Current Integration Sensing for LDPC Memory Read Probabilities
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
The existing methods for generating probabilistic information in low-density parity check (LDPC) codes for error correction in memory devices require multiple read strobes, which increase power consumption, latency, and interface bandwidth, negatively affecting memory performance.
Innovation Solution
A method involving a sense circuit that performs multiple detections using current integration techniques to generate one-sided and two-sided probabilistic information by varying the sense voltage and timing, allowing for refined probabilistic information generation with reduced number of strobes, thereby improving error correction efficiency.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Reliability
If multiple read strobes are performed to generate probabilistic information for LDPC decoding, then error correction capability is improved, but power consumption increases
Solution Approach 1:
The patent changes the parameter of sensing operations by performing multiple detections with varying sense voltages and timing rather than multiple full read strobes. This allows generation of probabilistic information with reduced power consumption by adjusting voltage levels and timing parameters of the sensing operations.
Solution Approach 2:
The patent applies partial action by performing multiple partial sensing operations that gather sufficient probabilistic information without completing full read strobes each time. The sense circuit performs detections that are sufficient for generating LLR values but less intensive than complete read operations, thereby reducing overall power consumption.
2Reliability
If multiple read strobes are performed to generate probabilistic information, then error correction capability is improved, but latency increases
Solution Approach 1:
The patent performs preliminary sensing operations that gather probabilistic information before the actual read operation. By conducting multiple detections with varying voltages and timing in advance, the system prepares LLR values that can be used immediately during decoding, reducing the latency of the overall read process.
Solution Approach 2:
The patent maintains continuous useful action by overlapping sensing operations with other memory operations. The sense circuit performs multiple detections during the read process itself, rather than requiring separate sequential operations, thereby maintaining continuous productive activity and reducing idle time.
3Reliability
If multiple read strobes are performed to generate probabilistic information, then error correction capability is improved, but interface bandwidth increases
Solution Approach 1:
The patent extracts only the necessary probabilistic information from multiple sensing operations without transferring all the intermediate data through the interface. The sense circuit processes the detection results locally to generate LLR values, extracting only the essential probabilistic data needed for decoding rather than transmitting all raw sensing data.
Solution Approach 2:
The sense circuit acts as an intermediary that processes sensing data locally and generates probabilistic information in a compressed form. Rather than passing all intermediate sensing data through the memory interface, the sense circuit mediates by converting raw detections into compact LLR values that require less bandwidth for transmission.
4Use of energy by moving object
If current integration sensing with varied voltages and timing is used to generate probabilistic information, then power consumption is reduced, but measurement precision requirements increase
Solution Approach 1:
The patent applies dynamics by varying the sense voltage and timing of detections adaptively. The sense circuit performs multiple detections with different voltage levels and timing parameters, dynamically adjusting the sensing conditions to optimize both power consumption and measurement precision. This dynamic approach allows the system to gather sufficient probabilistic information while maintaining acceptable power levels.
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 power consumption and latency while maintaining high error correction efficiency by refining probabilistic information generation, thus improving memory performance and compliance with timing requirements.
Implementation Method 1
A method involving a sense circuit that performs multiple detections using current integration techniques to generate one-sided and two-sided probabilistic information
Data Source
AI summary
An example method for determining likelihood of erroneous data bits stored in memory cells may include sensing a first plurality of memory cells based on a first sense thresholds. Responsive to sensing the first plurality of cells, a first set of probabilistic information may be associated with the first plurality of memory cells. A second plurality of memory cells may be sensed based on a second sense threshold. Responsive to sensing the second plurality of memory cells, a second set of probabilistic information may be associated with the second plurality of memory cells. An error correction operation may be performed on the first and second pluralities of memory cells based, at least in part, on the first and second values.