Four AI-infrastructure stocks gaining ground in 2026

Seagate, Ciena, Bloom Energy and Constellation Energy have drawn investor interest in 2026 as demand for AI infrastructure extends beyond processors into drives, optical gear and power solutions.

Seagate has seen rising attention as data volumes from AI training and inference increase the need for large-capacity drives and storage systems. Barclays analyst Tom O’Malley described memory and storage as “the most attractive vertical below accelerators.” At the same time, Fox Advisors downgraded Seagate to Equal-Weight, warning that hard-disk pricing expectations “may be getting ahead” of likely increases.

Ciena supplies optical networking equipment used to move large volumes of data between data centers and cloud regions. Company executives raised their 2026 revenue outlook and reported a growing order backlog as cloud and AI demand firmed. Morgan Stanley adjusted its price target after Ciena’s earnings while maintaining an Equal-Weight rating, noting much of the positive outlook is reflected in the share price.

Power and electricity constraints are affecting data-center planning. Bloom Energy’s on-site fuel-cell systems offer modular, steady power at customer sites, attracting interest from AI customers seeking alternatives to long grid lead times. RBC has expressed a bullish view on Bloom, while another research house began coverage with a Market Perform rating, reflecting investor expectations after recent gains. Bloom has expanded partnerships with large cloud customers seeking dedicated energy solutions.

Constellation Energy operates a fleet that includes nuclear plants and expanded its portfolio after acquiring Calpine. The company has been highlighted by analysts for its capacity to provide steady, low-emission baseload power for large data-center customers seeking long-term contracts. Melius Research’s James West described Constellation as “well-positioned” to meet rising data-center demand in 2026.

Analysts and investors tracking the sector point to demand for storage systems to hold training datasets and archives, optical networking to move data quickly across regions, and dedicated power options to ensure continuous operation. Large cloud customers and hyperscalers are negotiating longer-term power arrangements as part of their infrastructure planning.