# US 12,008,240 B2 · Random write method and apparatus > Agents: this data is available as JSON from the patentgrep API and MCP server — see https://patentgrep.com/llms.txt - Number: US 12,008,240 B2 - Document ID: US12008240B2 - Status: Granted - Kind: B2 - Application number: 17/925,687 - Filed: 2021-01-23 - Published: 2023-10-26 - Granted: 2024-06-11 - Earliest claimed: 2020-05-29 - Assignees: Inspur Suzhou Intelligent Technology Co., Ltd. - Applicants: Inspur Suzhou Intelligent Technology Co., Ltd. - Inventors: Qi Song; Baolin Zhao - Web page: https://patentgrep.com/patent/US12008240B2 ## CPC - G06F 3/0613: {in relation to throughput} - G06F 3/0616: {in relation to life time, e.g. increasing Mean Time Between Failures [MTBF]} - G06F 3/064: {Management of blocks} - G06F 3/0653: {Monitoring storage devices or systems} - G06F 3/0659: {Command handling arrangements, e.g. command buffers, queues, command scheduling} - G06F 3/0679: {Non-volatile semiconductor memory device, e.g. flash memory, one time programmable memory [OTP]} - G06F 12/0246: {in block erasable memory, e.g. flash memory} - G06N 3/0442: characterised by memory or gating, e.g. long short-term memory [LSTM] or gated recurrent units [GRU] - G06F 2212/7211: Wear leveling - G06F: ELECTRIC DIGITAL DATA PROCESSING (computer systems based on specific computational models G06N) - G06N: COMPUTING ARRANGEMENTS BASED ON SPECIFIC COMPUTATIONAL MODELS ## Abstract A random write method includes: using a wear-leveling module to scan the number of free blocks and the number of bad blocks in a target super logic unit; using a lookup management module to iteratively update, according to the number of current remaining solid-state disk data frames, the number of historically weighted solid-state disk data frames in a long short-term memory network manner; using dynamic write arbitration to determine an adjustment stage based on the number of historically weighted solid-state disk data frames, and determining the expected number of read and write operations per second based on the adjustment stage; and re-updating the number of historically weighted solid-state disk data frames, and adjusting the actual number of read and write operations per second based on the re-updated number of historically weighted solid-state disk data frames and the expected number of read and write operations per second. ## Claims ### Claim 1 A random write method, comprising: determining, in a manner of leveling a remaining space, a target super logic unit from among a plurality of super logic units in a solid-state disk; scanning a quantity of free blocks and a quantity of bad blocks in the target super logic unit so as to determine a quantity of current remaining solid-state disk data frames; iteratively updating a quantity of historically weighted solid-state disk data frames in a long short-term memory network manner according to the quantity of the current remaining solid-state disk data frames; by dynamic write arbitration, determining an adjustment stage on the basis of the quantity of the historically weighted solid-state disk data frames, and determining an expected number of times of read and write operations per second on the basis of the adjustment stage; re-updating the quantity of the historically weighted solid-state disk data frames, and adjusting an actual number of times of read and write operations per second on the basis of re-updated quantity of the historically weighted solid-state disk data frames and the expected number of times of read and write operations per second; wherein the dynamic write arbitration comprises three stages, each stage comprises a corresponding relationship between the quantity of the free blocks, the quantity of the historically weighted solid-state disk data frames, and the number of times of read and write operations per second. ### Claim 2 (depends on 1) The method according to claim 1, wherein determining, in a manner of leveling a remaining space, a target super logic unit from among a plurality of super logic units in a solid-state disk comprises: determining whether a quantity of the super logic units with a minimum quantity of the free blocks in the solid-state disk exceeds a predetermined threshold value; and selecting the super logic unit with the minimum quantity of the free blocks as the target super logic unit in response to exceeding the predetermined threshold value. ### Claim 3 (depends on 2) The method according to claim 2, wherein determining, in a manner of leveling a remaining space, a target super logic unit from among a plurality of super logic units in a solid-state disk further comprises: in response to not exceeding the predetermined threshold value, scanning whether there is at least one super logic unit in a free state in the solid-state disk; and determining one of the super logic units as the target super logic unit in response to the presence of at least one super logic unit in the free state. ### Claim 4 (depends on 1) The method according to claim 1, wherein iteratively updating a quantity of historically weighted solid-state disk data frames in a long short-term memory network manner according to the quantity of the current remaining solid-state disk data frames comprises: weighting a quantity of old weighted solid-state disk data frames according to a predetermined historical average weighted parameter; and executing a normalized weighted average on the weighted quantity of the old weighted solid-state disk data frames and the quantity of standard weighted current remaining solid-state disk data frames to obtain an updated quantity of the historically weighted solid-state disk data frames. ### Claim 5 A random write apparatus, comprising: a processor; and a memory storing a program code executable by the processor, wherein the program code, when executed, executes operations as follows: determining, in a manner of leveling a remaining space, a target super logic unit from among a plurality of super logic units in a solid-state disk; scanning, a quantity of free blocks and a quantity of bad blocks in the target super logic unit so as to determine a quantity of current remaining solid-state disk data frames; iteratively updating a quantity of historically weighted solid-state disk data frames in a long short-term memory network manner according to the quantity of the current remaining solid-state disk data frames; by dynamic write arbitration, determining an adjustment stage on the basis of the quantity of the historically weighted solid-state disk data frames, and determining an expected number of times of read and write operations per second