Vertiv reported AI datacenter earnings growth acceleration, validating power and thermal infrastructure demand.

AI infrastructure sits at the intersection of several macro themes: mixed but resilient trade, tight central bank policy, and energy and commodity shifts. While growth expectations remain steady and inflation signals uneven, there is clear, real-world demand for the physical backbone enabling AI to operate at scale. Three stocks merit particular attention for their exposure to this buildout: Lumentum Holdings, Advanced Energy Industries, and Vertiv Holdings Co.

**Lumentum Holdings** manufactures photonics components—lasers, optical chips, modules, and subsystems—supplying cloud data center and AI infrastructure providers, telecom networks, and industrial customers in semiconductor and electric vehicle manufacturing. The company generates revenue across a globally diversified footprint: the United States ($523.1m), Thailand ($494.4m), Hong Kong ($460.7m), Mexico ($318m), Other Asia Pacific ($256.1m), Europe, the Middle East and Africa ($178.7m), Japan ($96.2m), and Other Americas ($18.8m).

Lumentum sits at the heart of the AI data center buildout, supplying the high-speed optics and lasers that chipmakers and cloud providers require. Revenue and earnings are forecast to grow rapidly at 53.8% and 79.08% annually. Analysts see meaningful upside with tightly clustered price targets, while recent developments—stronger guidance, index inclusion, and Nvidia-linked demand—have provided support. The stock screens as significantly undervalued on Simply Wall St's fair value model. However, a rich valuation multiple, volatile share price, insider selling, and reliance on external funding present material risks. The key question for investors is whether Lumentum can convert AI-driven momentum into durable cash flow and returns before the cycle turns.

**Advanced Energy Industries** provides precision power conversion, measurement, and control systems for semiconductor tools, AI-ready data centers, industrial production lines, and medical equipment. The company generates approximately $1.9b in revenue from Power Electronics Conversion Products, with geographic exposure across the United States ($539.7m), Mexico ($344.1m), and other regions ($673m).

Advanced Energy is tightly linked to AI data center and advanced semiconductor spending, with earnings forecast to grow far faster than the broader US market and recent year-on-year earnings growth exceeding 100%. The company is advancing into higher-margin platforms, reporting expanding profit margins and strong order books, while returning cash through dividends and buybacks. Yet a high valuation multiple, heavy reliance on a small group of large customers, tariff exposure, and funding entirely through higher-risk borrowing introduce offsetting concerns. The critical question is whether this AI and semiconductor power specialist can sustain its conversion of design wins and backlog into durable, high-quality earnings as conditions evolve.

**Vertiv Holdings Co** provides the power, cooling, and thermal management equipment that keeps AI-intensive data centers and communication networks operational—from liquid-cooled racks for high-power GPUs to uninterruptible power supplies and integrated infrastructure. Its products and services support artificial intelligence, online gaming, e-commerce, and banking across more than 100 countries.

Vertiv generates most of its revenue in the Americas at $7.0b, with $2.4b from Asia Pacific and $2.3b from Europe, the Middle East and Africa, offset by $973.8m of intersegment sales. The company is tightly linked to the AI buildout, supplying liquid cooling, power systems, and data center infrastructure into a $15b backlog. Its earnings growth, margin expansion, and high projected ROE stand out even within AI infrastructure. Vertiv is working directly with Nvidia on next-generation thermal designs and has expanded in Malaysia and through acquisitions like PurgeRite and ThermoKey to deepen its liquid cooling and services capabilities. However, a high valuation multiple, reliance on a small group of hyperscale customers, and funding entirely through higher-risk borrowing mean expectations are already elevated and execution must remain strong.