What it means
AI capex is the upfront investment that stands up AI capacity: accelerators and servers, the buildings and land that house them, the electrical supply and cooling, and the high-speed networking that ties clusters together. It is where spending from across the supply chain accumulates, which also makes it the point where upstream shocks land, since a shortage of chips, transformers or grid connections can delay or strand a planned build. In the economics of AI, capex is the downstream sink: the AI buildout risk is precisely the chance that data centers on the drawing board are delayed or cannot be finished. Its sheer scale and long lead times make timing, power availability and construction capacity central to whether the buildout pays off.
Why it matters to investors
AI capex is the downstream sink where upstream shocks accumulate, so its pace decides who captures the buildout, from power and construction to infrastructure, and who is exposed if projects stall. Names spanning power, construction and compute such as Vistra, Quanta Services, IREN and Crusoe sit directly in that spending path.
Companies on this part of the chain
Named to show where the term sits in the AI supply chain — research, not advice, and never a recommendation to buy or sell.
Related terms
See AI capex in the live AI chain.
THE ENTITY maps every constraint onto one live model — which part is tight now, who owns it, and who gets squeezed when it moves. Plain-English reads you can check.
THE ENTITY is an educational read on the AI supply chain — research, not investment advice. It explains how the chain works and who sits where, never price targets or buy/sell calls.