Data centers: Power guzzlers in the AI ​​revolution need renewable energy!

Data centers: Power guzzlers in the AI ​​revolution need renewable energy!

The increasing energy demand due to artificial intelligence (AI) represents a significant challenge for sustainable energy supply. According to digitalbusiness-cloud.de The electricity consumption of data centers will increase to over 1,000 terawatt hours (TWh) by 2030. This massive increase could account for 2% of global electricity demand, putting severe strain on energy systems.

AI investments in data centers are expected to reach $375 billion as early as 2024, representing a 44% increase compared to 2023. Hyperscalers such as Microsoft and Meta are contributing to about 55% of these investments, while AI chip suppliers including Nvidia, AMD and Intel are expected to see revenue grow from $37 billion in 2021 to $340 billion by 2027.

Climate change and energy sources

The situation becomes particularly critical in Europe, where energy consumption by data centers for AI applications could rise to over 150 TWh by 2030. This would represent almost a threefold increase and around 5% of total European electricity consumption, compared to the current level of around 2%. If this increase is not covered by renewable energy, this could further accelerate climate change tagesschau.de warns.

Much of the electricity for these data centers currently comes from fossil fuels, despite the commitment of major companies such as Amazon, Microsoft and Google to switch their energy sources. Google, for example, reported a 13% increase in greenhouse gas emissions to over 14.3 million tons of carbon dioxide, largely driven by increased energy consumption in its data centers.

Innovation in energy production

In order to meet increasing energy needs, companies are increasingly turning to innovative solutions. Google has signed a deal with Kairos Power to acquire small modular nuclear reactors (SMRs). These are scheduled to be put into operation as early as 2030 to cover energy needs. Amazon and Microsoft are also planning to use nuclear energy for their data centers to reduce their dependence on fossil fuels.

The need for a paradigm shift in energy production is obvious, as existing fossil fuels continue to dominate. Data center operators are encouraged to integrate low-carbon energy sources, including local solar and wind power generation and battery storage systems.

In addition to increasing energy consumption, there are also efforts to improve energy efficiency. AI can be used in various sectors, including agriculture and energy efficiency, to optimize the carbon footprint. But the challenges remain: experts warn that even with more efficient algorithms, energy demand will continue to rise.

In order to sustainably relieve pressure on the energy systems, massive investment must be made in renewable energies and in the expansion of the electricity infrastructure. The success of these challenges will be crucial to achieving energy goals in the context of the exponential growth of AI technologies.