Phase-Change Memory Pulse Scheduling for Parallel Cell Programming
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing Phase Change Memory (PCM) technologies face challenges in efficiently performing program operations, particularly in two-cells-per-bit architectures, as they do not allow simultaneous programming of direct and complementary cells, leading to prolonged programming times.
Innovation Solution
A method for managing sequences of current pulses to optimize programming operations by maximizing tile parallelism and performing operations on direct and complementary cells simultaneously, while adhering to constraints such as IR drops and program pump capabilities.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Loss of time
If sequential programming of direct and complementary cells is performed, then programming operations can be completed with simple control logic, but programming time is prolonged
Solution Approach 1:
The set of cells to be programmed is divided into multiple subsets, where each subset contains cells that can be programmed simultaneously. The programming operation is segmented into multiple parallel operations across different subsets, allowing direct and complementary cells to be programmed in parallel while maintaining manageable control logic for each subset.
Solution Approach 2:
The control logic dynamically determines the optimal subset configuration based on the specific programming requirements. The system adaptively selects which cells to group together in each subset and determines the programming sequence, transforming the static sequential approach into a dynamic parallel processing scheme that reduces overall programming time.
2Productivity
If maximum parallelism is achieved by programming all cells simultaneously, then programming speed is maximized, but IR drops and program pump capabilities are exceeded
Solution Approach 1:
Instead of programming all cells simultaneously (excessive parallelism), the method applies partial parallelism by programming a carefully selected subset of cells at each time step. This partial action approach ensures that the current demand remains within the program pump's capability and IR drop constraints while still achieving significant speedup compared to sequential programming.
Solution Approach 2:
The control logic incorporates feedback regarding IR drops and program pump capabilities to dynamically adjust the subset selection and programming pace. By monitoring system constraints and adapting the parallelism level accordingly, the system maintains reliability while maximizing programming speed within available resource limits.
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 reduces the time required for programming by optimizing the sequence of pulses and parallelism, enhancing the efficiency and speed of program operations in PCM devices.
Implementation Method 1
the heat produced by an electric current flowing through a heating material called phase-change material such as, for instance, a chalcogenide glass, is used to melt and quench the phase-change material, making it amorphous, or to hold such phase-change material in its crystallization temperature range, thereby switching it to a crystalline state
Implementation Method 2
the heat produced by an electric current flowing through a heating material called phase-change material
Data Source
AI summary
A PCM device includes cells arranged in tiles where a set of cells is written via respective write operations. A determination is made of: a first value indicating cells in a same tile to be written in respective write operations at a common write time and a second value indicating cells to be written in respective write operations at the common write time. Based on the first and second values and whether a set write or reset write operation is performed, a number of subsets of the set of cells is selected, where each subset includes at least one cell candidate for write using application of a respective current pulse. The write operation, for each subset, includes generating a respective current pulse which is applied at the common write time to the at least one candidate cell in the subset.