Storage devices, controllers, and non-volatile storage devices

By introducing command queuing and merging operations into the storage device, the problem of limited command processing performance optimization in non-volatile memory is solved, and memory bank conflicts are reduced and performance is improved.

CN123114359APending Publication Date: 2026-11-17SAMSUNG ELECTRONICS CO LTD
View PDF 5 Cites 0 Cited by

Patent Information

Application Number
CN202610643562.0
Authority / Receiving Office
CN · China
Patent Type
Applications(China)
Current Assignee / Owner
Priority Date
2025-12-09
Filing Date
2026-05-11
Publication Date
2026-11-17

AI Technical Summary

Technical Problem

In existing technologies for non-volatile memory, the timing complexity and dependencies of programming, reading, and erasing operations limit command processing performance optimization and prevent efficient utilization of memory resources.

Method used

By introducing command queue circuits, queue manager circuits, and non-volatile memory interface circuits into the storage device, command queuing and merging operations are realized. The controller and non-volatile storage devices work together to merge commands from the same storage bank and execute them concurrently.

Benefits of technology

It reduces memory conflicts, improves storage device performance, enhances input/output per second (IOPS) and throughput, and optimizes overall operation latency.

✦ Generated by Eureka AI based on patent content.

Smart Images

  • Figure CN123114359A_ABST
    Figure CN123114359A_ABST
Patent Text Reader

Abstract

A storage device, a controller, and a non-volatile storage device are provided. The storage device includes non-volatile memory and a controller. The controller includes: a command queuing circuit that stores multiple queues corresponding to commands received from a host; a queue manager circuit configured to select a first queue corresponding to a first physical address and a second queue corresponding to a second physical address based on the physical addresses of the multiple queues, and to merge the second queue into the first queue; and a non-volatile memory interface circuit configured to sequentially send a first queuing command corresponding to the first queue, a second queuing command corresponding to the second queue, and a merge command corresponding to the first and second queues to the non-volatile memory. In response to the merge command, the non-volatile memory is configured to concurrently execute a first storage operation corresponding to the first physical address and a second storage operation corresponding to the second physical address.
Need to check novelty before this filing date? Find Prior Art

Description

Technical Field

[0001] This disclosure relates to storage devices, and more specifically, to storage devices, controllers, and non-volatile storage devices that support command queuing operations. Background Technology

[0002] In semiconductor memory devices, high-speed data processing requires efficient command processing architectures. Typically, memory devices primarily process commands input sequentially from external sources or by storing commands in simple internal buffers. However, in non-volatile memories such as NAND flash memory, programming, reading, and erasing operations involve complex timing and dependencies. Therefore, performance optimization relying solely on simple sequential processing schemes has limitations. Consequently, techniques are needed to enhance command processing performance while efficiently utilizing the resources within the memory device. Summary of the Invention

[0003] According to one aspect of this disclosure, a storage device is provided, the storage device comprising: a non-volatile memory; and a controller configured to control the non-volatile memory. The controller includes: a command queue circuit configured to store a plurality of queues corresponding to a plurality of commands received from a host; a queue manager circuit configured to select a first queue corresponding to a first physical address and a second queue corresponding to a second physical address based on the physical addresses of the plurality of queues, and to merge the second queue into the first queue; and a non-volatile memory interface circuit configured to sequentially send a first queuing command corresponding to the first queue, a second queuing command corresponding to the second queue, and a merge command corresponding to the first queue and the second queue to the non-volatile memory, wherein the non-volatile memory is configured to concurrently execute a first storage operation corresponding to the first physical address and a second storage operation corresponding to the second physical address in response to the merge command.

[0004] According to another aspect of this disclosure, a controller is provided, the controller comprising: a command queue circuit configured to store a plurality of queues corresponding to a plurality of commands received from a host; a queue manager circuit configured to merge a first queue and a second queue based on the physical addresses of the plurality of queues, and to set the first queue as a representative queue, and to release the second queue from the command queue circuit; and a host interface circuit configured to send a request for additional commands to the host based on the release operation of releasing the second queue from the command queue circuit, and to subsequently receive the additional commands from the host, wherein the command queue circuit is configured to further store the additional commands received from the host.

