Scientists find a way to slash computer memory energy use by orders of magnitude
Researchers at the University of Edinburgh have developed a theoretical framework using Optimal Control Theory to make magnetic memory switching more energy-efficient. This approach aims to reduce the electricity consumption of data centers as AI and digital data processing demands grow.
Why it matters
Improving memory efficiency is essential to curbing the rising carbon footprint and energy demands of global information technology infrastructure.
Artificial intelligence (AI) and other information and communication technologies (ICTs) are producing and processing data at a scale never seen before.
Internet searches, AI-generated images, recommendation systems, scientific simulations, and large language models all depend on enormous amounts of information being created, moved, stored, and analyzed. As AI becomes more deeply integrated into everyday life, industry, and science, the global need for computing power and data storage continues to climb.
That expansion also brings a major challenge: electricity use. Data centers already require huge amounts of power, and their energy demands are expected to rise substantially in the coming decades. Without significant improvements in efficiency, ICTs could eventually represent a sizable share of worldwide electricity consumption and carbon emissions.
Finding ways to make computing more energy efficient is therefore becoming increasingly important as demand for digital services accelerates.
Get smarter about the news
Sign up free for a feed built around what you actually care about, Dive Deeper research on any story, and the full text of every article.
Create free accountAlready have an account? Sign in