Artificial Intelligence Development, New Environmental Challenges

Artificial intelligence (AI) is significantly accelerating the demand for high-performance data centers, but the growth of these facilities brings with it new challenges in terms of energy efficiency and sustainability. According to a Goldman Sachs analysis, the need for data centers could increase by up to 160% by 2030, mainly due to the high-performance chips needed for AI, such as those produced by Nvidia. These chips are computationally intensive, produce more heat and require a highly efficient cooling system, preferably based on liquid cooling, which is proven to be more efficient than air cooling. +Liquid cooling is a more efficient method of managing temperatures in data centers, but it also comes with new environmental challenges, such as maintaining a low water temperature. Michael Winterson, president of the European Data Center Association, warns that lowering water temperatures could conflict with the European Union (EU) Energy Efficiency Directive, which requires detailed reporting of energy consumption. This requirement complicates the EU's decarbonization and sustainability plans, as large data centers are required to reduce their energy footprint to meet the region's climate goals.

• Impact on decarbonization goals

With increasing demands for AI, data centers are being forced to expand their capacity, which could put pressure on Europe's decarbonization goals. Schneider Electric is working with the EU to develop efficient energy optimization solutions. According to Schneider Electric Vice President Steven Carlini, cooling is the second largest energy consumer in data centers, and lowering water temperatures can reduce energy efficiency, increasing the overall energy consumption of facilities. In addition to liquid cooling, technological innovations in cooling equipment and efficient management of energy resources are essential for a balance between performance and sustainability. Equinix's Ferhan Gunen emphasizes that liquid cooling is crucial to control the high energy needs of dense servers and to maintain the balance between AI requirements and energy efficiency. Nebius, a company recently involved in massive investments in AI infrastructure in Europe, sees liquid cooling as "a first step" towards a more sustainable data center model. It emphasizes the importance of prioritizing energy efficiency in the growth and development of these centers, being aware of the need to protect the environment in the long term. As AI continues to drive demand for processing power, European data centers face a paradox: the need for high-performance infrastructure and the environmental imperatives imposed by decarbonization goals. Collaboration between companies and European regulators to find sustainable cooling solutions is essential. Sustainability must be a central element to enable the accelerated development of artificial intelligence without compromising climate commitments.