Spin Torque MRAM Write Pulse Optimization for Error Rate Control
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Spin-torque magnetoresistive random access memory (MRAM) faces challenges in achieving a low write error rate without compromising memory speed and endurance, particularly due to high variability in magnetic tunnel junction resistance and the need for sustained high switching currents, which limits scalability and increases write voltage requirements.
Innovation Solution
A method is developed to determine an optimized write pattern by adjusting the number and duration of write pulses, including the use of reversed polarity pulses, to minimize write error rates while maintaining memory speed, involving setting the write voltage to a minimum and adjusting the number of pulses and pulse duration based on desired error rates and operating conditions.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Reliability
If spin-torque MRAM uses high switching currents to achieve low write error rates, then write reliability is improved, but scalability and memory speed are compromised
Solution Approach 1:
The write operation is segmented into multiple sequential pulses instead of using a single high-current pulse. The method applies a first write pulse with a first current, followed by a second write pulse with a second current, where the second current is lower than the first current. This segmentation allows the system to achieve reliable writing while reducing the peak current requirements, thereby improving scalability and maintaining memory speed.
2Reliability
If spin-torque MRAM uses sustained high switching currents, then write reliability is improved, but device complexity and power consumption increase
Solution Approach 1:
The write operation is divided into multiple pulses with decreasing current magnitudes. The first pulse uses a higher current to initiate the switching process, while subsequent pulses use progressively lower currents to complete the transition. This segmented approach reduces total energy consumption compared to sustained high currents, as the system leverages the initial high-current pulse to start the switching and then uses lower currents to finish the operation.
Solution Approach 2:
The first high-current pulse performs a preliminary action by initiating the magnetization switching process and overcoming the energy barrier. This preliminary action prepares the magnetic moment for the subsequent lower-current pulses, which then complete the switching with less energy expenditure. This approach reduces overall power consumption while maintaining write reliability.
3Productivity
If spin-torque MRAM reduces write pulse duration to improve memory speed, then productivity is improved, but write error rate increases
Solution Approach 1:
Instead of using a single short pulse that may insufficiently switch the magnetization, the method segments the write operation into multiple pulses with controlled durations. The first pulse has a first duration and the second pulse has a second duration, where the sum of these durations achieves the required switching reliability. This segmentation allows the system to maintain shorter overall write times for improved speed while ensuring each pulse contributes effectively to the switching process, thereby maintaining low write error rates.
Applied Scientific Principles
This section explains which scientific principles are used to turn an abstract innovation direction into a practical engineering solution.
Function Achieved in This Case
This approach significantly improves write error rates without impacting memory speed or endurance, allowing for more efficient data storage and retrieval by optimizing write pulse patterns to accommodate non-ideal switching distributions and varying operating conditions.
Implementation Method 1
The angular momentum carried by the spin-polarized tunneling current causes reversal of the free layer
Implementation Method 2
exhibits an electrical resistance that depends on the magnetic state of the device
Data Source
AI summary
A method for determining an optimized write pattern for low write error rate operation of a spin torque magnetic random access memory. The method provides a way to optimize the write error rate without affecting the memory speed. The method comprises one or more write pulses. The pulses may be independent in amplitude, duration and shape. Various exemplary embodiments adjust the write pattern based on the memory operating conditions, for example, operating temperature.