Here’s a fun story: the waste energy thrown off by a small compute cluster is being used to heat a public swimming pool. Zoe Kleinman, writing for the BBC:

The heat generated by a washing-machine-sized data centre is being used to heat a Devon public swimming pool.

The computers inside the white box are surrounded by oil to capture the heat—enough to heat the pool to about 30C 60% of the time, saving Exmouth Leisure Centre thousands of pounds.

The data centre is provided to the council-run centre free of charge.

Start-up Deep Green charges clients to use its computing power for artificial intelligence and machine learning.

Founder Mark Bjornsgaard said the company would also refund the leisure centre’s electricity costs for running the “digital boiler”—and seven other England pools had signed up to the scheme.

The concept, developed over five years, is relatively straight forward - the hot oil is pumped into a heat exchanger to warm the water in the pool.

As a one-off proof of concept, this is great. As a new paradigm to think about public resources and energy management, I love it.

Deep Green’s pitch is to create “climate and society friendly compute,” essentially replacing giant, exurban datacenters with small, local, urban, carbon-neutral “data furnaces”. They’re creating a more distributed AWS with a built-in heat management solution.

What if cities themselves built these out and provided the compute resources as a sort of civic benefit. Imagine dozens (hudreds? thousands??) of these “washing-machine-sized data centers” scattered across cities, connected not just to swimming pools but boilers in school basements, heat pumps in office buildings, water heaters in fire departments. They could be powered by rooftop solar panels, accessible to residents via a nominal compute fee, and reuse the heat that usually goes to waste. A local, sustainable public cloud!

Of course, you can’t just give people free compute cycles, people would immediately do terrible things with them. Surely, though, there are services you could provide — easy backups, or local caches to speed up home internet, or just power a city’s ever growing need for computers.

The compute pod as mini heat furnace is also a fun way to reconsider energy storage. A few years ago, I started thinking about how I wanted to redesign the energy usage for my house. Solar panels were an obvious choice, but what about storage? Backup batteries are expensive and strike me as wasteful, but then I started expanding my thinking beyond just electrons to energy. A water heater is a store of energy, after all, so if you use your solar panels to fill up a heat-pump-powered water heater connected to radiant floor heating, you’ve obviated the need for one of your home’s biggest consumers of electricity. A bit more obviously, the giant batteries that millions of people drive around town are now also able to power their homes. With a bit of planning, and some Bidenbucks, you could retrofit to netzero without having to install a bank of lithium in your garage.

One change I’ve noticed over the past decade or so that I find generally pretty exciting is new ways to think about climate solutions. For the longest time, the conversation got defined by the forces aligned against change — this will all be too expensive, and it won’t work anyway, we’ll just be shivering through brownouts in our solar-powered shacks. Now, try to tell me a network of distributed data furnaces isn’t cool as hell. Obviously, we can’t just keep going with unchecked growth, nor are we doomed to a grim future of limited imagination.