Bio-inspired computing

Bio-inspired computing operates spike signals faster and highly power efficient through stochastic resonance which uses stochastic fluctuations to amplify signals like the human brain.

As flash memory is scaled down, the probability of electron injection into the floating gate fluctuates by the quantum mechanical tunnel current, and the time to reach the threshold voltage fluctuates. These neurons exhibit stochastic resonance phenomena when applied to neural circuits in a Spiking Neural Network (SNN). Variations in memory cell characteristics and fluctuations of electron injection probability result in the capability to compute weak signals with high speed, high accuracy and high energy efficiency.

HOME RESEARCH Bio-inspired computing