[0005] According to another aspect of this disclosure, a non-volatile storage device is provided, the non-volatile storage device comprising: a memory cell array including a plurality of planes, each of the plurality of planes including a plurality of memory cells; input / output circuitry configured to sequentially receive from a controller a first queuing command corresponding to a first queue for a first physical address, a second queuing command corresponding to a second queue for a second physical address, and a merge command corresponding to the first queue and the second queue; and control logic circuitry including internal queueing circuitry configured to store the first command and the second command, wherein the first command corresponds to the first queuing command and the second command corresponds to the second queuing command. The control logic circuitry is configured to control the memory cell array in response to the merge command, such that the non-volatile storage device concurrently: performs a first storage operation corresponding to the first command at the first physical address; and performs a second storage operation corresponding to the second command at the second physical address. Attached Figure Description

[0006] The embodiments will be more clearly understood from the following detailed description taken in conjunction with the accompanying drawings, in which: Figure 1 This is a block diagram illustrating a storage device according to an embodiment; Figure 2 This is a block diagram illustrating a storage device including a first plane to a fourth plane according to an embodiment; Figure 3 This is a block diagram showing the storage device according to an embodiment in more detail; Figure 4 The command merging operation according to an embodiment is illustrated; Figure 5 An additional queue update operation according to an embodiment is shown; Figure 6 This is a timing diagram illustrating the command queuing operation and the merged execution operation according to an embodiment; Figure 7 This is a block diagram illustrating a storage device according to an embodiment; Figure 8 This is a block diagram illustrating the configuration of a memory controller according to an embodiment; Figure 9 The operation of the memory controller according to an embodiment is illustrated; Figure 10 The run queue and enqueue queue according to an embodiment are shown; Figure 11 A read operation performed using command queuing and merging is illustrated according to an embodiment; Figure 12 The illustration shows programming operations that utilize command queuing and merging for execution according to an embodiment; Figure 13 This is a flowchart illustrating the operation between the host and the storage device according to an embodiment; Figure 14 This is a flowchart illustrating the operation between the memory controller and the non-volatile memory according to an embodiment; Figure 15 This is a flowchart illustrating an operation method of a storage device according to an embodiment; and Figure 16 This is a block diagram illustrating a system according to an embodiment. Detailed Implementation

[0007] This disclosure provides storage devices, controllers, and non-volatile storage devices with enhanced performance through command queuing operation.

[0008] In the following description, embodiments are illustrated in detail with reference to the accompanying drawings. In the drawings, the same reference numerals are given to the same elements, and repeated descriptions thereof are omitted.

[0009] Figure 1 This is a block diagram illustrating a storage device 10 according to an embodiment.

[0010] refer to Figure 1 The storage device 10 may include a controller 100 and a non-volatile memory 200. In response to a request or command received from the host 20, the controller 100 may control the non-volatile memory 200 such that data stored in the non-volatile memory 200 is read or data is programmed into the non-volatile memory 200. According to an embodiment, the controller 100 may be referred to as a memory controller or storage controller. According to an embodiment, the storage device 10 may be referred to as a storage system.

[0011] Controller 100 may include command queuing circuitry 120, which may enqueue or store commands received sequentially from host 20. Non-volatile memory 200 may include internal queuing circuitry 231, which may enqueue or store commands received sequentially from controller 100. Controller 100 may utilize the command queuing function of non-volatile memory 200 to merge multiple commands stored in command queuing circuitry 120, and remove or release merged commands from command queuing circuitry 120.

[0012] Therefore, the lifetime of each queue in the command queue circuit 120 can be reduced, and the number of queues operating concurrently per unit time can be increased. As a result, memory bank conflicts in the non-volatile memory 200 can be reduced, and the performance of the storage device 10 can be improved. The configuration and operation of the controller 100 and the configuration and operation of the non-volatile memory 200 are described in detail below.

[0013] Controller 100 may include host interface circuitry 110, command queue circuitry 120, queue manager circuitry 130, and non-volatile memory interface circuitry 140. Host interface circuitry 110 may sequentially receive multiple requests or commands from host 20. For example, host interface circuitry 110 may send data packets to and receive data packets from host 20. Data packets sent from host 20 to host interface circuitry 110 may include commands or data to be written to non-volatile memory 200, and data packets sent from host interface circuitry 110 to host 20 may include responses to commands or data read from non-volatile memory 200. Example commands include memory commands such as programming, reading, writing, and erasing operations.

