Pascal Saikaly retweetledi
ON THE COVER
Infinite currents.
Artificial intelligence is often imagined as weightless—a flow of data moving through invisible digital systems. In reality, that flow depends on data centres that consume vast amounts of electricity and freshwater, much of it to remove heat.
Could that heat—and the water used to carry it away—come from somewhere already going to waste?
Just a few kilometres away, in many cities, wastewater treatment plants discharge enormous volumes of treated effluent that can serve as a coolant. Their sludge drying and anaerobic digestion, in turn, need heat to run—heat a data centre has too much of.
In this issue, Aijie Wang and colleagues show how these two infrastructures can be coupled rather than managed in isolation. By mapping 4,775 data centres and 57,547 municipal wastewater treatment plants across 98 countries, they identify opportunities for treated effluent to cool data centres, while return heat supports sludge drying and anaerobic digestion. Their optimized data–water symbiosis could reduce greenhouse-gas emissions by about 84 million tonnes of CO2 equivalent each year, conserve about 1,300 million m3 of freshwater and deliver annual net cost savings of US$95.4 billion.
As AI expands, its infrastructure need not draw endlessly on separate streams of energy and water; it can be redesigned around exchange.
Environmental Science and Ecotechnology@ESEjournal
AI may feel weightless, but its growth depends on very physical systems: electricity, cooling water, heat exchange, and urban infrastructure. ESE's new cover article about “Global data-centre–wastewater symbiosis” reimagines this challenge as a circular infrastructure opportunity. Read the article here: doi.org/10.1016/j.ese.….
English
