모바일 메뉴 닫기
 

연구

Research & Laboratory

제목
세미나 [09/07] Nanoporous Silicon Oxide based resistive memory devices: performance, integration and advanced architectures
작성일
2016.09.06
작성자
최고관리자
게시글 내용

< BK21 플러스 BEST 정보기술 사업단 세미나 개최 안내 >

 


개최일시 : 2016 09 07일 수요일 11:00 ~ 12:00

개최장소 : 공학원 379A

세미나 제목 : Nanoporous Silicon Oxide based resistive memory devices: performance, integration and advanced architectures

발표초록 :

As a next generation non-volatile memory (NVM), resistive random access memory (ReRAM) has been highlighted because of its excellent scalability within a simple structure, high switching speed, low programming energy per bit, and high endurance. A wide variety of materials exhibiting resistive switching behavior such as metal oxides, silicon oxide, amorphous Si, organic materials, and nanocomposites have been investigated as candidates for inclusion in future memory devices. Furthermore, the most desirable device structure for a high density non-volatile memory is the crossbar architecture composed of two electrode lines (bit- and word-lines) perpendicular to each other with memory materials sandwiched in between. This provides the highest density attainable in a two dimensional (2D) planar crossbar architecture. However, to realize highly dense three dimensional (3D) memory array, still there are several kind of critical issues such as leakage current, crosstalk and high electrochemical forming voltage.

In this talk, I will report the fabrication of 1D1R crossbar devices consisting of nanoporous SiOxunipolar memories and Pt/TiO2 Schottky diodes in order to over come the issues. The demonstrated memory device shows very high ON/OFF ratios(upto5×105), low power operation(~1.2×10-5/bit) and excellent retention(>104s), good endurance, and the capability of exhibiting multi-bit switching behavior at each cell. Each Si diode in the 1D1R cells can effectively suppress the leakage current from neighboring cells. The SiOx-based 1D1R device system has promise to satisfy the prerequisite conditions for next generation non-volatile memory applications, and is being considered for merging with probabilistic computing hardware design.

 


강연자 이승기 선임연구원 / 한국과학기술연구원 양자응용복합소재연구센터

초청자전기전자공학과 교수 안종현