[0014] Command queue circuit 120 can receive multiple commands from host interface circuit 110 and store multiple queues corresponding to the received commands. Queue manager circuit 130 can select at least two queues based on the physical addresses of the multiple queues stored in command queue circuit 120, and can merge the selected at least two queues. In addition, queue manager circuit 130 can release the merged queue from command queue circuit 120 based on the merge operation, and thus, command queue resources can be secured.

[0015] In embodiments, queue manager circuitry 130 can merge queues having physical addresses corresponding to the same memory bank. In some embodiments, a memory bank can correspond to a chip, die, logical unit number (LUN), or plane. For example, each chip may include one or more dies, each die may include one or more LUNs, and each LUN may include one or more planes. Herein, LUN can refer to the smallest logical unit that controller 100 can independently execute commands on. For example, controller 100 can schedule programming operations, read operations, or erase operations on a per-LUN basis.

[0016] In an embodiment, queue manager circuit 130 can select queues with target addresses to be operated on concurrently, and can merge at least one selected queue into a representative queue. For example, queue manager circuit 130 can select a first queue corresponding to a first physical address and a second queue corresponding to a second physical address from command queue circuit 120, and can merge the second queue into the first queue. Additionally, queue manager circuit 130 can release the second queue from command queue circuit 120 based on the merge operation of merging the second queue into the first queue, and thus, command queue resources can be secured. As used herein, the first queue may be referred to as the "representative queue," and the second queue may be referred to as the "merged queue."

[0017] The non-volatile memory interface circuit 140 can send data to be programmed into the non-volatile memory 200, or receive data read from the non-volatile memory 200. For example, the non-volatile memory interface circuit 140 can send commands CMD and addresses ADDR to the non-volatile memory 200 to indicate storage operations such as read, program, or erase operations, and can send and receive data DATA from the non-volatile memory 200. The non-volatile memory interface circuit 140 can be implemented to comply with standards such as Toggle and Open NAND Flash Interface (ONFI).

[0018] In an embodiment, the non-volatile memory interface circuit 140 can sequentially send a first queuing command corresponding to a first queue, a second queuing command corresponding to a second queue, and a queue execution command or merge command corresponding to the first and second queues to the non-volatile memory 200. For example, the queue execution command or merge command can be sent after the first and second queuing commands are sent. In an embodiment, the host interface circuit 110 can send a request for additional commands to the host 20 based on a release operation that releases the second queue from the command queue circuit 120.

[0019] The non-volatile memory 200 may include input / output circuitry 210, a memory cell array 220, and control logic circuitry 230. The control logic circuitry 230 may include an internal queue circuitry 231, which may store commands received sequentially from the controller 100. However, this disclosure is not limited thereto, and in some embodiments, the internal queue circuitry 231 may be located outside the control logic circuitry 230. Additionally, in some embodiments, the internal queue circuitry 231 may be included within the input / output circuitry 210.

[0020] The input / output circuit 210 can sequentially receive from the controller 100 a first queuing command corresponding to a first queue for a first physical address, a second queuing command corresponding to a second queue for a second physical address, and a queue execution command or merge command corresponding to the first queue and the second queue. The memory cell array 220 may include multiple planes PL, and each of the multiple planes PL may include multiple memory cells.

[0021] Internal queuing circuitry 231 can enqueue or store multiple commands received sequentially from controller 100. According to an embodiment, internal queuing circuitry 231 can be referred to as a command queue or on-chip queue. In an embodiment, internal queuing circuitry 231 can enqueue or store a first command corresponding to a first queuing command and a second command corresponding to a second queuing command. In response to a queue execution command or a merge command, control logic circuitry 230 can control memory array 220 to concurrently execute a first storage operation and a second storage operation corresponding to the first command and the second command, respectively, at a first physical address and a second physical address.

[0022] However, this disclosure is not limited to this, and even when no merge command is received, the control logic circuit 230 can control the memory cell array 220 to perform a first storage operation and a second storage operation corresponding to the first command and the second command, respectively, on the first physical address and the second physical address. For example, the control logic circuit 230 may include a queue manager circuit, which can merge commands enqueued in the internal queue circuit 231. Therefore, the control logic circuit 230 can initiate operations corresponding to commands enqueued in the internal queue circuit 231. Additionally, for example, the control logic circuit 230 can periodically perform queue execution operations. For example, the control logic circuit 230 can periodically perform storage operations corresponding to commands enqueued in the internal queue circuit 231.

