Dual Nonvolatile Memory Segmentation for Data Rewrite Tolerance
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Conventional nonvolatile memories have limited data rewriting tolerance, leading to issues with storing data that requires frequent updates, which can result in data errors and memory failures.
Innovation Solution
Implementing a dual-memory system where high-tolerance nonvolatile memory with a higher number of rewrites is used for high-frequency data, while conventional nonvolatile memory is used for low-frequency data, thereby enhancing data rewriting tolerance and reducing storage costs.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Reliability
If conventional nonvolatile memory is used to store data with high update frequency, then storage cost is reduced, but data rewriting tolerance becomes insufficient and data errors occur
Solution Approach 1:
The memory system is segmented into two distinct nonvolatile memory devices: a first nonvolatile memory for storing data with low update frequency and a second nonvolatile memory for storing data with high update frequency. This segmentation allows each memory device to be optimized for its specific use case, with the second memory providing enhanced rewrite tolerance specifically for frequently updated data without requiring the entire system to use high-tolerance (and more complex/expensive) memory throughout.
Solution Approach 2:
Different quality levels of nonvolatile memory are applied locally based on data characteristics. The second nonvolatile memory provides higher rewrite tolerance specifically where needed (for high-frequency update data), while the first nonvolatile memory uses standard tolerance appropriate for low-frequency data. This local differentiation optimizes overall system reliability without uniformly increasing complexity across all storage operations.
2Reliability
If conventional nonvolatile memory is used for all data storage, then device complexity is minimized, but data errors occur when rewrite limit is exceeded
Solution Approach 1:
The storage resources are segmented into two nonvolatile memory devices with different rewrite tolerance capabilities. The second nonvolatile memory is specifically allocated for high-frequency update data, providing enhanced reliability where needed, while the first memory handles low-frequency data with standard reliability requirements. This segmentation prevents data errors in critical high-update scenarios without unnecessarily increasing memory resource requirements for all data.
3Productivity
If single nonvolatile memory is used, then device complexity is low, but data rewriting tolerance is insufficient for high-frequency data
Solution Approach 1:
The memory system is divided into two specialized nonvolatile memory devices, allowing high-frequency data to be stored in the second memory with higher rewrite tolerance, thus supporting faster and more reliable data updates without being constrained by the limitations of conventional single-memory architectures.
Solution Approach 2:
The second nonvolatile memory provides locally enhanced rewrite tolerance specifically for high-frequency data storage operations, enabling higher productivity in data update scenarios without requiring the entire memory system to operate at the higher tolerance level, thus balancing reliability and performance.
Data Source
AI summary
It is an object to provide a technique that can enhance data rewriting tolerance. An information processing apparatus includes a first nonvolatile memory and a second nonvolatile memory. The first nonvolatile memory stores first data. The second nonvolatile memory stores second data that is higher in update frequency than the first data. This allows the data rewriting tolerance in the device to be enhanced. For example, data of a dynamic map that is a dynamically changeable map can be applied to at least the second data among the first data and the second data.