Huang's $6.7 Trillion AI Build-Out Could Create Six-Figure Trades-If the Grid Lasts

Jensen Huang's thesis: AI spending is reaching beyond chips

Jensen Huang is talking about more than semiconductors. He says the build-out is producing six-figure salaries for trade workers such as electricians, plumbers, steelworkers, and network technicians. That points to a broader shift: AI demand is not only supporting silicon. It is also supporting the physical infrastructure needed to house and power it.

The spending wave is moving downstream

Data centers may need $6.7 trillion investment by 2030, according to recent projections, and the spending is already visible. The top five hyperscalers are expected to spend over $600 billion on infrastructure in 2026, with roughly $450 billion aimed at AI infrastructure. That points to more demand for copper, steel, power equipment, construction, and related services.

Grid delays are the main constraint

The key risk is not a lack of interest. It is the ability to complete projects on time. Nearly half of the US data centers planned for 2026 are facing delays or cancellations as power-grid bottlenecks worsen. That does not necessarily kill the theme. It does mean opportunities may shift toward companies that can help secure power access, install equipment, and keep projects moving.

For investors, the watchpoint is straightforward: follow both spending and bottlenecks. The theme works only if the grid and supply chain stay functional long enough for planned spending to turn into realized revenue.

Why investors are looking past GPUs

Nvidia's record fourth-quarter revenue of $68.1 billion, up 20% from the prior quarter and 73% from a year ago, shows that demand at the upstream end of the chain remains strong. That helps explain why investors are now looking harder at the next layers of the build-out: power, grid equipment, construction, and labor.

Utilities may have the longest revenue runway

Utilities may offer the clearest extension of the theme. American Electric Power said data centers helped drive load growth in our commercial class and said it could serve more than 20 gigawatts of new load growth by the end of the decade. If that demand becomes firm, utility cash flows could be supported by long-term electrification needs rather than a brief sentiment cycle.

Transformer shortages are a tighter bottleneck

Equipment scarcity may matter more than the broader power-demand story. Reuters reports lead times for some high-voltage transformers have increased to multiple years from around a year in 2020 and 2021. That suggests suppliers may benefit from tighter supply and longer order backlogs, even as developers face slower project timelines.

Labor constraints are becoming more visible

The labor market is also starting to reflect the build-out. Huang's comment that this work is creating six-figure salaries for trade workers highlights how physical infrastructure roles are becoming more valuable as demand rises.

The practical signal is simple: if power demand keeps climbing and transformer lead times remain long, benefits may continue to spread beyond GPUs. If those constraints ease, that broader monetization thesis weakens.

Capital spending may stay high, but the market will get tougher

The debate is shifting from whether AI spending is real to how long it can stay aggressive.

Free-cash-flow pressure is now explicit

At the current trajectory, the biggest tech spenders could spend more combined on capital expenditures than they generate in free cash flow by 2027. That does not guarantee a slowdown, but it does mean investors are likely to scrutinize returns more closely as spending rises.

Slower growth would hit a crowded trade

UBS sees hyperscaler capex growth slowing to 25% in 2027 and 6% in 2028. Even if spending still rises, that kind of deceleration can matter for valuations, especially in a market that has heavily rewarded the AI theme.

Watch three things:

• Whether grid and equipment bottlenecks stay tight
• Whether utilities and infrastructure partners show firmer demand
• Whether hyperscaler spending keeps growing fast enough to justify expectations

If those signals hold, the trade can keep extending across the infrastructure chain. If they do not, the market may focus more on spending discipline than on future capacity.