on the basis of the adjustment stage; re-updating the quantity of the historically weighted solid-state disk data frames, and adjusting an actual number of times of read and write operations per second on the basis of the quantity of the historically weighted solid-state disk data frames and the expected number of times of read and write operations per second; wherein the dynamic write arbitration comprises three stages, each stage comprises a corresponding relationship between a quantity of free blocks, the quantity of the historically weighted solid-state disk data frames, and the number of times of read and write operations per second. ### Claim 6 (depends on 5) The apparatus according to claim 5, wherein determining, in a manner of leveling a remaining space, a target super logic unit from among a plurality of super logic units in a solid-state disk further comprises: determining whether a quantity of the super logic units with a minimum quantity of the free blocks in the solid-state disk exceeds a predetermined threshold value; and selecting the super logic unit with the minimum quantity of the free blocks as the target super logic unit in response to exceeding the predetermined threshold value. ### Claim 7 (depends on 6) The apparatus according to claim 6, wherein determining, in a manner of leveling a remaining space, a target super logic unit from among a plurality of super logic units in a solid-state disk further comprises: in response to not exceeding the predetermined threshold value, scanning whether there is at least one super logic unit in a free state in the solid-state disk; and determining one of the super logic units as the target super logic unit in response to the presence of at least one super logic unit in the free state. ### Claim 8 (depends on 5) The apparatus according to claim 5, wherein iteratively updating a quantity of historically weighted solid-state disk data frames in a long short-term memory network manner according to the quantity of the current remaining solid-state disk data frames comprises: weighting a quantity of old weighted solid-state disk data frames according to a predetermined historical average weighted parameter; and executing a normalized weighted average on the weighted quantity of the old weighted solid-state disk data frames and the quantity of standard weighted current remaining solid-state disk data frames to obtain an updated quantity of the historically weighted solid-state disk data frames. ### Claim 9 A non-transitory computer-readable storage medium, storing a computer program that is executed executable by a processor, and upon execution by the processor, is configured to cause the processor to implement operations as follows: determining, in a manner of leveling a remaining space, a target super logic unit from among a plurality of super logic units in a solid-state disk; scanning a quantity of free blocks and a quantity of bad blocks in the target super logic unit so as to determine a quantity of current remaining solid-state disk data frames; iteratively updating a quantity of historically weighted solid-state disk data frames in a long short-term memory network manner according to the quantity of the current remaining solid-state disk data frames; by dynamic write arbitration, determining an adjustment stage on the basis of the quantity of the historically weighted solid-state disk data frames, and determining an expected number of times of read and write operations per second on the basis of the adjustment stage; re-updating the quantity of the historically weighted solid-state disk data frames, and adjusting an actual number of times of read and write operations per second on the basis of the quantity of the historically weighted solid-state disk data frames and the expected number of times of read and write operations per second; wherein the dynamic write arbitration comprises three stages, each stage comprises a corresponding relationship between a quantity of free blocks, the quantity of the historically weighted solid-state disk data frames, and the number of times of read and write operations per second. ### Claim 10 (depends on 9) The non-transitory computer-readable storage medium according to claim 9, wherein determining, in a manner of leveling a remaining space, a target super logic unit from among a plurality of super logic units in a solid-state disk further comprises: determining whether a quantity of the super logic units with a minimum quantity of the free blocks in the solid-state disk exceeds a predetermined threshold value. ### Claim 11 (depends on 10) The non-transitory computer-readable storage medium according to claim 10, wherein determining, in a manner of leveling a remaining space, a target super logic unit from among a plurality of super logic units in a solid-state disk further comprises: selecting the super logic unit with the minimum quantity of the free blocks as the target super logic unit in response to exceeding the predetermined threshold value. ### Claim 12 (depends on 10) The non-transitory computer-readable storage medium according to claim 10, wherein determining, in a manner of leveling a remaining space, a target super logic unit from among a plurality of super logic units in a solid-state disk further comprises: in response to not exceeding the predetermined threshold value, scanning whether there is at least one super logic unit in a free state in the solid-state disk. ### Claim 13 (depends on 12) The non-transitory computer-readable storage medium according to claim 12, wherein determining, in a manner of leveling a remaining space, a target super logic unit from among a plurality of super logic units in a solid-state disk further comprises: determining one of the super logic units as the target super logic unit in response to the presence of at least one super logic unit in the free state. ### Claim 14 (depends on 12) The non-transitory computer-readable storage medium according to claim 12, wherein iteratively updating a quantity of historically weighted solid-state disk data frames in a long short-term memory network manner according to the quantity of the current remaining solid-state disk data frames comprises: weighting a quantity of old weighted solid-state disk data frames according to a predetermined historical average weighted parameter; and executing a normalized weighted average on the weighted quantity of the old weighted solid-state disk data frames and the quantity of standard weighted current remaining solid-state disk data frames to obtain an updated quantity of the historically weighted solid-state disk data frames.