Memristors

We develop high-reliable and ultralow energy consumption artificial synaptic devices based on memristive materials. The metal/insulator/metal-structured devices show highly linear synaptic plasticity to implement in the next-generation crossbar array architecture. Various neural networks are conducted to implement neuromorphic computing. Our memristor-based artificial synapses pave the way toward neuromorphic computing systems.

The Lithium ion-based two-terminal synaptic device is under research for improved weight updating in hardware-based neuromorphic systems. Under programming voltage pulse or bias sweep, the resistance state of Li-based oxide thin film can be adjusted continuously which corresponds to synaptic weight, hence emulating biological synapse plasticity.

To eliminate the data bus, a new hardware which makes possible in-memory operations is being studied. In the case of redox RSMs, the redox reaction rate and ion migration speed decides the switching speed. Halide perovskites are promising materials for resistive switching. Also, they allow facile fabrications and are easy to tune the bandgap through controlling the types and the ratio of Halogen elements. We develop copper based HPs, which are eco-friendly and relatively environmentally stable. These are suggested as artificial synapses.

We have aimed at development of resistive switching devices based on halide perovskites which attracts the attention in ReRAM engineering for new materials. Futhermore, we fave focused on fabrication of crossbar array through micro & nano patterning with photo-lithography and e-beam lithography to integrate them in an array and attempt to redesign their structure. Finally, Our goals are their application in storage class memory and demonstration of neuromorphic system.

We fabricate a memristor, based on halide perovskites such as CsPbI_3. Especially, making halide perovskites into quantum dot is a hot topic nowadays since quantum dot opens a new door to new materials with high stability, and much more improved efficiency. Applying halide perovskite quantum dots to memristor is also a very interesting field. We expect our research into halide perovskite quantum dots in addition to resistance switching memory with high endurance will become a crucial step toward a next generation memory industry.