[0023] In one embodiment, the first command and the second command correspond to read commands, and the control logic circuit 230 can concurrently perform read operations on the first physical address and the second physical address in response to a queue execution command or a merge command. In another embodiment, the first command and the second command correspond to programming commands, and the control logic circuit 230 can concurrently perform programming operations on the first physical address and the second physical address in response to a queue execution command or a merge command. In yet another embodiment, the first physical address and the second physical address may correspond to the same plane, the same block, and the same page.

[0024] In an embodiment, control logic circuitry 230 may activate a queue mode or a queuing mode based on a setting feature command received from controller 100. In this specification, the terms "queue mode" and "queuing mode" have substantially the same meaning and can be used interchangeably. For example, a setting feature command may include a setting value for enabling command queuing functionality of control logic circuitry 230. In an embodiment, control logic circuitry 230 may receive a separate command from controller 100 indicating a queue mode.

[0025] In this embodiment, when the queue mode is activated, the control logic circuit 230 can store user commands received from the controller 100 into the internal queue circuit 231 without immediately executing the storage operation corresponding to the user command. Therefore, the user command can serve as a queuing command. In this embodiment, based on a queue execution command or a merge command received from the controller 100, the control logic circuit 230 can merge commands enqueued in the internal queue circuit 231 and perform a storage operation according to the merge command. In this embodiment, the controller 100 can periodically issue queue execution commands and periodically send queue execution commands to the non-volatile memory 200.

[0026] In the non-volatile memory 200, each memory bank can process only one command at a time. Therefore, memory bank conflicts may occur when different commands concurrently request the same memory bank. When a memory bank conflict occurs, subsequent commands must wait until the earlier command completes. As a result, command latency increases, and overall operation latency becomes larger. This accumulation of latency can be a major cause of limiting the input / output per second (IOPS) and throughput of the storage device 10.

[0027] However, the controller 100 according to this disclosure can sequentially send queuing commands to enqueue commands as physical operation units during busy periods of the non-volatile memory 200, and can send merging commands to instruct the execution of enqueued commands during ready periods of the non-volatile memory 200, thereby processing multiple commands as a single execution unit. In this case, commands entered at different times can be grouped together and executed as a single memory unit, thereby preventing access to the same memory from being scattered in time. That is, because commands for the same memory can be processed in a centralized execution manner through command queuing and merging operations, memory conflicts that occur due to commands arriving at scattered times can be effectively mitigated. As a result, unnecessary waiting time caused by memory conflicts is significantly reduced, and the effective utilization time of the non-volatile memory 200 is increased, thereby improving overall IOPS and throughput.

[0028] Figure 2 This is a block diagram illustrating a non-volatile memory 200 including a first plane P0 to a fourth plane P3 according to an embodiment.

[0029] Let's refer to each other. Figure 1 and Figure 2 The storage cell array 220 may include a first plane P0 to a fourth plane P3. For example, the first plane P0 to the fourth plane P3 may correspond to the first storage array block MAT0 to the fourth storage array block MAT3, respectively. However, this disclosure is not limited thereto, and the number of planes included in the storage cell array 220 may vary depending on the embodiment.

[0030] The control logic circuit 230 can sequentially receive multiple commands from the controller 100, including a first command C1 and a second command C2. For example, the first command C1 may be a read command or a programming command corresponding to the fourth plane P3. For example, the second command C2 may be a read command or a programming command corresponding to the fourth plane P3.

[0031] Internal queue circuit 231 can store a first command C1 in a first queue Qa in response to a first queuing command, and store a second command C2 in a second queue Qb in response to a second queuing command. In response to a queue execution command or a merge command, control logic circuit 230 can merge the first queue Qa and the second queue Qb and concurrently execute the storage operation corresponding to the fourth plane P3.

[0032] Figure 3 This is a block diagram showing the non-volatile memory 200a according to an embodiment in more detail.

[0033] refer to Figure 3The non-volatile memory 200a may include input / output circuitry 210, a memory cell array 220a, control logic circuitry 230, a row decoder 240, and a page buffer circuitry 250. For example, the non-volatile memory 200a may correspond to... Figure 1 and Figure 2 An example of a non-volatile memory 200 is shown. See below for reference. Figure 1 and Figure 3 The components of the non-volatile memory 200a are described. The input / output circuit 210 can provide the control logic circuit 230 with the command CMD and address ADDR received from the controller 100, and can provide the page buffer circuit 250 with the data DATA.

