Host-Managed Data Integrity Thresholds for Storage Reads
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing data storage devices manage data integrity through error handling and correction algorithms, which consume time and resources, impacting quality of service and performance, especially in mixed host workloads with quick data transfer needs.
Innovation Solution
Implementing a host-managed data integrity system where the host determines a data integrity level threshold and error correction is performed iteratively until the data meets the specified threshold, with a new communication protocol for quick syndrome weight estimation and error correction, allowing for faster data delivery.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Reliability
If error correction algorithms are performed during read operations to ensure data integrity, then data reliability is improved, but read latency increases
Solution Approach 1:
The patent applies partial action by performing syndrome weight estimation without completing full error correction decoding. The controller estimates the data integrity level using a simplified syndrome weight calculation, and only performs full error correction when the estimated integrity fails to meet the threshold. This partial approach significantly reduces read latency while maintaining data reliability.
2Reliability
If read-retry commands are issued to re-sense memory for further decoding, then data reliability is improved, but quality of service deteriorates
Solution Approach 1:
The patent performs preliminary syndrome weight estimation before issuing read-retry commands. By calculating the syndrome weight in advance and comparing it against the required threshold, the system can determine whether full error correction is needed. This preliminary action prevents unnecessary read-retry operations, improving quality of service while maintaining data reliability.
3Reliability
If excess program time is spent on program failure handling, then data reliability is improved, but quality of service for mixed workloads deteriorates
Solution Approach 1:
The patent applies partial action during program operations by performing syndrome weight estimation without completing full error correction decoding. The controller estimates the data integrity level using a simplified syndrome weight calculation, and only performs full error correction when the estimated integrity fails to meet the threshold. This partial approach significantly reduces program time while maintaining data reliability.
Data Source
AI summary
In one embodiment, a data storage device is provided comprising a memory and a controller. The controller is configured to: receive, from a host, a command to read data from the memory, wherein the command comprises a data integrity level threshold; read the data from the memory; determine a data integrity level of the data; in response to the data integrity level of the data being above the threshold, send the data to the host; and in response to the data integrity level of the data not being above the threshold: perform at least one iteration of an error correction operation on the data until the data integrity level of the data is above the threshold; and send the data to the host. Other embodiments are provided, and each of the embodiments can be used alone or in combination.