Urgent.News

What's breaking now, across thousands of outlets.

Tech

What If Your Prompt Could Feed Someone a Strawberry?

Every prompt ends as heat, and data centers pay to throw it away. That heat could grow food year-round at the fence line. Both political parties are missing it.

What If Your Prompt Could Feed Someone a Strawberry?

Every query you type is converted into heat. Inside a data center, the server racks turn electricity into warmth, which then dissipates through cooling systems. Unfortunately, most of that energy is wasted. In America, nearly all of the thermal output from data centers is released as heat, much like a 100-megawatt furnace that happens to compute, with the excess heat released through chillers, cooling towers, and fans.

The industry has improved efficiency over the years, but it has failed to consider whether the heat generated can be useful after leaving the rack. Liquid cooling is changing the dynamic, as it converts the heat into a more useful product, such as warm water. This warm water can be used in various ways, including growing crops in a greenhouse or heating nearby homes.

By redirecting the heat, AI-powered data centers could provide year-round produce, such as strawberries in December, greens throughout the year, and mushrooms in a dark room. The unused heat can also support fish houses, store sales, and even radiant floor heating. By incorporating this heat utilization into the rezoning process, data centers can generate additional revenue while reducing their environmental impact.

Local opposition and political campaigns are focusing on the data centers' construction, but the real issue is the waste heat they generate. Addressing the heat output is no longer a communications problem; it is now an engineering constraint that developers must consider when designing and building new facilities.

Written by urgent.news from HackerNoon's reporting — not their text. Machine-written — may contain errors; check the original before relying on it.

Read the original at hackernoon.com →

More in Tech

I am often wrong

Article URL: https://borischerny.com/management,/product/2026/09/19/I-am-often-wrong.html Comments URL: https://news.ycombinator.com/item?id=49777467 Points: 176 # Comments: 141

Faster Maps: Chasing Swiss Speed

After looking at Go's collector, we turned to its maps. Swiss tables have an appealing premise: keep entries compact, use a small amount of metadata to narrow the search, and avoid chasing a separate…

  • ParparVM improved map performance by adopting Go's compact map design
  • Benchmark showed missing key search time reduced from 32.7 seconds to 44.9 ms
  • ParparVM added support for Short, Character, Float, Long, and Double types

More from Sunday 20 September →