How AI Data Centers Stay Cool | Explained
The short version: A tidy explainer on chillers and liquid cooling skips the part that matters most to residents: where all that heat-rejecting water and power actually comes from.
Our Take
Tech Explained Simply does a clean job walking through the engineering — CDUs, direct-to-chip liquid cooling, cooling towers — that keeps racks of GPUs from melting themselves. It's a solid primer on the plumbing. Credit where it's due: understanding the hardware helps everyone have a smarter conversation about what these buildings actually need to run.
But 'how it works' is a different question from 'what it costs the town next door.' Cooling towers evaporate water, sometimes hundreds of thousands of gallons a day at a single large campus, often drawn from municipal supplies that also serve homes and farms. Chillers and pumps run on electricity pulled from the same grid residents pay into, and utilities routinely ask ratepayers to help fund the substations and transmission upgrades that make a new campus possible. None of that shows up in a video about thermodynamics, but it shows up on local water bills and in permitting hearings.
If you want to see what this looks like on the ground — which facilities are drawing from which aquifers or rivers, and what utility deals were signed to power them — check our facility map. And if a project is moving into your county, our take-action page has the questions worth asking before the concrete gets poured.
This is GridWatch the USA’s original commentary. The video above is the work of Tech Explained Simply, published on YouTube — full credit to the creator.