Memristive Crossbar Arrays for Storage and Computing Applications

Abstract

The emergence of memristors with potential applications in data storage and artificial intelligence has attracted wide attentions. Memristors are assembled in crossbar arrays with data bits encoded by the resistance of individual cells. Despite the proposed high density and excellent scalability, the sneak‐path current causing cross interference impedes their practical applications. Therefore, developing novel architectures to mitigate sneak‐path current and improve efficiency, reliability, and stability may benefit next‐generation storage‐class memory (SCM). Moreover, conventional digital computers face the von‐Neumann bottleneck and the slowdown of transistors’ scaling, imposing a big challenge to hardware artificial intelligence. Memristive crossbar features colocation of memory and processing units, as well as superior scalability, making it a promising candidate for hardware accelerating machine learning and neuromorphic computing. Herein, first, crossbar architecture is introduced. Then, for storage, the origin of sneak‐path current is reviewed and techniques to mitigate this issue from the angle of materials and circuits are discussed. Computing wise, the applications of memristive crossbars in both machine learning and neuromorphic computing are surveyed, focusing on the structure of unit cells, the network topology, and the learning types. Finally, a perspective on future engineering and applications of memristive crossbars is discussed.

A Neuro-vector-symbolic Architecture for Solving Raven's Progressive Matrices

“…To reduce its computational complexity, one notable option is to use in-memory computing that executes searches in an analogue manner. It has recently been shown that the associative memory can be realized by analog in-memory computing based on crossbar arrays of emerging non-volatile memories [36][37][38][39] . Besides improving the computational density and energy efficiency, this paves the way for reducing the computational complexity of the associative memory to O(1).…”

Effects of the Operating Ambiance and Active Layer Treatments on the Performance of Magnesium Fluoride Based Bipolar RRAM

“…Resistive switching random access memory (RRAM) devices are one of the emerging non-volatile memory (NVM) technologies with two terminal metal/insulator/metal (MIM) structures [1,2]. The simple MIM structures make RRAMs integrated into dense crossbar arrays and traditional, complementary metal-oxide-semiconductors (CMOS) [2]. RRAM stores data by using different resistance states.…”

“…These similarities make the RRAM-based neuromorphic computing a promising technology for future artificial intelligence [4]. Additionally, RRAMs show potential for next-generation high-density NVMs, cryogenic computing, and artificial neural computing due to their high programming speed, low-voltage operation, high scalability and simple fabrication/integration processes [1][2][3][4][5]. Even though significant performance improvements in the RRAM device have been achieved, one remaining drawback is the large parameter variability, whose cause has been ascribed to moisture present in the atmospheric environment [6].…”

2D Heterostructure for High‐Order Spatiotemporal Information Processing

“…The software-driven neuromorphic computing has realized advanced artificial intelligence with supercomputers. [1][2][3][4][5][6][7][8] However, this technology demands huge space volume and superhigh energy dissipation compared with human brain, seriously limits their potential for large-scale application. Such issue is owing to the fundamental constraints in von Neumann computer architecture and the sequential Boolean logic.…”