[0034] In one embodiment, the command CMD, address ADDR, and data DATA can be transmitted via a data signal DQ, and the data signal DQ can be synchronized with a data strobe signal and transmitted via a data signal line. However, this disclosure is not limited thereto, and in some embodiments, the controller 100 and the non-volatile memory 200a can communicate using a separated command address (SCA) method in which the command CMD and address ADDR are transmitted separately from the data DATA. For example, the command CMD and address ADDR can be synchronized with a command / address clock signal and transmitted via a command / address signal line, and the data DATA can be synchronized with a data strobe signal and transmitted via a data signal line.

[0035] For example, memory cell array 220a may include a fourth plane P3, and the fourth plane P3 may include multiple memory blocks BLK1 to BLKn, where n is a natural number. Each memory block (e.g., memory block BLK1) may include multiple pages PG1 to PGm, where m is a natural number. For example, a memory block may be a unit of erasure, and a page may be a unit of programming and reading. Memory cell array 220a may be connected to row decoder 240 via word line WL, serial select line SSL, and ground select line GSL, and may be connected to page buffer circuitry 250 via bit line BL. The description of the fourth plane P3 can also be applied to the first planes P0 to the third planes P2.

[0036] For example, the memory cells included in the memory cell array 220a may be NAND flash memory cells. However, this disclosure is not limited thereto, and in some embodiments, the memory cells may also include magnetic RAM (MRAM), spin-transfer torque MRAM (STT-MRAM), conductive bridged RAM (CBRAM), ferroelectric RAM (FeRAM), phase RAM (PRAM), resistive RAM (RRAM), and various other types of memory.

[0037] In some embodiments, the memory cell array 220a may include a three-dimensional (3D) memory cell array, and the 3D memory cell array may include a plurality of NAND strings. Each NAND string may include memory cells respectively connected to word lines vertically stacked on a substrate. The disclosures of U.S. Patent Nos. 7,679,133, 8,553,466, 8,654,587, and 8,559,235 and U.S. Application Publication No. 2011 / 0233648 are incorporated herein by reference in their entirety. In some embodiments, the memory cell array 220a may include a two-dimensional (2D) memory cell array, and the 2D memory cell array may include a plurality of NAND strings arranged in both row and column directions.

[0038] Control logic circuitry 230 can control all various operations in non-volatile memory 200a. Specifically, control logic circuitry 230 can be configured to control storage operations for a plurality of storage cells included in storage cell array 220a. For example, storage operations may include at least one of programming operations, reading operations, and erasing operations. In some embodiments, control logic circuitry 230 may receive command CMD and address ADDR from input / output circuitry 210, and in response to the received command CMD and address ADDR, may output various control signals, such as row address X_ADDR and column address Y_ADDR.

[0039] In this embodiment, the control logic circuit 230 can perform on-chip command queuing, thereby enhancing the performance of the non-volatile memory 200a. Specifically, the control logic circuit 230 can enqueue commands in response to a queuing command, and can concurrently execute storage operations corresponding to multiple enqueue commands in response to a queue execution command or a merge command. Therefore, even in random workload environments, the utilization of overall planar resources can be increased, thereby improving the core throughput of the non-volatile memory 200a.

[0040] In response to the row address X_ADDR, the row decoder 240 can select one word line from multiple word lines WL and one string select line from multiple string select lines SSL. The page buffer circuit 250 may include multiple page buffers, each connected to a memory cell via multiple bit lines BL. The page buffer circuit 250 can select at least one bit line from the multiple bit lines BL in response to the column address Y_ADDR. The page buffer circuit 250 can operate as a write driver or a sense amplifier depending on the operating mode.

[0041] Figure 4 The command merging operation according to an embodiment is illustrated.

[0042] Let's refer to each other. Figure 1 and Figure 4The queue manager circuit 130 can perform a command merge operation to merge some queues from multiple queues stored in the command queue circuit 120, thereby generating a merged resource 40b from the original resource 40a. For example, when the maximum queue depth of the command queue circuit 120 is 8, the command queue circuit 120 can store up to 8 queues. The maximum queue depth of the command queue circuit 120 can vary depending on the embodiment. The queue manager circuit 130 can perform a command merge operation to release some queues from the command queue circuit 120, thereby securing command queue resources. For example, when four queues are released from the command queue circuit 120, four command queue resources can be secured in the command queue circuit 120. The command merge operation is described in detail below.

