Hybrid Memory Controller Refresh Management
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Conventional memory devices using only volatile memories like DIMMs have high power consumption due to the need for frequent refresh operations to retain data, while nonvolatile memories, which do not require refresh operations, have lower access speeds.
Innovation Solution
A memory device that combines both volatile and nonvolatile memories, where volatile memory is used for high-access-frequency data and nonvolatile memory for low-access-frequency data, with a controller managing data transfer and refresh operations to optimize power consumption without compromising access speed.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Speed
If volatile memory is used to achieve high access speed, then access speed is improved, but power consumption increases due to frequent refresh operations
Solution Approach 1:
The memory system is segmented into two distinct parts: volatile memory for high-access-frequency data and nonvolatile memory for low-access-frequency data. This segmentation allows each memory type to operate in its optimal mode, with the volatile memory serving only the most frequently accessed data that requires rapid access, while less frequently accessed data resides in nonvolatile memory that does not require refresh operations, thereby reducing overall power consumption while maintaining high access speeds for critical data.
Solution Approach 2:
Different quality characteristics are assigned to different parts of the memory system. The volatile memory portion is optimized for speed with high-access-frequency data, while the nonvolatile memory portion is optimized for power efficiency with low-access-frequency data. This local quality differentiation resolves the contradiction by allowing each segment to exhibit its superior characteristic without being constrained by the other.
2Use of energy by moving object
If nonvolatile memory is used to reduce power consumption, then power consumption is reduced, but access speed decreases
Solution Approach 1:
The memory system divides data storage into two segments based on access frequency patterns. Frequently accessed data is placed in volatile memory to maintain high access speeds, while infrequently accessed data is placed in nonvolatile memory to reduce power consumption. This segmentation strategy ensures that the system achieves low power consumption overall while maintaining high access speeds for the critical subset of data that is accessed frequently.
Solution Approach 2:
The system applies different quality characteristics to different data segments: volatile memory provides high-speed access for hot data, while nonvolatile memory provides power-efficient storage for cold data. This local quality approach resolves the contradiction by allowing the system to optimize for power consumption in the nonvolatile segment while preserving high access speed in the volatile segment.
3Speed
If volatile memory is used for all data, then access speed is maintained, but refresh operations increase power consumption
Solution Approach 1:
The invention extracts the refresh operation requirement from the overall memory system by moving low-access-frequency data from volatile memory to nonvolatile memory. This extraction eliminates the need for refresh operations on the extracted data portion, thereby removing the associated energy loss while maintaining high access speeds for the remaining volatile memory data that requires frequent access.
Solution Approach 2:
The memory system segments data based on access frequency, placing only the necessary portion in volatile memory that requires refresh operations. By segmenting the data storage this way, the system minimizes the amount of data subject to refresh operations, thereby reducing energy loss while maintaining adequate access speed for frequently accessed data.
4Use of energy by moving object
If nonvolatile memory is used for all data, then power consumption is reduced, but access speed for frequent data decreases
Solution Approach 1:
The memory system segments storage based on access frequency requirements, placing high-access-frequency data in volatile memory and low-access-frequency data in nonvolatile memory. This segmentation ensures that power consumption is reduced by utilizing nonvolatile memory for the majority of data, while access speed for frequent data is maintained by keeping that specific segment in volatile memory.
Solution Approach 2:
The system applies different quality characteristics to different data segments: volatile memory provides high-speed access for frequently accessed data, while nonvolatile memory provides power-efficient storage for infrequently accessed data. This local quality differentiation resolves the contradiction by allowing the system to optimize for power consumption overall while preserving high access speed where it is most needed.
Data Source
AI summary
According to one embodiment, a memory device is connectable to a host, and includes a nonvolatile memory, a volatile memory which is used as a cache of the nonvolatile memory and has a higher access speed than the nonvolatile memory, and a controller which controls access to the nonvolatile memory and the volatile memory. The controller increments, when the controller receives a refresh command for the volatile memory from the host, a value of a refresh counter, and executes, when the value of the refresh counter exceeds a threshold, no refresh operation corresponding to the refresh command.