DRAM Data Flipping for Error Probability Reduction
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Dynamic random access memory (DRAM) experiences high error probabilities due to electric charge leakage, leading to frequent refresh operations that increase power consumption and system bandwidth usage.
Innovation Solution
A memory access method that determines whether to flip data based on the quantities of '1' and '0' in the data and the storage mode of the DRAM, reducing the number of storage cells with high electric charges by flipping data when necessary, thereby reducing error probabilities.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Reliability
If the refresh time interval is reduced to handle high error probability, then data reliability is improved, but power consumption increases and system bandwidth is occupied
Solution Approach 1:
The patent applies local quality by differentiating refresh operations based on row characteristics. Instead of uniformly refreshing all rows at fixed intervals, the system identifies rows with high error probability (worst cells) and applies targeted refresh operations only to those specific rows. This selective approach reduces overall power consumption while maintaining data reliability for critical rows.
Solution Approach 2:
The patent implements dynamic refresh intervals by adjusting the refresh time interval based on detected error probabilities. The system monitors data errors in real-time and adapts the refresh frequency dynamically - increasing refresh rate for rows showing high error rates and reducing it for stable rows. This dynamic adaptation optimizes both reliability and power consumption.
2Reliability
If the refresh time interval is reduced to handle high error probability, then data reliability is improved, but system bandwidth is occupied
Solution Approach 1:
The patent applies local quality by differentiating refresh operations based on row characteristics. Instead of uniformly refreshing all rows at fixed intervals, the system identifies rows with high error probability (worst cells) and applies targeted refresh operations only to those specific rows. This selective approach reduces overall power consumption while maintaining data reliability for critical rows.
Solution Approach 2:
The patent implements dynamic refresh intervals by adjusting the refresh time interval based on detected error probabilities. The system monitors data errors in real-time and adapts the refresh frequency dynamically - increasing refresh rate for rows showing high error rates and reducing it for stable rows. This dynamic adaptation optimizes both reliability and power consumption.
3Device complexity
If data is stored without flipping, then storage simplicity is maintained, but error probability increases due to high electric charges in storage cells
Solution Approach 1:
The patent applies preliminary action by performing data flipping before storage when the data contains a high proportion of '1' bits. The system analyzes the data pattern beforehand and proactively inverts the data to reduce the number of storage cells that would hold high electric charges. This preventive measure reduces error probability before the data is even written to memory, maintaining storage simplicity while improving reliability.
Data Source
AI summary
A memory access method and a computer system are provided. According to the memory access method, whether to flip the to-be-stored data for storage may be determined based on quantities of “1” and “0” in data to be written into a dynamic random access memory (DRAM) and a storage mode of the DRAM, to reduce a quantity of storage cells with high electric charges in the DRAM, thereby reducing a data error probability.