Hybrid Silicon Substrate FinFET-Metal Insulator Metal (MIM) Memristor Based Sense Amplifier Design for the...

Area of Science:

• Semiconductor device physics
• Non-volatile memory technologies
• Integrated circuit design

Background:

• Downscaled CMOS technologies face power consumption challenges, impacting portable device longevity.
• Demand for low-cost, high-density, low-power, high-performance memory is increasing for IoT and consumer electronics.
• Memristors offer non-volatility, high switching frequency, low power, and compact size, benefiting memory device design.

Purpose of the Study:

• To propose and evaluate a novel non-volatile Static Random-Access Memory (SRAM) cell design.
• To integrate silicon substrate-based Field-Effect Transistors (FinFETs) with Metal-Insulator-Metal (MIM) memristors.
• To compare the performance of the proposed hybrid SRAM design against traditional SRAM architectures.

Main Methods:

• Design and simulation of a hybrid non-volatile SRAM cell using FinFET and MIM memristors.
• Comparison of the proposed design with standard Silicon substrate FinFET-based SRAM.
• Utilization of Monte Carlo (MC) simulations to assess device stability.
• Incorporation of high-K insulation at metal interfaces for enhanced performance.

Main Results:

• The proposed hybrid SRAM design achieved a 91.8% improvement in power consumption compared to standard Silicon substrate FinFET designs.
• The hybrid design demonstrated a significantly reduced delay of approximately 1.989 ps.
• High-K insulation integration enhanced the practicality of the architecture for low-power, high-speed memory applications.

Conclusions:

• The proposed FinFET and memristor-based non-volatile SRAM cell offers substantial power savings and speed improvements.
• This hybrid architecture is a promising solution for next-generation low-power, high-speed memory systems.
• Further Monte Carlo simulations confirmed the stability of the 6T-SRAM designs.