Can Antarctica host data centers? 🏔️
The short version: A YouTuber's Antarctica thought experiment is a fun no, but it accidentally explains why hyperscalers dump real costs on ordinary American grids, water tables, and towns instead.
Our Take
Credit to Akshay Dixit for the thought experiment: if cold alone solved data center cooling, Antarctica would be a hyperscale paradise. Instead, as the video lays out, you need cheap reliable power, fiber, fuel logistics, and skilled crews on-site around the clock — which is exactly why nobody's breaking ground on the ice shelf. But flip that logic around and it becomes an uncomfortable mirror for the U.S. buildout. The reason Virginia, Iowa, Texas, and Georgia keep winning these projects isn't that they're the best stewards of land and water — it's that they already have the grid interconnects, fiber trunks, and tax incentives ready to sign. Antarctica fails the checklist because it has no captive ratepayers to absorb the bill. Plenty of U.S. counties pass that test a little too easily.
That's the part the cooling-climate framing skips: the real constraint on hyperscale siting was never just temperature, it's who's willing to eat the transmission upgrades, the water withdrawals, and the substation noise next to their backyard. No penguins are picking up that tab here — homeowners and local ratepayers are. If a facility is proposed or already humming near you, check our facility map to see what's been approved and what hasn't, and use our take-action page if you want a say before the deal is final.
This is GridWatch the USA’s original commentary. The video above is the work of Akshay Dixit, published on YouTube — full credit to the creator.