Jensen Huang is betting that global data center capital expenditures will hit $4 trillion annually by 2030, a fivefold jump from an estimated $800 billion next year, and that trajectory could push Nvidia's market value to $20 trillion from $5 trillion today.
The math is straightforward, if you accept the premise. The four largest AI hyperscalers have signaled roughly $650 billion in data center spending for 2026, a figure that excludes neocloud operators, Chinese buyers, and model builders such as Anthropic and OpenAI. Adding those pieces gets you to the $800 billion baseline. Huang's $4 trillion target implies that spend quintuples in four years. If Nvidia merely holds its current share of that expanding pie, revenue and profits scale accordingly. The company only needs to quadruple its market cap to reach $20 trillion, so it could even cede some ground and still hit the number.
The top line is still accelerating. Revenue rose 85 percent year over year in the latest quarter, and analysts are modeling nearly 100 percent growth for the next one. All of that has happened without meaningful chip sales to China, where U.S. export controls have kept the highest-end processors out of reach.
A recent comment from a U.S. official that "very few" H200 chips have shipped to China sounds like a dead end, but it can be read the other way: evidence that the door is cracking open. The Trump administration eased some restrictions, and if Beijing's retaliatory roadblocks are also falling, a revenue stream currently excluded from guidance could materialize. That is the upside case. The downside is simpler, capex forecasts have a habit of missing by wide margins, and a $4 trillion figure assumes an investment cycle with few historical parallels.
The market is pricing a lot of certainty into a spend forecast that sits at the edge of plausible. Nvidia's moat is real, its GPUs remain the standard, and switching costs rise with every new rack installed, but a $20 trillion valuation requires the data center buildout to proceed without a hiccup, a trade war detente, or a competitive breakthrough. That is a lot of "ifs" for a company already worth $5 trillion.