Nonvolatile CAM for Memory Remap Information Storage
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Nonvolatile memory devices experience errors over time due to deterioration and manufacturing defects, making error correction increasingly difficult as error rates rise, leading to potential data loss and reduced memory device lifespan.
Innovation Solution
Implementing a nonvolatile, discretely-addressable memory to store remap information, allowing for finer-grained control and flexible memory management, including the use of spare memory portions to replace error-prone areas, thereby maintaining memory capacity and reducing latency.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Duration of action of stationary object
If nonvolatile memory devices are used to store data, then data persistence is achieved, but memory cells deteriorate over time leading to increased error rates
Solution Approach 1:
The system performs preliminary scanning of memory cells to identify error-prone cells before they cause data corruption. Remap information is prepared in advance and stored in a remap table, so that when errors occur, the system can immediately redirect accesses to healthy memory cells without data loss.
Solution Approach 2:
A remap table acts as an intermediary structure between the processor and the physical memory cells. When a memory cell deteriorates, the remap table stores alternative addresses that redirect accesses away from failing cells to healthy ones, mediating the interaction between the processor and deteriorating memory hardware.
2Reliability
If error correction is implemented in memory devices, then data integrity is improved, but error correction becomes difficult and impossible as error rates increase
Solution Approach 1:
The memory system is segmented into healthy and error-prone regions. The remap table divides the address space into original addresses and remapped addresses, allowing the system to handle errors at the cell level rather than requiring complex whole-block error correction mechanisms.
Solution Approach 2:
The system creates a copy of the remap information in a nonvolatile memory portion, ensuring that remap data persists even when power is lost. This copying mechanism simplifies error handling by providing redundant access paths without requiring complex real-time correction algorithms.
3Speed
If remap information is stored in volatile memory, then fast access is achieved, but remap information is lost during power-down
Solution Approach 1:
The system nests a small nonvolatile memory portion within the volatile memory structure to store remap information. This nested nonvolatile storage provides persistence without significantly impacting the overall fast access performance of the volatile memory system, as remap lookups are rare compared to normal data accesses.
Solution Approach 2:
Only the critical remap information is stored in nonvolatile memory, while the majority of memory operations continue to use fast volatile memory. This local application of nonvolatility to specific data structures maintains overall system speed while preventing remap information loss.
Data Source
AI summary
A memory device may comprise a port to receive remap information regarding a memory device and may comprise a content-addressable memory (CAM) to store the remap information, wherein the CAM may comprise a nonvolatile, discretely-addressable memory.