Article published online by Construction Digital 6 November 2024
The growing impact of AI on data centre design and construction is reshaping the industry, driven by AI’s high computational, power, and cooling demands. Traditional data centres, once sufficient for enterprise applications, are now under strain due to AI’s requirements, which can exceed 30kW per rack compared to the 7 – 10kW for standard applications, as noted by Anna Kristín Pálsdóttir, Chief Development Officer at atNorth. This necessitates changes to power distribution and cooling strategies, including advanced technologies like liquid cooling, as emphasized by experts such as Alex Brew from Vertiv.
Strategic site selection is also crucial, with AI’s computational intensity requiring facilities in cooler climates like the Nordics for efficient cooling. Anna Kristín highlights that AI’s flexibility with uptime allows data centres to be placed near data sources, reducing latency through “edge AI.” Environmental sustainability is a key consideration, with solutions like circular economy practices and green energy to mitigate AI’s environmental impact. Rajesh Sennik from KPMG stresses the importance of incorporating sustainability from the outset to create more environmentally friendly and operationally efficient data centres.
Read the whole article here