Vanadium dioxide nanosheets grown directly onto standard computer chips can mimic human brain cells while consuming just 18 picojoules of energy per spike.
April 24, 2026
Original Paper
Monolithically Integrated VO$_2$ Mott Oscillators for Energy-Efficient Spiking Neurons
arXiv · 2604.21487
The Takeaway
Vanadium dioxide nanosheets integrated onto standard CMOS platforms create artificial neurons that fire with just 18 picojoules of energy. Current AI hardware struggles with massive energy consumption because it separates memory and processing. These Mott oscillators move brain-like computing from a physics concept into actual hardware compatible with today's chip factories. This hardware mimics the spiking behavior of biological neurons at a fraction of the power used by traditional digital circuits. Developing these chips could lead to ultra-low-power AI that runs locally on tiny devices instead of massive server farms.
From the abstract
Brain-inspired non-Boolean computing offers intrinsic error tolerance and parallelism, but its practical deployment is limited by the lack of compact, energy-efficient spiking hardware compatible with large-scale integration. Mott phase-transition materials provide a promising route, as their abrupt insulator-to-metal transitions enable neuron-like thresholding and oscillatory dynamics in compact devices. Among these, vanadium dioxide (VO$_2$) stands out for its near-room-temperature transition,