Decision level prediction method, storage medium and SSD equipment

Abstract

The invention provides a decision level prediction method, a storage medium and SSD (Solid State Disk) equipment. The method comprises the steps: when data reading is carried out, acquiring the use state information, the read data temperature and the write data temperature of a storage unit, wherein the read data temperature is the current temperature of the storage unit, and the write data temperature is the temperature of the storage unit when data is written; predicting a decision level of the storage unit according to the use state information; and determining a decision level compensation value according to the read data temperature and the write data temperature, and correcting the predicted decision level according to the decision level compensation value. According to the invention, the use state information of the storage unit in the temperature-crossing scene is calculated, and the optimal decision level is predicted based on the use state information, so an optimal judgment level is adopted for data reading during data reading, the RBER of read data caused by temperature-crossing reading and writing is effectively reduced, the iteration time and the corresponding power consumption of an ECC decoder are reduced, and the service life of the whole SSD equipment is prolonged.

Description

technical field [0001] The invention relates to the technical field of data storage, in particular to a decision level prediction method, a storage medium and an SSD device. Background technique [0002] The current mainstream storage devices of computer servers are mainly divided into two types: HD (Hard Disk, mechanical hard disk) and SSD (Solid State Drive, solid state drive). Solid-state drives and mechanical hard drives are essentially hardware used for data storage, and the essential difference lies in their different storage media. The traditional mechanical hard disk uses a mechanical disk as the storage medium, and stores and reads data through the mechanical structure between the magnetic arm, the magnetic head, and the disk; while the solid-state hard disk uses NAND flash memory (non-volatile memory) as the storage medium. Data is read and written through the number of charges inside the memory, that is, the power on and off of the cell, and then data storage is ...

AI Technical Summary

Problems solved by technology

However, due to the influence of dwell time and cross-temperature reading and writing, the voltage difference between the default decision level and the optimal decision level is relatively large, resulting in a large number of RBER (Raw Bit Error Rate, original error bit number) in the read data sequence. )

In order to correct these erroneous bits, the ECC decoder needs to perform multiple iterative calculations, resulting in a reduction in data reading speed

Moreover, if the RBER remains high, on the one hand, it will significantly increase the delay of reading data and the corresponding chip energy consumption; on the other hand, when the number of error bits reaches a certain level, it will directly lead to the failure of decoding, making the block marked as a bad block. Waste of resources, shorten product life

Images