Digital data use, driven by social networks and online commerce, is rapidly increasing global energy consumption. More than 10% of the world’s electricity is now used to power data centers for companies like Google and Facebook. As the digital economy is expanding, the need for energy-efficient data storage and processing is becoming urgent.
Brain-inspired computing
Researchers from the Institute for Molecules and Materials (IMM) at Radboud University are trying to tackle this challenge with brain-inspired computing. Traditional data storage devices, such as hard drives, rely on magnetic states to store bits of information as either a 0 or 1. This technology is taking much energy. Therefore, physicists in Nijmegen are looking for ways to advance data processing and storage by developing materials that mimic how the brain works. While supercomputers consume around 50 MW of power, the human brain requires just 20 W to perform its computing tasks, combining storage and processing in parallel. By imitating biological systems, the scientists aim to develop energy-efficient computing that can adapt and learn like neural networks, potentially advancing data processing while significantly reducing energy consumption to meet future demands.