Published:

AI and Crypto Mining Fuel a Surge in Data Center Energy Consumption

The International Energy Agency’s (IEA) latest report paints a startling picture — the energy consumption of global data centres, which stood at 460 terawatt-hours (TWh) in 2022, is set to skyrocket, driven predominantly by AI and crypto mining.

journey back to 2010 shows data centers accounting for approximately 1% of global electricity use, a figure that remained fairly stable over the decade. Fast forward to today, and the landscape has drastically transformed. Crypto mining alone gobbled up a quarter of this energy in 2022. With the IEA projecting a surge to between 620 and 1,050 TWh by 2026, we’re looking at an additional consumption equivalent to the entire power use of countries like Germany or Sweden.

The report highlights that the rapid expansion of the data center sector and its high electricity demand could pose challenges to the electrical system. Emphasizing the importance of efficiency improvements and regulations, the IEA suggests these measures will be crucial in managing consumption levels.

Focusing on the specifics, the IEA estimates that the electricity used for computing operations in data centers accounts for 40% of their total demand, with another 40% used for cooling the equipment. The remaining 20% is associated with other information technology devices.

The growth in electricity consumption related specifically to cryptocurrency mining is expected to increase by at least 40% by 2026, compared to the previous year. The report cites the case of Ethereum, the second-largest cryptocurrency, which reduced its electricity demand by 99% in 2022 after changing its mining mechanism. In contrast, Bitcoin, responsible for over 90% of the energy consumption in cryptocurrency mining, has not made a similar transition.

The projections for AI are even more substantial, with an anticipated tenfold increase in the next two years. This forecast considers the number of AI servers expected to be sold and their power ratings. NVIDIA, a leading AI processor manufacturer, is projected to ship around 1.5 million units by 2027, despite an expected decrease in market share.

Google, a major player in the AI space, is also mentioned in the IEA report. The total electricity consumption of the search giant was 18.3 TWh in 2021, with AI representing between 10% and 15% of this total. With AI increasingly being incorporated into most of Google’s products, a significant rise in consumption is anticipated. John Hennessy, Chairman of Alphabet (Google’s parent company), stated to Reuters that interacting with a large language model (LLM) could be “probably 10 times more expensive than a standard keyword search.”

The IEA report suggests that the key to managing this growing demand could lie in the adoption of clean energy technologies. In 2023, clean energies saw a 50% growth, reaching nearly 510 gigawatts, marking the highest increase in two decades.

Amaya Simon
Amaya Simon
Amaya Simon is a seasoned programmer specializing in smart devices, AI, and machine learning. With a Master's in Computer Science and a decade of experience, she writes to demystify complex tech. Outside work, she's a robot-building, origami-loving mom.

LEAVE A REPLY

Please enter your comment!
Please enter your name here

This site uses Akismet to reduce spam. Learn how your comment data is processed.

More from this stream