Memory Compression Control for Low-Power Non-Volatile Writes
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
The existing memory systems face challenges in optimizing power consumption while achieving efficient data compression for non-volatile memory systems, as the energy required for data compression processes can outweigh the benefits of reduced write amplification.
Innovation Solution
A memory system that incorporates a compressor with a dictionary coding unit and an entropy coding unit, along with a compression efficiency monitoring unit and a gating control unit, which calculates and adjusts the operation of these units based on compression efficiencies and power reduction levels to optimize power usage and compression performance.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Productivity
If data compression is performed to reduce write amplification and effectively use memory area, then compression efficiency is improved, but power consumption increases
Solution Approach 1:
The patent implements dynamic control of the compression unit operation based on real-time monitoring of compression efficiency and power consumption. The control unit adjusts whether to perform compression dynamically by comparing current compression efficiency against threshold values, enabling the system to adapt its compression behavior to current data characteristics and power conditions, thus resolving the contradiction between maintaining compression efficiency and reducing power consumption.
Solution Approach 2:
The patent changes the operational parameters of the compression unit by introducing controllable operation modes (enabled/disabled states) based on monitored compression efficiency and power consumption levels. By adjusting these operational parameters dynamically, the system optimizes the balance between compression performance and energy usage, allowing the compression unit to operate only when beneficial.
2Productivity
If compression units always operate to maintain high compression efficiency, then data compression performance is improved, but power consumption increases
Solution Approach 1:
The patent implements a feedback mechanism where the control unit continuously monitors compression efficiency and power consumption metrics, then uses this feedback to dynamically adjust the operation of compression units. The control unit disables compression when efficiency falls below thresholds or power consumption is excessive, and re-enables it when conditions improve, creating a closed-loop system that optimizes the balance between compression performance and energy cost.
Solution Approach 2:
The patent applies partial compression action by selectively enabling compression units only for specific data blocks or operations where compression efficiency exceeds predefined thresholds. Instead of always performing full compression, the system applies compression partially and selectively, reducing overall energy consumption while maintaining adequate compression performance for suitable data types.
Data Source
AI summary
According to one embodiment, a memory system includes a compressor configured to output second data obtained by compressing input first data and a non-volatile memory to which third data based on the second data output from the compressor is written. The compressor includes a dictionary coding unit configured to perform dictionary coding on the first data, an entropy coding unit configured to perform entropy coding on the result of the dictionary coding, a first calculation unit configured to calculate compression efficiencies of the dictionary coding and the entropy coding, and a first control unit configured to control an operation of at least one of the dictionary coding unit and the entropy coding unit based on the compression efficiencies and a power reduction level.