Back to News
Market Impact: 0.6

3 AI Infrastructure Stocks Solving the Power Crisis

NVDAVRTETNPWR
Artificial IntelligenceTechnology & InnovationCompany FundamentalsCorporate EarningsCorporate Guidance & OutlookInfrastructure & Defense
3 AI Infrastructure Stocks Solving the Power Crisis

The escalating energy demands of AI, with server racks consuming 20-100 kilowatts, are creating a critical infrastructure bottleneck, forcing hyperscalers to prioritize power grid capacity for data center locations. This trend is significantly benefiting three key players: Vertiv (VRT), which provides thermal management and power distribution solutions for high-density AI environments; Eaton (ETN), specializing in electrical power distribution equipment, including systems to manage GPU power surges; and Quanta Services (PWR), a grid infrastructure builder constructing transmission lines and substations for AI campuses. These companies are reporting strong financial performance and raised outlooks, capitalizing on the unavoidable and long-term power and cooling requirements that scale with every new AI cluster, irrespective of future chip architectures.

Analysis

The escalating energy demands of Artificial Intelligence, with server racks consuming 20 to 100 kilowatts compared to 5-15 kW for traditional servers, are creating a significant infrastructure bottleneck. This forces hyperscalers to prioritize power grid capacity for data center locations, shifting focus from traditional incentives like tax breaks. This critical constraint is generating substantial opportunities for specialized infrastructure providers. Three key players are directly benefiting: Vertiv (VRT), Eaton (ETN), and Quanta Services (PWR). Vertiv, a thermal management specialist, reported strong Q3 2025 results and raised full-year guidance, driven by AI infrastructure backlogs. Eaton, an electrical distribution leader, has launched systems to manage sudden GPU power surges, leveraging its integrated power solutions. Quanta Services, a grid infrastructure builder, shows recent revenue growth and a raised 2025 outlook from contracts tied to grid modernization for AI campuses. These companies capture unavoidable costs that scale with every new AI cluster, irrespective of future chip architectures. Their multiyear project timelines and recurring service revenues provide significant visibility and stability. The AI power bottleneck is identified as a long-term, structural issue, ensuring sustained demand for these essential infrastructure solutions.