[0043] Command queuing circuit 120 can store multiple queues 41 to 48 based on multiple commands received from host 20, and the multiple queues 41 to 48 can constitute the original resource 40a. In an embodiment, command queuing circuit 120 can store the physical address PA corresponding to the queue ID QID and validity information VI. For example, the physical address PA may include a physical page number (PPN), which includes plane information, block information, page information, and sector information corresponding to the queue ID QID. For example, validity information VI may include a validity flag indicating whether the corresponding queue is valid.

[0044] Queue manager circuit 130 can select at least two queues based on the physical addresses PA of multiple queues 41 to 48 stored in command queue circuit 120, and generate merged resource 40b by merging the selected at least two queues. For example, queue manager circuit 130 can select queues 41 and 45, and merge queue 45 into queue 41 to generate representative queue 41R. For example, queue manager circuit 130 can select queues 42 and 47, and merge queue 47 into queue 42 to generate representative queue 42R. For example, queue manager circuit 130 can select queues 44, 46, and 48, and merge queues 46 and 48 into queue 44 to generate representative queue 44R.

[0045] For example, the physical address corresponding to queue 41 with QID1 could be plane 0, BLK1, PG1, and S0, and the physical address corresponding to queue 45 with QID3 could be plane 0, BLK1, PG1, and S5. In this case, the physical addresses corresponding to queues 41 and 45 share the same plane, the same block, and the same page, and therefore, concurrent operations are possible. Thus, queue manager circuit 130 can select queues 41 and 45 and merge queue 45 into queue 41 to generate a representative queue 41R. In this case, queue manager circuit 130 can release the merged queue 45 from command queue circuit 120, and therefore, command queue resource 45E can be secured. The valid flag for the secured command queue resource 45E can be set to 0.

[0046] For example, the physical address corresponding to queue 42 with QID2 could be plane 1, BLK2, PG2, and S2, and the physical address corresponding to queue 47 with QID7 could be plane 1, BLK2, PG2, and S0. In this case, the physical addresses corresponding to queues 42 and 47 share the same plane, the same block, and the same page, and therefore, concurrent operations are possible. Thus, queue manager circuit 130 can select queues 42 and 47 and merge queue 47 into queue 42 to generate a representative queue 42R. In this case, queue manager circuit 130 can release the merged queue 47 from command queue circuit 120, and therefore, command queue resource 47E can be secured. The valid flag for the secured command queue resource 47E can be set to 0.

[0047] For example, the physical address corresponding to queue 44 with QID0 could be plane 3, BLK4, PG4, and S5; the physical address corresponding to queue 46 with QID5 could be plane 3, BLK4, PG4, and S1; and the physical address corresponding to queue 48 with QID6 could be plane 3, BLK4, PG4, and S7. In this case, the physical addresses corresponding to queues 44, 46, and 48 share the same plane, the same block, and the same page, and therefore, concurrent operations are possible. Thus, queue manager circuit 130 can select queues 44, 46, and 48 and merge queues 46 and 48 into queue 44 to generate a representative queue 44R. In this case, queue manager circuit 130 can release the merged queues 46 and 48 from command queue circuit 120, and therefore, command queue resources 46E and 48E can be secured. The valid flags for the secured command queue resources 46E and 48E can be set to 0.

[0048] Queue manager circuit 130 can update or generate a merge flag MF based on a command merge operation and store the merge flag MF in command queue circuit 120. In queue manager circuit 130, the merge flag MF of each of the representative queues 41R, 42R, and 44R to which a command merge operation has been performed can be set to a first value (e.g., 1). Therefore, when the merge flag MF is set to 1, each of the representative queues 41R, 42R, and 44R can include at least one merge queue ID QID. For example, representative queue 41R can include Q0 with QID0, representative queue 42R can include Q7 with QID7, and representative queue 44R can include Q5 with QID5 and Q6 with QID6.

[0049] Additionally, in the queue manager circuit 130, the merge flag MF of queue 43 (which is a regular queue to which a command merge operation has not yet been performed) can be set to a second value (e.g., 0) that is different from the first value. Therefore, when the merge flag MF is 0, queue 43 (i.e., the general queue) can store the physical address PA corresponding to queue ID QID. For example, queue 43 may include planes 1, BLK3, PG3, and S1 as physical addresses corresponding to QID4.

[0050] Figure 5 An additional queue update operation according to an embodiment is illustrated.

