New Cambridge human brain-inspired chip could slash AI energy use — new type of memristor has roughly a million times lower switching current than conventional devices

TL;DR AI
2 min readKey summary
Researchers at the University of Cambridge reported a hafnium-oxide memristor that switches at currents near or below 10−8 A.
The device uses a multicomponent Hf(Sr,Ti)O2 film that self-assembles a p-n interface to change resistance without filament formation.
Experiments showed hundreds of conductance levels, retention over 105 seconds, and endurance beyond 50,000 cycles.
Synaptic update energy measured between ~2.5 pJ and ~45 fJ, and the devices reproduced spike timing-dependent plasticity.
The fabrication requires ~700°C deposition, which currently exceeds standard CMOS temperature limits.
