Dynamic Memory Redundancy Configuration for Power and Reliability
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Conventional memory redundancy systems are inflexible, leading to inefficient memory usage and increased power consumption due to fixed memory configurations and redundancy levels, which do not allow for real-time configuration or adaptation to application requirements, and lack security against physical attacks and data retention errors.
Innovation Solution
A dynamically real-time configurable redundancy concept for writeable memory clusters, where the main memory and redundancy memory units are not separated by fixed size, allowing flexible configuration of memory parts and enhanced security through independent encryption keys and parallel data storage across multiple memory modules, enabling real-time adaptation of redundancy levels and detection of data integrity errors.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Reliability
If fixed memory configuration with predefined redundancy is used, then data protection and redundancy are ensured, but memory usage efficiency decreases and power consumption increases
Solution Approach 1:
The patent implements dynamic configuration of memory redundancy where the redundancy factor R can be changed at runtime. The system transitions from static fixed redundancy to dynamic adaptive redundancy, allowing the memory controller to adjust the number of redundant memory modules based on current application requirements, thereby reducing power consumption when full redundancy is not needed while maintaining data protection when required.
Solution Approach 2:
The system changes the redundancy parameter R dynamically by selecting different numbers of memory modules for redundant storage. The memory controller can adjust the redundancy level by modifying the mapping between logical addresses and physical memory locations, enabling the system to optimize between data protection and power consumption by selecting appropriate redundancy levels for different operational states.
2Reliability
If fixed memory configuration with predefined redundancy is used, then data protection and redundancy are ensured, but memory usage efficiency decreases
Solution Approach 1:
The patent enables dynamic reallocation of memory resources by allowing the redundancy configuration to change at runtime. The memory controller can adaptively adjust which memory modules are used for primary storage and which are used for redundancy based on current workload requirements, thereby improving memory usage efficiency while maintaining data protection when needed.
Solution Approach 2:
The memory system achieves multi-functionality by allowing the same physical memory modules to serve different purposes at different times - they can be used for primary data storage, redundant storage, or left unused depending on the current redundancy configuration. This universal usage approach improves overall memory efficiency while ensuring data protection capabilities are maintained.
3Device complexity
If separated memory units with fixed size are used for main memory and redundancy memory, then redundancy control is simplified, but flexibility in configuration is lost
Solution Approach 1:
The patent merges the main memory and redundancy memory functions into a unified memory system where the same physical memory modules can serve both purposes dynamically. Instead of having permanently separated memory units, the system uses a single pool of memory modules that can be configured in different redundancy arrangements, simplifying the overall memory architecture while maintaining configuration flexibility through software-controlled mapping.
Solution Approach 2:
The system implements dynamic configuration capabilities that allow the memory controller to reorganize memory modules between main storage and redundant storage roles at runtime. This dynamic reconfiguration is achieved through programmable address mapping that can adapt to different application requirements, providing both simplified control through centralized management and high flexibility in configuration.
4Speed
If parallel write/read access to redundant memory areas is implemented, then acceptable access times are achieved, but power consumption increases
Solution Approach 1:
The patent applies partial redundancy by allowing the system to use only the necessary number of redundant memory modules for each access operation rather than always activating all redundant modules. The memory controller can selectively engage R memory modules for parallel access based on the current redundancy requirement, achieving acceptable access times when needed while reducing power consumption when full parallel access is not required.
Data Source
Figure 1A
Figure 1B
Figure 2A
AI summary
In order to further develop a circuit arrangement (100) as well as a method of processing data to be protected against unauthorized access by means of encryption or decryption, by means of which method the data are stored in at least two memory modules (10, 12) in such way that a flexible configuration of any memory parts as main memory or redundancy memory is enabled, it is proposed to provide at least one real-time configurable redundancy concept for the memory modules (10, 12), by which the data can be stored redundantly in physically separate memory modules (10, 12).