Abstract: The performance gap between the processors and the main memory is continuously widening, known as the memory wall or von Neumann bottleneck. Emerging nonvolatile devices have the ability of in-memory processing, and thus, have the potential to partially alleviate the memory wall bottleneck. Processing-in-memory has emerged as one of the promising techniques in the post-Moore era. In this talk, I will present 1) a basic introduction to processing-in-memory, 2) the concepts of processing-in-memory using emerging nonvolatile devices, and 2) three representative works (multi-functional processing-in-memory array, computation-memory exchangeable architecture, and processing-in-memory linear solver) of processing-in-memory in our lab.