Adaptive Soft-Read Suspend Scheme for SSD Command Latency
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
The high latency caused by multiple soft reads required for error correction in flash-memory SSDs reduces the Quality of Service (QoS) and can block the execution of other read commands, leading to increased overall latency and reduced performance.
Innovation Solution
A memory system with a memory controller that predicts whether and how many soft reads are needed for error correction, allowing for adaptive soft-read suspend schemes to prioritize other commands and reduce latency.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Reliability
If multiple soft reads are performed for error correction decoding, then error recovery capability is improved, but command latency increases and QoS deteriorates
Solution Approach 1:
The patent implements a dynamic soft-read suspend scheme that adaptively adjusts the number of soft reads based on real-time predictions. The memory controller predicts the number of soft reads needed for successful decoding and dynamically suspends excess soft reads, transforming the static multi-read process into a dynamic, adaptive process that balances error recovery with latency reduction.
Solution Approach 2:
The patent changes the parameter of soft-read count from a fixed value (typically 7 reads) to a variable parameter determined by prediction algorithms. By modifying this parameter dynamically based on predicted decoding requirements, the system achieves error recovery with minimal necessary reads, reducing latency while maintaining reliability.
2Measurement precision
If multiple soft reads are performed for error correction, then decoding accuracy is improved, but overall system throughput decreases
Solution Approach 1:
The patent performs preliminary prediction of the number of soft reads needed before actually executing the soft reads. This preliminary action allows the system to prepare and execute only the necessary number of reads for successful decoding, avoiding unnecessary reads that would reduce throughput while ensuring sufficient reads for accurate decoding.
Solution Approach 2:
The patent implements a feedback mechanism where the prediction algorithm learns from past decoding outcomes to improve future predictions. By incorporating feedback from actual decoding results, the system refines its prediction accuracy, thereby optimizing the balance between decoding success rate and throughput over time.
3Reliability
If soft decoding is performed with multiple reads, then error correction capability is improved, but command queue execution is blocked
Solution Approach 1:
The patent makes the soft-read process dynamic by suspending reads based on predictions, allowing the command queue to proceed with other commands while the suspended soft reads are handled selectively. This dynamic approach prevents the command queue from being completely blocked, improving ease of operation while maintaining error correction capability for critical reads.
Data Source
AI summary
Soft read suspend schemes for a memory system reduce overall command latency and improve QoS of the memory system, which includes a memory device and a memory controller. The memory controller controls the memory device to perform, in response to a command, a hard read to generate hard information for hard decoding, and predict whether soft decoding of the data is to be performed, and if so, how many soft reads are to be performed. When hard decoding fails and soft decoding and at least one soft read is to be performed, the memory device is controlled to perform an initial soft read to generate soft information for soft decoding, predict whether, and if so, how many, subsequent soft reads are to be performed, and determine whether to perform a first subsequent soft read on the data based on the prediction.