Memory Address Marking for Failed Storage Unit Isolation
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Semiconductor memories face reliability and service life issues due to poor or damaged storage units, which existing error correction codes cannot effectively address, leading to data errors and potential memory disablement.
Innovation Solution
A read-write method that marks failed storage units in real time using a lookup table with enabled/disabled marks, ensuring operations are only performed on functional units, thereby preventing data errors and extending memory lifespan.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Reliability
If error correction codes are used to address storage unit failures, then data integrity is improved, but the complexity of the memory system increases and cannot effectively prevent data errors from damaged storage units
Solution Approach 1:
The patent applies preliminary action by marking storage units as failed immediately when errors are detected during read operations, before these failed units can cause data loss in subsequent write operations. The lookup table pre-records the status of each storage unit, enabling proactive prevention rather than reactive correction.
Solution Approach 2:
The patent introduces a lookup table as an intermediary component that mediates between the storage units and the control logic. This lookup table stores the operational status of each storage unit and provides quick reference during read-write operations, simplifying the overall system complexity while improving reliability.
2Reliability
If storage units are monitored and marked as failed, then memory reliability is improved, but the device complexity increases due to additional marking mechanisms
Solution Approach 1:
The patent merges the failure marking function with the existing address translation and control logic of the memory system. The lookup table is integrated into the memory controller, combining multiple functions (address mapping, status tracking, operation control) into a unified structure, thereby minimizing additional complexity while maximizing reliability improvements.
Solution Approach 2:
The lookup table serves multiple functions simultaneously: it tracks the operational status of storage units, guides read operations to valid data locations, and prevents write operations to failed units. This multi-functionality reduces the need for separate dedicated marking mechanisms, thereby limiting the increase in device complexity.
3Reliability
If real-time marking of failed storage units is implemented, then data loss is prevented, but the operation time increases due to additional marking checks
Solution Approach 1:
The lookup table is pre-populated with the status of all storage units before normal operations begin. During read-write operations, the system performs a quick lookup in this pre-prepared table rather than performing complex real-time diagnostics, significantly reducing the time overhead while maintaining real-time protection against data loss.
Solution Approach 2:
The patent segments the memory address space into individual addressable units, each with its own status entry in the lookup table. This segmentation allows the system to perform targeted, efficient checks on only the specific storage unit being accessed, rather than scanning or checking the entire memory space, thereby minimizing the time penalty for reliability monitoring.
Data Source
Figure 1
Figure 2~3
Figure 4~5
AI summary
A read-write method and a memory are provided. The read-write method includes: issuing a read command to a memory, wherein the read command points to an address; reading a to-be-read data from a storage unit corresponding to the address to which the read command points; and in response to an error occurring in the to-be-read data, marking the address to which the read command points as disabled. When executing a read-write operation on the memory, the address of the storage unit is marked to distinguish enabled storage units from failed storage units in real time. A data error or a data loss can be avoided, thereby greatly improving the reliability and the service life of the memory.