Vikas Bhatnagar, Adesh Kumar
Phase-change memory (PCM) is a promising nonvolatile memory technology that stores information through reversible, thermally driven phase transitions in chalcogenide materials, particularly Ge 2 Sb 2 Te 5 (GST225). The significant resistance contrast between the crystalline and amorphous states enables data storage, while SET (crystallization) and RESET (amorphization) operations rely on localized Joule heating. As a result, temperature critically influences data retention, switching speed, programming energy, and long-term device reliability. This review analyses the multifaceted impact of temperature on PCM across material, device, and array scales. It highlights key thermal challenges, including high RESET power, thermal crosstalk, resistance drift, and endurance degradation thatintensify as device dimensions continue to scale down. PCM behavior is shown to depend strongly on phase-change kinetics, device geometry, thermal boundary resistance, and the thermal history of programming cycles. Recent innovations in material engineering, such as dopant-stabilized chalcogenide alloys and superlattice-like structures, are discussed for their potential to improve thermal stability and reduce switching energy. Additionally, architectural strategies, including confined-cell design and interface engineering, are evaluated for their ability to enhance thermal confinement. Algorithmic and circuit-level compensation techniques addressing temperature-induced variability are also surveyed. The novelty of this review lies in its integrated perspective , c ombining materials, device physics, circuit strategies, and emerging computing paradigms to present a unified understanding of temperature-driven behavior in PCM. The work further examines temperature-related implications for advanced applications, including in-memory and neuromorphic computing. Overall, it outlines key pathways toward thermally robust, energy-efficient PCM technologies.