GE Vernova to lift output toward 24 gigawatts in 2028

The article frames the AI buildout as a power problem as much as a chip problem. Data centers need huge amounts of electricity, but generating, transmitting, and connecting that power can take years, which pushes power infrastructure into the center of the AI investment cycle. GE Vernova and Quanta Services are presented as two different ways to play that trend, one selling equipment and the other building the connections.

GE Vernova's case rests on scarce equipment. Beyond gas turbines, its electrification equipment backlog has climbed above $40 billion, more than four times the segment's 2025 revenue of about $9.6 billion. That backlog gives the company visibility as demand for transformers, switchgear and other grid equipment rises, and its growing installed base could create a second source of value through long-term servicing.

Quanta's case rests on execution. Its CEO points to skilled labor as the differentiator, with Quanta self-performing 80% to 85% of its work and more than a decade spent building its craft workforce. That matters when utilities, data centers and power developers compete for the same workers. The outcome for GE Vernova appears to hinge on whether it can keep expanding capacity without eroding the pricing and margins that make the opportunity attractive, while for Quanta the question is whether its workforce advantage holds as competition for labor continues.