Balanced Programming Method for Multi-Layer Metal Oxide Memory Cells
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
The integration of multi-level metal oxide memory cells in semiconductor integrated circuits faces challenges in programming due to the need for precise control of voltage and current to achieve desired resistance states, especially with the technology node scaled down, where minimum endurance requirements are higher than for single-level cell memories.
Innovation Solution
The method involves determining the electrical characteristics of programming pulses, including voltage, current limits, and duration, based on the current and objective resistance states of the memory cells, using a table lookup operation to apply pulses that transform the resistance state effectively, allowing for direct write multi-level cell programming without stressing the memory elements.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Speed
If programming pulses with high voltage and current are applied to transform resistance state of metal oxide memory cells, then programming speed is improved, but the memory cells experience severe stress and reduced endurance
Solution Approach 1:
The patent applies dynamics by making the programming pulse characteristics adaptive rather than fixed. The controller dynamically adjusts voltage and current levels based on real-time feedback about the memory cell's resistance state and the desired target state. This dynamic adjustment allows the system to achieve fast programming when conditions permit while reducing stress when precision is needed, thereby resolving the contradiction between programming speed and cell endurance.
Solution Approach 2:
The patent implements feedback by continuously monitoring the resistance state of the metal oxide memory cell during programming operations. The controller uses this feedback information to determine whether the cell has reached the desired resistance state or if additional programming pulses are needed. This feedback mechanism prevents excessive voltage and current application by stopping programming once the target state is achieved, thus protecting cell endurance while maintaining fast programming speeds.
2Manufacturing precision
If multiple programming pulses are applied sequentially to achieve precise resistance state, then programming precision is improved, but programming time increases
Solution Approach 1:
The patent applies partial action by delivering programming pulses in controlled increments rather than applying maximum voltage and current continuously. Each pulse is designed to achieve a portion of the total required resistance change, with the controller adjusting subsequent pulse parameters based on the cell's response. This approach achieves precise resistance state transformation through multiple smaller steps rather than one large excessive action, balancing precision with acceptable programming time.
Solution Approach 2:
The patent utilizes parameter changes by varying voltage, current, and pulse duration based on the specific programming requirements and the cell's current state. The controller changes these parameters adaptively during the programming sequence, using larger pulses when rapid transition is needed and smaller, more precise pulses when fine-tuning the resistance state. This dynamic parameter adjustment achieves high precision without excessive programming time.
3Productivity
If voltage and current are increased to program memory cells faster, then productivity is improved, but runaway heating risk increases
Solution Approach 1:
The patent implements feedback by monitoring the resistance state of the memory cell during programming and using this information to control the voltage and current levels. This feedback mechanism prevents runaway heating by detecting changes in cell state that indicate approaching thermal limits and by adjusting pulse parameters to stay within safe operating margins, thus enabling high productivity without excessive heating risk.
Solution Approach 2:
The patent applies dynamics by making the programming pulse characteristics adaptive rather than fixed at maximum levels. The controller dynamically adjusts voltage and current based on real-time conditions, reducing power delivery when the cell approaches its thermal limits or when the resistance state indicates near-completion of programming. This dynamic control maintains high productivity while preventing runaway heating through continuous adaptation to cell conditions.
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 enables improved performance by balancing voltage and current during programming, maintaining operation power close to constant, reducing the risk of runaway heating, and increasing memory cell lifespan, thus achieving reliable and efficient multi-level cell programming.
Implementation Method 1
determining electrical characteristics for a programming pulse that will transform the resistance state of the selected memory cell to the objective resistance state
Implementation Method 2
balancing voltage and current during programming, maintaining operation power close to constant, reducing the risk of runaway heating
Data Source
AI summary
Improved methods for programming multi-level metal oxide memory cells balance applied voltage and current to provide improved performance. Set programming, which transitions the memory cell to a lower resistance state, is accomplished by determining an appropriate programming voltage and current limit for the objective resistance state to be achieved in the programming and then applying a pulse having the determined set electrical characteristics. Reset programming, which transitions the memory cell to a higher resistance state, is accomplished by determining an appropriate programming voltage and optionally current limit for the state to be achieved in the programming and then applying a pulse having the determined electrical characteristics. The algorithm used to determine the appropriate set or reset programming voltage and current values provides for effective programming without stressing the memory element. The electrical characteristics for programming pulses may be stored in a data table used in a table look up algorithm.