AI In Data Centers Alters The Way Of Clean Energy Transition

By Shubhendra Anand , 19 June, 2024

In the era of the digital economy, data centers are an important aspect of helping businesses grow efficiently. The digital economy is creating more jobs in developing nations to beat current unemployment rates. However, the increasing dependency on the latest technologies is changing the industry dynamics. Hence, the digital economy supports the growth of artificial intelligence. Data centers depend heavily on artificial intelligence to enhance their work. Based on a survey in 2024, data centers are supposed to consume large amounts of fresh water. However, experts have overlooked this factor, citing only the positive aspects.

The survey shows that data centers would use more water due to on-site cooling. The data in the 2024 survey considers inputs from 2017 to 2022 regarding water usage. It shows that water consumption in the data centers has increased by 6 percent annually. This water consumption is expected to grow by 450 million gallons daily by the end of 2030. Hence, 681 gallons of daily water will be required to cool data centers. Experts suggest that this may be the reality in the next 4.5 years. The introduction of artificial intelligence in data centers arose from the clean energy transition phenomenon. This is due to artificial intelligence's use in data centers to produce clean electricity. In this technological era, companies use software to find clean electricity worldwide. The areas abundant in wind and solar energy are mostly chosen to generate electricity from renewables. This eventually leads to a reduction in carbon emissions and other infrastructural costs. However, it has severe impacts on water usage.

Increasing water demand in the data centers can cause many issues and force the closure of areas with water scarcity. Using artificial intelligence to boost the data center has its pros and cons. According to the International Energy Agency (IEA), an artificial intelligence model requires more power than 100 households to receive its training. Therefore, the shift from conventional fuels to renewables will be useless if most of its energy is drained by artificial intelligence. With the increasing use of artificial intelligence, it is challenging for data centers to go green as they require more power, which can lead to rising emissions. Solar and wind power plants alone cannot provide large amounts of energy. Thus, a greener way with renewables for the power supply in data centers seems complicated.