Segmented Memory ECC Coding for DRAM Error Correction
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
DRAM faults in data centers can lead to system crashes and performance degradation, and existing error correction codes (ECCs) may not effectively correct errors beyond a certain limit, resulting in uncorrectable errors and system shutdown.
Innovation Solution
A method and memory system that segment data into multiple beats and apply distinct error correction codes (ECCs) to each segment, using a Reed Solomon ECC scheme to detect and correct errors, while also freeing up ECC bits for other functions.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Reliability
If the number of ECC bits is increased to correct more errors, then error correction capability is improved, but DRAM capacity is reduced due to additional storage requirements
Solution Approach 1:
The patent divides a 16-beat burst access into two separate 8-beat segments, applying distinct ECC codes to each segment. This segmentation allows the system to maintain strong error correction capability for each segment while reducing the total ECC overhead compared to a single comprehensive ECC code, thereby preserving more DRAM capacity for actual data storage.
Solution Approach 2:
The patent changes the ECC parameters by using different ECC code configurations for different beat segments. Specifically, it applies Reed-Solomon ECC with different code rates and symbol allocations to the first and second sets of beats, optimizing the balance between error correction capability and storage efficiency for each segment.
2Reliability
If a single comprehensive ECC code is used to cover all beats, then error detection capability is improved, but the complexity of the ECC scheme increases
Solution Approach 1:
The patent segments the burst access into multiple beats and applies different ECC codes to different segments. This segmentation simplifies the overall ECC scheme by breaking down a complex single-code problem into multiple simpler, independent ECC applications, each optimized for its specific segment.
Solution Approach 2:
The patent applies different ECC code characteristics to different beat segments based on their specific requirements. The first set of beats receives one ECC configuration while the second set receives another, allowing each segment to have locally optimized error protection tailored to its characteristics.
Data Source
AI summary
Aspects of the disclosed technology include techniques and mechanisms for an efficient error correction coding scheme that can detect and correct data errors that may occur in a memory. In general, the scheme comprises segmenting the data that would be transferred as part of a data request into different parts and applying error correction codes to the separate parts.