The waste heat paradox: How data centers can cool AI with the heat AI creates
AI infrastructure has a resource problem that gets worse the more you try to solve it.
AI infrastructure has a resource problem that gets worse the more you try to solve it.
The short version
- Every generation of GPU accelerator draws more power and rejects more heat than the last — next-generation AI racks are projected to reach densities north of 300 kW , up from roughly 6 kW just a few years ago.
- The industry’s answer has been liquid cooling: direct-to-chip cold plates, rear-door heat exchangers, immersion tanks.
- But it also creates a second, quieter problem that gets far less boardroom attention than the first: liquid cooling is extremely good at capturing heat and comparatively wasteful about what happens to that heat afterward.
What happened
Most of it is still rejected to a cooling tower or dry cooler and paid for twice — once to remove it from the chip, and again to run the electric chiller plant that keeps the facility cold. A 40-year-old technology built for this exact moment Absorption chillers produce chilled water the way a conventional chiller does, but they use heat instead of electricity to drive the refrigeration cycle.
Why it matters
A generator uses hot water, steam, or exhaust gas to separate refrigerant vapor from a lithium bromide solution; the vapor condenses, evaporates under low pressure to produce the cooling effect, and is reabsorbed to close the loop.
Summary by Nerd News Network. Read the full article at Network World via the links above and below.