Flash Memory Seed Calculation for Faster Random Page Reads
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Flash memory devices face delays and inefficiencies in random read operations due to the need to drive a sequence generator for generating seeds, especially when accessing data in segments smaller than a full page, which can slow down data de-randomization processes.
Innovation Solution
A method and circuit for random data reading in nonvolatile memory devices that uses finite field arithmetic to generate seeds for de-randomization without requiring a sequence generator, by utilizing an initial seed and location information to quickly calculate seeds for random read operations, thereby reducing processing time.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Reliability
If a sequence generator is driven to generate seeds for random read operations, then the randomization process can be performed, but the read operation speed decreases due to the additional processing time required
Solution Approach 1:
The patent extracts the seed generation function from the sequence generator by directly calculating seeds using finite field arithmetic operations on the initial seed and location information. This eliminates the need to drive the sequence generator for seed generation, thereby removing the processing delay while maintaining the necessary randomization capability.
Solution Approach 2:
The patent replaces the mechanical sequence generator operation with a mathematical computation approach using finite field arithmetic. Instead of sequentially generating seeds through a hardware sequence generator, the system computes seeds directly using Galois field operations, which are faster and can be performed in parallel.
2Reliability
If a sequence generator is used to generate seeds for each random read operation, then proper de-randomization can be achieved, but the device complexity increases
Solution Approach 1:
The patent extracts the essential function of seed generation from the complex sequence generator hardware. By using finite field arithmetic operations on the initial seed and location information, the system achieves the same de-randomization accuracy without requiring the full sequence generator infrastructure.
Solution Approach 2:
The patent creates a mathematical model (finite field arithmetic system) that replicates the seed generation function of the sequence generator. This computational copy produces equivalent seeds for de-randomization without requiring the actual sequence generator hardware to be driven.
3Adaptability or versatility
If random read operations are performed on segments smaller than a full page, then data access flexibility is improved, but the processing time increases due to additional seed generation requirements
Solution Approach 1:
The patent changes the parameter generation method from sequential sequence generator output to direct finite field arithmetic computation. This allows seeds to be generated on-demand for any segment size or location without the time penalty of driving the sequence generator, enabling flexible random access to any data segment regardless of size.
Data Source
AI summary
A random data reading method of a nonvolatile memory device includes receiving an initial seed corresponding to a selected page of the nonvolatile memory device and relative location information of read-requested random data in the selected page. The method further includes generating a seed for randomizing the random data by subjecting the initial seed and the location information to a finite field arithmetic operation, and de-randomizing the random data based on a random sequence generated from the seed.