In late 2022, Wall Street noticed that the first iteration of OpenAI's AI chatbot, ChatGPT, was trained and powered by thousands of Nvidia's (NVDA 1.53%) data center graphics processing units (GPUs). And a spark was lit that quickly made the cutting-edge chipmaker the largest company in the world, with a market cap of $4.55 trillion.
The good news is that companies continue to clamor for Nvidia's hardware. But how sustainable is the demand propping up Nvidia's huge valuation? Let's dig deeper into what could come next as the chipmaker seeks to maintain its dominance over the next decade and beyond.
Image source: Getty Images.
How much is big tech spending on chips?
As of early 2026, the AI race shows no signs of slowing down. This month, cloud computing giant Amazon announced plans to increase its full-year capital expenditures by 50% to $200 billion, with much of that going to data center spending. Alphabet has similar goals, with $175 billion to $185 billion earmarked this year. CNBC estimates that total AI spending could hit an eye-popping $700 billion this year alone.
The big-spending hyperscalers can afford to pour so much money into AI because of their diversified and highly profitable businesses. But that doesn't necessarily mean it's a good idea. Capital expenditures come with an opportunity cost because they represent cash flow that could have been used for other things or returned to shareholders through buybacks or dividends. Meanwhile, the value of the AI spending looks uncertain.
While AI technology continues to improve at a rapid rate, it still consistently underperforms humans on basic labor tasks and generates huge losses for the consumer-facing companies like OpenAI and Anthropic that rent computing power from Nvidia's clients. Investors are also starting to balk at the data center spending, with Amazon's share price falling almost 20% in the week following its capex announcement.
Ultimately, public companies exist to serve their shareholders. And if shareholders start consistently punishing Nvidia's clients for their huge AI data center spending, it could force them to cut back or turn to cheaper, homegrown solutions like custom chips.
Can new opportunities save the day?
Nvidia earned around 90% of its revenue from its data center segment, which means the company is extremely overexposed to a potential slowdown in this market. However, over the coming years or decades, management aims to pioneer new markets for its hardware.
Nvidia has already begun aggressively investing in quantum computing. And in October, it announced NVQlink, an architecture designed to couple its GPUs with quantum computers. It is too early to know how things will play out, but analysts expect quantum computing to be commercially viable by the end of the decade. And Nvidia could be positioned to dominate the opportunity because of its expertise in chip design and partnerships with advanced manufacturing partners like Taiwan Semiconductor Manufacturing.
Nvidia also has an opportunity in automotive hardware and robotics, where it provides chips to assist with self-driving cars. This segment represented a relatively modest $592 million in third-quarter revenue, but it grew by 32% from the prior-year period. And there is plenty of room for continued acceleration as self-driving becomes more mainstream.
NASDAQ: NVDA
Key Data Points
How will Nvidia's stock perform over the next 10 years?
Over the next decade, new opportunities like quantum computing and automotive could help make up for a possible slowdown in AI data center demand. But with around 90% of Nvidia's sales coming from its data center segment, new revenue streams could struggle to move the needle.
That said, the company's current valuation seems to price in these long-term challenges. With a forward price-to-earnings (P/E) multiple of just 24, the stock is very cheap for a company that grew earnings per share by 67% year over year to $1.30 in its most recent quarter. The low valuation gives the stock a margin of safety as it navigates these uncertain times. Nvidia looks like a hold until more information becomes available.
