New brain-inspired ‘memristors’ promise to cut AI energy use
The human brain takes a different approach to traditional computers. Every synapse — the junctions where neurons communicate — both stores and processes information locally, which means memory and processing are fused inside the same biological hardware.
| Photo Credit: DOI: 10.1126/sciadv.aec2324
Researchers from the University of Cambridge have reported a new kind of brain-inspired nanodevice, a hafnium-oxide memristor, that could dramatically cut artificial intelligence’s (AI) energy bill by bringing memory and computation into the same component.
The findings were published in Science Advances on March 20.
Published – April 16, 2026 07:45 am IST
