Modeling and Simulating Emerging Memory Technologies: A Tutorial

2025-02-14

AI Summary

Due to the scarcity of real non-volatile memory (NVM) devices, architectural exploration and hardware-software co-design are severely constrained. To address this, we propose and open-source a cycle-accurate, parameterized simulation toolchain tailored for emerging NVM technologies. Our toolchain is the first to systematically integrate four representative NVM use cases: hybrid main memory/cache hierarchies, in-memory computing, wear leveling, and hierarchical storage systems—enabling highly customizable, system-level modeling. Implemented in C++ and Python with modular interfaces, it balances simulation fidelity and usability. Experimental validation confirms its accuracy in reproducing key NVM characteristics—including latency, power consumption, and endurance—across diverse configurations. The toolchain has already enabled multiple cutting-edge co-design studies, effectively bridging the gap between architectural research and practical NVM deployment in the absence of widespread physical devices.

Technology Category

• Machine Learning: Hardware-aware ML
• Cognitive Modeling & Cognitive Systems: Simulating Human Behavior
• Multiagent Systems: Agent-Based Simulation and Emergent Behavior

Application Category

• Systems and Infrastructure for Web, Mobile and WoT: Web performance, measurement, and characterization
• User Modeling, Personalization and Recommendation: User modeling and simulation for interactive and conversational systems
• Web Mining and Content Analysis: Web data generation and simulation

Abstract

Non-volatile Memory (NVM) technologies present a promising alternative to traditional volatile memories such as SRAM and DRAM. Due to the limited availability of real NVM devices, simulators play a crucial role in architectural exploration and hardware-software co-design. This tutorial presents a simulation toolchain through four detailed case studies, showcasing its applicability to various domains of system design, including hybrid main-memory and cache, compute-in-memory, and wear-leveling design. These case studies provide the reader with practical insights on customizing the toolchain for their specific research needs. The source code is open-sourced.

Problem

• Modeling emerging Non-volatile Memory technologies
• Simulating NVM for architectural exploration
• Customizing toolchain for system design

Innovation

• Non-volatile Memory simulation
• Toolchain for system design
• Open-source simulation toolchain