[0051] Let's refer to each other. Figure 1 and Figure 5 The host interface circuit 110 can generate a queue entry table 50a based on multiple requests or commands received sequentially from the host 20. Specifically, the host interface circuit 110 can register the requests or commands received from the host 20 as queue entries and assign queue IDs (QIDs). For example, the queue entry table 50a can be implemented as a link risk queue. For example, the link list queue may include an index representing the queue entry number, a previous pointer indicating the previous destination in the link list, a next pointer indicating the next entry in the link list, and a queue ID (QID) representing the number of the command to which the corresponding entry belongs.

[0052] Host 20 can send multiple commands based on the maximum queue depth of command queue circuit 120, and host interface circuit 110 can generate a queue entry table 50a containing multiple queue entries Q_ENTRY0 to Q_ENTRY7 based on the multiple commands. When command queue resources are secured according to a command merging operation, controller 100 can request additional commands from host 20 and subsequently receive additional commands from host 20.

[0053] For example, host interface circuit 110 can update queue entry table 50a with additional queue entries Q_ENTRY8 and Q_ENTRY9 corresponding to additional commands received from host 20. Queue manager circuit 130 can store new queues 46N and 47N in the allocated command queue resources based on the additional queue entries Q_ENTRY8 and Q_ENTRY9 received from host interface circuit 110. Therefore, queue manager circuit 130 can generate updated resource 50b by updating command queue circuit 120 with additional queues.

[0054] Figure 6 This is a timing diagram illustrating the command queuing operation and the merged execution operation according to an embodiment.

[0055] Let's refer to each other. Figure 1 , Figure 4 and Figure 6 In the comparative example 60a of command queuing and merged execution operations without using non-volatile memory, the controller can send normal read commands only during the ready period of the non-volatile memory based on the ready / busy signal RnB from the non-volatile memory. For example, the controller can sequentially send normal read commands 61a corresponding to queue Q9, 61b corresponding to queue Q0, 61c corresponding to queue Q5, and 61d corresponding to queue Q6 to the non-volatile memory.

[0056] On the other hand, according to the embodiment, the storage device 10 can perform command queuing and merged execution operations of the non-volatile memory 200. Specifically, the controller 100 can determine that storage operations corresponding to queues Q0, Q5, and Q6 can be executed concurrently based on the fact that the physical addresses of queues Q0, Q5, and Q6 correspond to the same plane, the same block, and the same page. Therefore, the controller 100 can sequentially send a first queuing command 62a corresponding to queue Q0, a second queuing command 62b corresponding to queue Q5, and a third queuing command 62c corresponding to queue Q6 to the non-volatile memory 200. In this case, the controller 100 can send the first queuing command 62a to the third queuing command 62c to the non-volatile memory 200 during busy periods.

[0057] The non-volatile memory 200 can store queues Q0, Q5, and Q6 in the internal queue circuit 231 in response to first queuing commands 62a through third queuing commands 62c. For example, each page may include sectors one through eight. For example, each page may correspond to 4KB, and each sector may correspond to 512B. The non-volatile memory 200 can enqueue queue Q0, corresponding to the sixth sector S5 of the fourth plane P3, BLK4, and PG4, queue Q5, corresponding to the second sector S1 of the fourth plane P3, BLK4, and PG4, and queue Q6, corresponding to the eighth sector S7 of the fourth plane P3, BLK4, and PG4, in the internal queue circuit 231.

[0058] Subsequently, the controller 100 may send a merge command 63 corresponding to queues Q0, Q5, and Q6 to the non-volatile memory 200. In this case, the controller 100 may send the merge command 63 to the non-volatile memory 200 during its ready period. The merge command 63 may correspond to a queue execution command instructing the execution of a storage operation corresponding to queues Q0, Q5, and Q6. The non-volatile memory 200 may concurrently execute the storage operations corresponding to queues Q0, Q5, and Q6 in response to the merge command 63, and may release queues Q0, Q5, and Q6 from the internal queue circuit 231. For example, the non-volatile memory 200 may concurrently perform read operations on fourth planes P3, BLK4, and PG4.

[0059] Figure 7 This is a block diagram illustrating a storage device 70 according to an embodiment.

