[Paper Review] Investigating Reliability Aspects of Memristor based RRAM with Reference to Write Voltage and Frequency

Tukaram D. Dongale, Kishorkumar V. Khot+9 more|arXiv (Cornell University)|Feb 5, 2016

Advanced Memory and Neural Computing3 citations

TL;DR

This paper investigates the impact of write voltage (0.2–1.2V) and frequency (1–200 Hz) on the reliability of memristor-based RRAM using a linear drift model. Results show that higher write voltage and lower frequency enhance memory window and lifetime (τ), with LRS dependent on both parameters, while HRS remains voltage-independent, indicating improved data retention and reduced losses under optimal conditions.

ABSTRACT

In this paper, we report the effect of write voltage and frequency on memristor based Resistive Random Access Memory (RRAM). The above said parameters have been investigated on the linear drift model of memristor. With a variation of write voltage from 0.2V to 1.2V and a subsequent frequency modulation from 1, 2, 4, 10, 100 and 200 Hz the corresponding effects on memory window, Low Resistance State (LRS) and High Resistance State (HRS) have been reported. Thus the lifetime (τ) reliability analysis of memristor based RRAM is carried out using above results. It is found that, the HRS is independent of write voltage, whereas LRS shows dependency on write voltage and frequency. The simulation results showcase that the memristor possess higher memory window and lifetime (τ) in the higher voltage with lower frequency region, which has been attributed to the fewer data losses in the memory architecture.

Motivation & Objective

• To evaluate the reliability of memristor-based RRAM under varying write voltage and frequency.
• To analyze the dependence of High Resistance State (HRS) and Low Resistance State (LRS) on write voltage and frequency.
• To estimate the lifetime (τ) of the memristor-based RRAM using simulation results.
• To identify optimal voltage and frequency combinations that maximize memory window and reliability.

Proposed method

• A linear drift model of the memristor is employed to simulate device behavior under different write voltage and frequency conditions.
• Write voltage is varied from 0.2V to 1.2V, and frequency is modulated from 1 Hz to 200 Hz.
• Memory window, LRS, and HRS are measured and analyzed across the parameter space.
• Lifetime (τ) is calculated based on the observed resistance states and their stability over time.
• Simulations are used to assess data retention and loss trends under different operational settings.
• The reliability of the RRAM architecture is evaluated based on resistance state stability and memory window size.

Experimental results

Research questions

• RQ1 How does write voltage affect the stability of HRS and LRS in memristor-based RRAM?
• RQ2 How does frequency variation influence the memory window and resistance states in the RRAM device?
• RQ3 What is the relationship between write voltage, frequency, and the estimated lifetime (τ) of the memristor?
• RQ4 Does the HRS remain independent of write voltage across different frequencies?
• RQ5 What combination of write voltage and frequency maximizes reliability and minimizes data loss?

Key findings

• The High Resistance State (HRS) remains independent of write voltage across all tested frequencies.
• The Low Resistance State (LRS) shows significant dependence on both write voltage and frequency.
• Higher write voltage combined with lower frequency results in a larger memory window and improved device lifetime (τ).
• The optimal performance, in terms of reduced data loss and enhanced reliability, is achieved at 1.2V write voltage and 1 Hz frequency.
• Simulation results confirm that fewer data losses occur in the higher voltage and lower frequency region, indicating better long-term stability.
• The lifetime (τ) of the memristor-based RRAM increases with higher write voltage and lower frequency, confirming improved reliability under these conditions.