Sustainability experts have raised concerns about the future of the rapidly expanding data Centre Industry, as increased energy consumption and a growing demand for artificial intelligence (AI) collide with a hotter and drier climate. Data centres have become a critical part of infrastructure as AI fuels growth and data demands, with investment managers and superannuation funds increasing their stakes in the sector.

The International Energy Agency estimates that software-related activities currently account for 5% of global greenhouse gas emissions, a figure that may rise to 14% by 2040. While data centres and transmission networks currently account for around 1% of these emissions, some estimates predict that this could increase to over 3% in the coming years.

Major Australian data centre providers are keen to emphasize their energy efficiency credentials, with companies such as NEXTDC claiming carbon neutrality, Vocus aiming to be net-zero by 2025, and Equinix striving for climate neutrality by 2030. However, a recent report from the University of Technology Sydney’s Institute for Sustainable Futures warned that the industry is exposed to significant environmental risks, including increased cooling needs and escalating data demands in a changing climate.

As record-breaking heatwaves impact Europe, data centres in Ireland have come under scrutiny for consuming 18% of the country’s electricity, equivalent to the energy consumed by homes. Politicians have raised concerns that these energy-intensive facilities are putting pressure on the national grid, raising electricity prices, and hindering emissions reduction goals.

AI technology is drawing attention from developers and investors but requires a significant amount of energy, with training a single medium-sized generative AI model potentially consuming the same amount of electricity and energy as five American cars throughout their lifetimes. Water is also used for cooling data centres, creating additional climate impacts.

While data centre providers can mitigate environmental impacts through energy-efficient technologies and offsets, responsibility also lies with developers and consumers of AI products. Users should consider utilizing low-power AI development in cooler or less water-intensive locations, and AI models should not be implemented for every problem. By adopting hybrid approaches and making more effective use of AI technology, the environmental impact can be reduced while still benefitting from the positive advancements it offers.

The post The Growing Data Centre Industry Faces Challenges in Sustainability appeared first on TS2 SPACE.