The rise of artificial intelligence is undoubtedly polarizing, with the technology's wide variety of pros and cons leading people on each end of the opinion spectrum to believe it could either save the world or bring it to an end.
But what is undeniable is that AI is extremely energy-intensive. According to the International Energy Agency, electricity consumption from the data centers that power AI is expected to double by 2026.
"After globally consuming an estimated 460 terawatt-hours (TWh) in 2022, data centers' total electricity consumption could reach more than 1,000 TWh in 2026," says the IEA. "This demand is roughly equivalent to the electricity consumption of Japan."
That's why scientists are desperately trying to make AI more energy-efficient, and researchers at the University of Minnesota Twin Cities have made a breakthrough in the path to that goal.
In a statement, the University detailed that it was able to create "state of the art" hardware that could reduce the energy consumption of AI applications by a factor of around 1,000. The details of the findings were published in the NPJ Unconventional Computing journal.
"This work is the first experimental demonstration of [computer random access memory, or CRAM], where the data can be processed entirely within the memory array without the need to leave the grid where a computer stores information," said the paper's first author Yang Lv, a University of Minnesota department of electrical and computer engineering postdoctoral researcher.
The process builds on studies into Magnetic Tunnel Junctions led by the study's lead author Jian-Ping Wang. The researchers said that reconfiguring CRAM to match the performance needs of AI algorithms can lead to significant energy savings.
"We were able to obtain positive results and now have demonstrated that this kind of technology is feasible and is ready to be incorporated into technology," Wang said.
While AI has been criticized for its production of art, literature, and fake pictures and videos, among other issues, it has practical applications that can truly make a difference in the world. For example, AI is being used to protect endangered species, help create more resilient crops, and reduce the risk of wildfires.
🗣️ Will America someday get all its energy from renewable sources?
🔘 Yes — very soon ☀️
🔘 Yes — by 2050 ⏰
🔘 Yes — by 2070 ⏳
🔘 Probably never ❌
🗳️ Click your choice to see results and speak your mind
But doing so at a significant environmental cost is a problem. The energy sector is one of the world's biggest producers of planet-warming pollution, and despite the rise in sustainable technologies, dirty fuel is still responsible for the creation of most of our power. In fact, some of the issues AI is being utilized to solve could be made worse by using it.
Finding a balance is essential, and the research from the team at the University of Minnesota Twin Cities has taken a significant step toward doing so.
Join our free newsletter for weekly updates on the latest innovations improving our lives and shaping our future, and don't miss this cool list of easy ways to help yourself while helping the planet.
Cool Picks
