Flexible RAID Parity in Mixed-Density Memory Blocks

Resolve Bottlenecks,
Find Innovative Solutions
Generate Solutions

Solution Overview

Problem

Existing data management systems face challenges in efficiently transferring data between memory blocks of different densities while maintaining reliability and bandwidth, particularly in solid-state drives (SSDs), as they often lack effective methods to leverage the lower read/write times of less-dense memory blocks for parity data storage.

Innovation Solution

The implementation of flexible RAID parity, where parity data is generated and stored in a less-dense memory block to ensure reliable data transfer to a more-dense block, utilizing the look-up table to maintain parity data accessibility for error correction, thereby improving data transfer reliability and bandwidth.

Engineering Contradictions & Design Principles

VSEngineering Contradiction Analysis

1Quantity of substance

If parity data is stored in the second memory block (higher density), then memory space utilization is improved, but read/write time for parity data increases

Engineering Contradiction:
Improvememory space utilizationVSAvoidread/write time for parity data
Core Design Contradiction:
Quantity of substanceVSSpeed

Solution Approach 1:

The memory device is segmented into two distinct memory blocks with different density characteristics. The first memory block stores frequently accessed parity data, while the second memory block stores data transfer operations. This segmentation allows each block to be optimized for its specific function, resolving the contradiction between space utilization and access speed.

Inventive Principle:
Principle #1Segmentation

Solution Approach 2:

Different memory blocks are assigned different quality characteristics tailored to their specific functions. The first memory block is optimized for fast read/write operations to store parity data, while the second memory block is optimized for data transfer operations. This local optimization ensures that each component performs optimally for its intended purpose.

Inventive Principle:
Principle #3Local quality

2Quantity of substance

If data is transferred from first memory block to second memory block, then data storage capacity is improved, but data transfer reliability may deteriorate

Engineering Contradiction:
Improvedata storage capacityVSAvoiddata transfer reliability
Core Design Contradiction:
Quantity of substanceVSReliability

Solution Approach 1:

Parity data is generated and stored in the first memory block before the actual data transfer operation begins. This preliminary preparation ensures that error correction information is readily available when needed, maintaining reliability during the data transfer process from the first to the second memory block.

Inventive Principle:
Principle #10Preliminary action

Solution Approach 2:

The system prepares parity data in advance and stores it in a readily accessible location (first memory block) to cushion against potential errors during data transfer. This beforehand cushioning ensures that if errors occur during the transfer to the second memory block, correction can be performed using the pre-stored parity data.

Inventive Principle:
Principle #11Beforehand cushioning (Prior cushioning)

3Speed

If parity data is stored in less-dense memory block, then read/write speed is improved, but memory density utilization decreases

Engineering Contradiction:
Improveread/write speed of parity dataVSAvoidmemory density utilization
Core Design Contradiction:
SpeedVSQuantity of substance

Solution Approach 1:

The memory device is divided into two segments with different density characteristics. The less-dense first memory block is dedicated to storing parity data that requires fast access, while the more-dense second memory block handles data transfer operations. This segmentation accepts the trade-off of lower overall density utilization in exchange for significantly improved parity data access speed.

Inventive Principle:
Principle #1Segmentation

Solution Approach 2:

The first memory block is locally optimized for speed rather than density, creating a specialized storage area for parity data. This local quality optimization acknowledges that not all memory needs to be densely packed - some areas can be optimized for specific performance characteristics.

Inventive Principle:
Principle #3Local quality

Data Source

PatentUS12481558B2Flexible raid parity application for memory management
Publication Date: 2025.11.25 SK HYNIX NAND PRODUCT SOLUTIONS CORP
  • US12481558B2 patent drawing
  • US12481558B2 patent drawing
  • US12481558B2 patent drawing

AI summary

A system and related method, the system including control circuitry and memory with a first memory block of a first memory density and a second memory block of a second memory density which is greater than the first memory density. Control circuitry, which is communicatively coupled to the memory, is configured to determine to transfer data from the first memory block to the second memory block, generate parity data based on the data and cause to store the parity data at a parity address corresponding to an available portion of the first memory block. Control circuitry is further to cause to update a look-up table with the parity data address and cause to copy the data from the first memory block to the second memory block.