Exploring Pareto-Optimal Hybrid Main Memory Configurations Using Different Emerging Memories

Main memory system design and corresponding technology requirements have become increasingly challenging for data-dominated high-performance applications. To address the leakage scalability issues of conventional DRAM-based memory, new technologies with ultra-low potential high been explored extensively over last decade. However, none them are mature enough to serve as a drop-in replacement DRAM. In this paper, we propose hybrid main solution utilizing specific features, based on target application characteristics configurations. end, examine two memories, 1S-1VCMA IGZO-based DRAM, along DRAM in context solutions high-capacity low-power Pareto-optimizations, respectively. better evaluate power performance, consider page-fault modeling our evaluations. The results simulation show that different combinations system, capacities, storage systems could provide promising regarding running applications system.