Two Industrial Infrastructure Stocks Poised to Benefit From AI Data Center Boom

For investors fatigued by AI stock headlines, the article proposes that the real opportunity lies not in the artificial intelligence companies themselves, but in the industrial companies building the physical infrastructure those companies need. Two stocks exemplify this thesis: Vertiv Holdings and Fluor.

Vertiv manufactures industrial-scale power supplies, HVAC systems, and facilities-monitoring technology—products that sound mundane on their surface. But the company's customer base tells a different story: AI data center operators who require extraordinary levels of power efficiency and heat management. Vertiv's Q1 revenue grew 30% year over year to $2.65 billion（約4200億円）. CEO Giordano Albertazzi attributed the surge to customers "prioritizing optimized design, deployment speed, and operational efficiency" as they reshape their data center strategies. The company is guiding for comparable sales growth for the rest of the year. According to Precedence Research, the data center cooling market is likely to grow at an average annual pace of nearly 12% through 2035, while the power management aspect will grow at a slower but steady 7% yearly average over the same period. Vertiv has capitalized on this trend through partnerships like its work with Nvidia to develop 800-volt DC platforms that improve efficiency, and through liquid-cooling technology that reduces cooling energy consumption by up to 80% by making direct contact with AI processing chips. Despite strong stock performance in recent years during the AI boom, the vast majority of analysts rate Vertiv as a strong buy, with a consensus price target of $376.99—24% above its current price.

Fluor operates in the construction and engineering sector, designing and building semiconductor manufacturing facilities, chemical factories, nuclear power plants, airport infrastructure, and other large-scale projects. The business is lucrative but complicated: projects require time to fund, years to plan, and years to execute, and unforeseen cost changes can pose risks. This challenge came to a head in 2023 and has periodically weighed on the stock since. However, Fluor has taken steps to mitigate future risk by negotiating more reimbursable contracts. As of March, 82% of its $25.7 billion（約4.1兆円） backlog is structured as reimbursable (meaning Fluor passes costs through to customers), compared to only 63% at the end of 2022. This should stabilize bottom-line results for shareholders. While this year's projected revenue of just under $15.9 billion（約2.5兆円） is only about 2% better than last year's, analysts are calling for 7.5% sales growth next year, which should boost per-share profits by 27%. The reason for this spike is that several sizable higher-margin projects are expected to come into execution while older lower-margin projects finish. Standard & Poor's notes that over the past few months Fluor announced awards it expects will accelerate growth in the second half of 2026 and 2027, including limited notice to proceed on LNG Canada Phase 2 and TeraWulf's large-scale data center campus. If these awards move to the execution phase, they could convert to multiyear mega projects fueling revenue growth over several years. The article cautions that this long-term potential is not yet fully priced into the stock and recommends viewing Fluor as a long-term, buy-and-hold prospect.

The article presents a contrarian investment thesis: while technology and AI stocks have dominated market attention and headlines, the industrial infrastructure supporting AI data centers remains underappreciated by mainstream investors. This argument rests on a structural mismatch—the AI infrastructure boom creates real, concrete demand for power and cooling systems at scale, yet the companies fulfilling that demand trade at prices that do not fully reflect the opportunity.

Vertiv exemplifies this dynamic. The company's 30% year-over-year revenue growth to $2.65 billion（約4200億円） reflects genuine customer need: AI data centers require both enormous amounts of electricity and sophisticated thermal management because modern AI accelerators generate extreme heat. The article cites Precedence Research projections showing the data center cooling market will grow at nearly 12% annually through 2035, a rate substantially above typical industrial growth—yet the analyst consensus target of $376.99 (24% above the current price) suggests the market has not fully priced this durability.

Fluor's case is more nuanced because its business is inherently complex and has faced execution challenges. However, the shift in contract structure—moving from 63% to 82% reimbursable backlog—materially reduces downside risk on its $25.7 billion（約4.1兆円） project pipeline. The specific announcements for LNG Canada Phase 2 and TeraWulf's data center campus represent tangible work that is expected to execute in 2026 and 2027, when analysts project 7.5% sales growth and a 27% boost in per-share profits. The article treats this as forward-looking but grounded in announced project awards rather than speculation.