Neuromorphic Computing Systems with Emerging Nonvolatile Memories: A Circuits and Systems Perspective

The renaissance of artificial intelligence highlights the tremendous need for computational power as well as higher computing efficiency in both high performance computing and embedded applications. [1] To meet this demand, neuromorphic computing systems (NCS) that are inspired by the biological neural systems circumvent the von Neumann bottleneck by integrating computation and memory in the same place with reduced data traffic. The NCS can be efficiently implemented using emerging nonvolatile memories such as memristor crossbar arrays, which provide fast, analog matrix-vector multiplications (MVM). This paradigm is also called “in-memory computing” or “processing in memory”. [2]