Intra-array Non-Idealities Modeling and Algorithm Optimization for RRAM-based Computing-in-Memory Applications

In recent years, Computing-in-Memory (CIM) has shown attractive advantages over CPU/FPGA/ASIC in terms of area, energy efficiency and latency for neural network acceleration in edge applications. Among them, RRAM-based CIM accelerators has gained lots of attentions with fast read access, and low leakage power. However, the inherent non-idealities of RRAM devices, such as read non-linearity and rea...