The inhibitory function of the neuron related to biological lateral inhibition can improve energy efficiency by firing only specific neurons and enhance learning efficiency by enabling winner-takes-all (WTA) mechanism ( 21– 23). Meanwhile, neuronal inhibition and tunability of firing threshold voltage are important for an energy-efficient and reliable neural network. Memristor-based neurons were realized with a single device, diffusive memristor (SiO x:Ag), or metal-insulator transition materials (NbO x and VO x). To overcome the limitations of circuit-based neurons, few works to cointegrate memristor-based artificial neuron devices and synaptic devices in a single crossbar array for a fully memristive neural network have been reported ( 18– 20).

However, circuit-based neurons are problematic for high packing density and power consumption with low cost because they are composed of a capacitor, integrator, and comparator including many transistors ( 16, 17). Neurons are mainly composed of complementary metal-oxide semiconductor (CMOS)–based circuits, while synapses primarily comprise memristors ( 9– 15). Image processing for letter pattern recognition and face image recognition is performed using experimental-based neuromorphic simulation. Spatiotemporal neuronal functionalities are demonstrated with fabricated single-transistor neurons and synapses. The multistate single-transistor neuron that can control neuronal inhibition and the firing threshold voltage was achieved for an energy-efficient and reliable neural network. Such cointegration can enhance packing density, reduce chip cost, and simplify fabrication procedures. By virtue of 100% CMOS compatibility, it was also realized to cointegrate the neurons and synapses with additional CMOS circuits. The neurons and synapses were integrated on the same plane with the same process because they have the same structure of a metal-oxide semiconductor field-effect transistor with different functions such as homotype.

Cointegration of multistate single-transistor neurons and synapses was demonstrated for highly scalable neuromorphic hardware, using nanoscale complementary metal-oxide semiconductor (CMOS) fabrication.

Cmos transistor