Processing-in-memory (PIM) architectures attempt to overcome the von Neumann bottleneck by combining computation and storage logic into a single component. The content-addressable parallel processing ...