RRAM-Based in-Memory Computing for Intelligent Applications

The rapid advancement of artificial intelligence has fueled a heightened demand for highly energy-efficient chips. Nevertheless, conventional von Neuman computation architectures encounter critical bandwidth bottlenecks when handling vast amounts of real-time data processing. In-memory computing (IMC) paradigms leveraging resistive random-access memory (RRAM) have demonstrated remarkable potential in enabling these burgeoning intelligent applications. In this paper, we have demonstrated our efforts on RRAM-based IMC systems, encompassing the design of RRAM devices and the implementation of RRAM-based IMC chips. We envision these works may set the stage for future applications in RRAM in-memory computing technology.