Background
摘要
This chapter provides background for the book. It begins with emerging non-volatile memories based on memristive devices. Switching physics is outlined, with focus on ReRAM and valence-change effects, plus key materials and models. Crossbar arrays, write schemes, and limits from sneak paths, IR drop, and leakage are described, along with 1T1R variants. Next, computing-in-memory is surveyed. Analog CIM performs parallel MAC through device conductance; DAC/ADC cost and device variability set bounds. Logic-in-memory uses binary states; notable families include IMPLY, MAGIC, and near-memory MRL. Reliability aspects follow. Fault, error, and failure are defined. Process defects across FEOL/BEOL are summarized. Traditional and ReRAM-specific fault models are listed. Fault injection methods cover hardware, software, simulation, emulation, and hybrid setups.The chapter concludes with an overview of hardware security. Fundamental properties are introduced through the CIA triad–confidentiality, integrity, and availability. Major threats are then outlined, including hardware Trojans, side-channel attacks, and active fault injection such as Rowhammer. These considerations frame the security challenges that accompany emerging memory technologies and provide essential context for the contributions presented in later chapters.