Blog

Our Google searches, generative AI prompts and the various social media and entertainment streams we consume daily all require energy. In fact, some current projections state that the energy needed for AI alone could equal that of what is annually needed by a nation like Sweden or Argentina! Add to that billions of searches, nonstop AI training runs and endless streaming options, and you start to get a glimpse of the hidden engine room behind our digital lives — we’re talking about data centers.

These massive facilities are also rapidly increasing electricity demand, with the recent U.S. Department of Energy report documenting usage rapidly rising from 58 TWh (terawatt hours) in 2014 to 176 TWh in 2023. They are also projecting annual usage of 325–580 TWh by 2028, potentially doubling or tripling their share of total U.S. electricity to as much as 12%.

This surge is largely driven by artificial intelligence technology, cloud computing and digital electrification, all of which require immense server capacity and cooling. Alongside energy use, water consumption has become a pressing issue, as many centers rely on evaporative cooling that strains scarce water supplies; Google, for instance, used 5.6 billion gallons globally in 2021.

While these issues are very real, solutions are emerging! For example, Meta has piloted programs aimed at cutting the energy needed for cooling by about 20% and would reduce water use by 4%. 

Google’s Hamina, Finland facility now has successfully demonstrated how seawater cooling and heat recovery can simultaneously power servers and — as a really awesome byproduct — warm homes in the region! The Google Hamina data center system heat recovery project alone is expected to supply roughly 80% of the local district’s heating needs.

Here in the United States, policy responses are also in motion, with the Department of Energy urging the deployment of enabling technologies such as geothermal energy, enhanced storage, more efficient energy transmission, and demand flexibility programs. 

Still, major challenges remain: balancing efficiency with equipment safety; limiting water footprints in vulnerable regions; scaling up sustainable infrastructure; aligning rapid AI growth with international climate goals; and demanding clearer environmental reporting from all the organizations involved.

Ultimately, data centers sit at the global crossroads between continued digital expansion and the need for scalable, sustainable practices. With a combination of tech innovation, infrastructure reuse, regulatory oversight and organizational transparency, it may be possible to scale up computing energy needs while reducing its environmental cost.

After all, humanity was creative and passionate enough to dream up and create the digital world. Now, the question is, how do we make it sustainable for generations to come?