Siemens Energy — AI supply-chain exposure

Siemens Energy sits at the power layer of the AI build-out, well upstream of any chip. It makes the heavy electrical apparatus that moves and conditions electricity at scale — most critically the large power transformers that step voltage up and down between generation and the grid connection a datacenter plugs into. Nothing computes until the site is energised, and energising a site of this size runs through exactly this kind of equipment.

Its structural hook is scarcity of the iron itself. Large transformers are slow to build, capacity is limited to a handful of serious producers, and order books have stretched as grid expansion and datacenter demand arrive at once. That makes transformer availability a genuine gating constraint on compute, and the firms that produce them hold quiet pricing power. The model places Siemens Energy centrally because grid capacity, not silicon, increasingly sets the pace of the build-out.