Data Center Capex to Hit $1 Trillion, Boosting Chipmakers
Nvidia management projects the four largest AI hyperscalers will spend $1 trillion on data centers next year, a surge that is straining global memory chip supply and driving up component prices.
Nvidia management expects the four largest AI hyperscalers to spend $1 trillion on data center capital expenditures next year, a sharp increase from $650 billion this year. Alphabet has independently validated this trajectory, telling investors to anticipate "significantly" higher capex in 2027. The figures indicate that the physical infrastructure buildout required for an AI-first economy is accelerating rather than plateauing.
Supply Constraints Emerge
This projected spending surge is translating into immediate financial results for core hardware providers. Nvidia posted an 85% revenue increase last quarter, with Wall Street analysts projecting nearly 100% growth for the current quarter. The company's GPUs dominate the AI computing market due to their flexibility and reliability, securing Nvidia's role at the center of nearly every new data center. Notably, these figures exclude chip sales to China, a market Nvidia could re-enter in the near future, adding potential upside to the current demand.
Despite the explosive growth, Nvidia's valuation remains closely aligned with the broader market. The company trades at 23.7 times forward earnings, representing only a slight premium to the S&P 500's forward price-to-earnings ratio of 21.7.
The infrastructure boom is creating acute supply constraints further down the semiconductor stack. Micron Technology, a manufacturer of NAND and DRAM memory chips, is struggling to meet soaring data center demand. These two memory types serve different use cases in servers, but both are experiencing intense procurement pressure. Micron and its industry peers currently lack sufficient production capacity to satisfy this demand. Because constructing new foundries requires significant time, this supply shortage is not easily resolved.
As a result of this imbalance, memory prices have skyrocketed. This pricing power is directly boosting Micron's revenue and profits. For market participants, the $1 trillion capex forecast serves as a concrete benchmark for the scale of capital flowing into AI infrastructure. The data suggests that while AI software adoption remains early, the hardware cycle required to support it is firmly underway and actively straining global semiconductor manufacturing capacity.
