this old only loses value. The market is paying up for it instead. pic.twitter.com/TNSqgys3vx
Image source: Ornn. Via X.
As you can see, rental prices per hour for an H100 GPU, which were first launched nearly four years ago, are up 22% over the last month, even as Nvidia is now launching the new Rubin platform. the Rubin GPU, or R100, will make the H100 two generations old.
Rental and purchase prices for the H100 have indeed come down substantially from their peak in 2023, when generative AI was just starting to take off, but it's noteworthy that they were still able to increase their value at this point, even as newer options on the market emerge. The R100 will have several times as much memory as the H100 and use a superior, updated architecture.
Even the A100, the generation before the H100, remains in high demand, as some customers prefer the cheaper per-hour rental costs of the A100. Rather than disrupting itself with newer chips, Nvidia seems to be benefiting from a multi-tiered pricing model in which premium customers can pay premium prices for the newest chips and budget customers can pay lower prices for older chips.
NASDAQ: NVDA
Key Data Points
What it means for Nvidia
The durability of H100 pricing offers yet another reason to be skeptical of the bearish thesis on the stock and to bet on the company's continued leadership in AI chips.
It's a reminder also that Nvidia, if anything, has been underestimated by Wall Street over the last few years, as the analyst consensus has been woefully short. Similarly, bearish predictions for the stock have fallen flat. Given the fact that the business just doubled in size in the most recent quarter, prices for chips that are now two generations old are holding up, and the stock is trading on par with the S&P 500, Nvidia continues to look like an excellent buy.