Semiconductor Capex Nears Oil Industry's Share of GDP as AI Buildout Accelerates

What the capex comparison shows

A new analysis finds that global capital spending on semiconductors is closing in on capital spending in the oil industry when both are measured as a share of world GDP. That is a striking shift for two sectors that have historically operated on very different scales, and it reflects just how much money hyperscalers, chipmakers, and equipment suppliers are now pouring into building out AI computing capacity.

Why it matters for chip and equipment stocks

Semiconductor capital spending covers everything from new fabrication plants to the equipment inside them, and a rising share of world GDP flowing into that category is a structural signal, not a one-quarter blip. For Nvidia and Broadcom, which sell the GPUs and networking silicon that hyperscale AI data centers are built around, sustained capex growth at this scale supports years of order visibility rather than a single upgrade cycle. For Applied Materials, which makes the deposition, etch, and inspection tools used to build the chip factories themselves, the connection is even more direct, since its equipment orders are effectively a line item inside that capex figure.

Which stocks, and why

Nvidia and Broadcom benefit as capex translates into continued demand for the compute and networking hardware inside new data centers, though the link runs through large customers deciding to keep spending rather than through either company's own announcements. Applied Materials sits closer to the source of the trend, since new fabrication capacity anywhere in the world generally requires its tools, making it one of the more direct beneficiaries of a semiconductor capex share of GDP that keeps climbing. None of these effects is guaranteed to continue at the current pace, since capex cycles in both energy and technology have historically turned when returns disappoint.

What to watch

Watch capital expenditure guidance from the largest hyperscalers and semiconductor equipment order backlogs in coming earnings reports, since those are the most direct readings on whether this capex trend is still accelerating or beginning to plateau. Any sign that AI infrastructure spending is being pulled back would be the clearest signal that this multi-year buildout is losing momentum.