AI Generated Image
Scientists say ‘ultrafast magnetic-field pulses’ could help AI data centers cut their energy usage by 100x - and bring sizable improvements to RAM and data storage
Yahoo Tech • • science
Key Points:
- Researchers at the University of Edinburgh have developed a theoretical method to drastically reduce energy consumption in magnetic memory by using ultrafast magnetic-field pulses, potentially cutting energy usage by up to 100 times.
- This breakthrough could significantly lower the energy demands of AI data centers, which currently consume large amounts of electricity and contribute substantially to global carbon emissions.
- The new approach could be adapted to other technologies such as electrical currents and ultrafast laser pulses, broadening its potential impact on future data storage and computing systems.
- Although still theoretical, the researchers have outlined practical steps for building prototypes, but widespread implementation is likely far off.
- If realized, this innovation could help address the growing environmental and energy challenges posed by the rapid expansion of AI and data center infrastructure.
