This chapter surveys prior research on neuromorphic computing with emphasis on reliability and hardware security of memristor-based systems. The first part examines reliability challenges in ReRAM devices and crossbar structures, including IR drop, read/write disturbances, and variability, along with error correction codes, circuit-level optimizations, and architectural strategies proposed to improve endurance and stability. Reliability aspects of computing-in-memory applications are then addressed, highlighting error correction, matrix transformations, mapping algorithms, and training frameworks that mitigate non-ideal effects. Simulation platforms such as RxNN, PytorX, NeuroSim, and SySCIM are reviewed for their ability to model performance, energy, variability, and fault behavior. The discussion then turns to hardware security, covering hardware Trojans, side-channel attacks, and fault injection attacks targeting memristive systems and accelerators. Finally, instrumentation platforms for testing and characterization of single devices and crossbar arrays are summarized. These surveys reveal open gaps in bridging reliability evaluation, hardware security, and practical system validation.