Rai, S.Talawar, B.2026-02-062025Communications in Computer and Information Science, 2025, Vol.2461 CCIS, , p. 395-40318650929https://doi.org/10.1007/978-3-031-96473-2_28https://idr.nitk.ac.in/handle/123456789/28642The continuous growth of data-intensive applications like big data analytics, machine learning and AI has led to an increasing demand for high-performance and energy-efficient memory systems. Traditional DRAM technology, while fast, suffers from significant power consumption and scalability issues. Conversely, Non-Volatile Memory (NVM) technologies, offer lower static power consumption and higher density but lag in access speed and endurance. Hybrid memory systems, which combine DRAM and NVM, aim to provide the high-speed capabilities of DRAM with the energy efficiency and non-volatility of NVM. The amalgamation of two technologies requires various modifications for the existing monolithic DRAM-based systems. In this paper, power-performance trade-offs, migration, page replacement and cache replacement for developing energy efficient hybrid memory are explored. © The Author(s), under exclusive license to Springer Nature Switzerland AG 2025.DRAMHybrid MemoryNVM