[0060] refer to Figure 7 The storage device 70 may include a storage device 200b and a controller 100. The storage device 70 may support multiple channels, including a first channel CH0 and a second channel CH1. The storage device 200b may include multiple non-volatile memories NVM11 to NVM2n (where n is a positive integer) and is therefore referred to as a "non-volatile storage device". In the storage device 200b, the non-volatile memories NVM11 to NVM1n may be connected to the first channel CH0 via paths WAY11 to WAY1n, and the non-volatile memories NVM21 to NVM2n may be connected to the second channel CH1 via paths WAY21 to WAY2n. For example, each of the multiple non-volatile memories NVM11 to NVM2n may be implemented as a chip or a die, but this disclosure is not limited thereto.

[0061] In embodiments, non-volatile memories NVM11 to NVM1n can be stacked vertically to form a first stack, and non-volatile memories NVM21 to NVM2n can be stacked vertically to form a second stack. In embodiments, the first and second stacks can be stacked on top of each other in a vertical direction. For example, the second stack can be disposed above the first stack and electrically connected to it via a through-silicon via (TSV) structure. However, this disclosure is not limited thereto, and in some embodiments, the first and second stacks can be arranged in a horizontal direction.

[0062] The controller 100 can select one non-volatile memory connected to each channel via each channel, and can send signals to and receive signals from the selected non-volatile memory. For example, the controller 100 can select non-volatile memory NVM11 from non-volatile memories NVM11 to NVM1n connected to the first channel CH0. The controller 100 can send command CMDa, address ADDRa, and data DATAa to the selected non-volatile memory NVM11 via the first channel CH0, or can receive data DATAa from the selected non-volatile memory NVM11.

[0063] Furthermore, the controller 100 can send signals to and receive signals from the storage device 200b in parallel via different channels. For example, the controller 100 can send command CMDb to the storage device 200b via the first channel CH0 while simultaneously sending command CMDb to the storage device 200b via the second channel CH1. Similarly, the controller 100 can receive data DATAa from the storage device 200b via the first channel CH0 while simultaneously receiving data DATAb from the storage device 200b via the second channel CH1.

[0064] Furthermore, the controller 100 can enhance the parallelism of data access to non-volatile memory connected to paths assigned to corresponding channels by using path interleaving technology. Specifically, the controller 100 can perform programming / reading / erasing operations on different paths in a time-overlapping manner, thereby enabling commands to be executed in parallel on one path while an internal operation on another path is in progress. Therefore, single-channel utilization efficiency is increased, and overall throughput and I / O performance are improved. For example, while a programming or reading operation is being performed on non-volatile memory NVM11 connected to path WAY11, the controller 100 can send command CMDa, address ADDRa, and data DATAa to non-volatile memory NVM12 connected to path WAY12 via the first channel CH0, or can receive data DATAa from non-volatile memory NVM12.

[0065] According to an embodiment, each of the plurality of non-volatile memories NVM11 to NVM2n may include internal queuing circuitry and can support command queuing operations by utilizing the internal queuing circuitry. Specifically, each of the plurality of non-volatile memories NVM11 to NVM2n can enqueue multiple commands into the internal queuing circuitry in response to a queuing command received from controller 100, and can concurrently execute storage operations corresponding to the enqueuing commands in response to a merge command received from controller 100. Controller 100 can merge at least two commands by utilizing the command queuing operations of the plurality of non-volatile memories NVM11 to NVM2n, and can remove merged commands from the command queue, thereby efficiently managing command queue resources.

[0066] Figure 8 This is a schematic block diagram illustrating a storage device 80 according to an embodiment.

[0067] refer to Figure 8 The storage device 80 may include a controller 100a and a plurality of non-volatile memories 200c. Each of the plurality of non-volatile memories 200c can support command queuing operations and may therefore be referred to as a queued non-volatile memory Q_NVM. The controller 100a may include a host interface layer 101 (hereinafter referred to as "HIL 101"), a flash translation layer 102 (hereinafter referred to as "FTL 102"), and a flash interface layer 103 (hereinafter referred to as "FIL 103"). The controller 100a corresponds to Figure 1 An example of controller 100 is provided, and host interface circuitry 110 may be included in HIL 101. Command queue circuitry 120 and queue manager circuitry 130 may be included in FTL 102, and non-volatile memory interface circuitry 140 may be included in FIL 103.

[0068] HIL 101 is the layer responsible for the interface between storage device 80 and the host. It receives commands, addresses, data, and control signals from the host and converts them into a format suitable for the internal processing architecture. HIL 101 can perform command reception, command parsing, queue management, and direct memory access (DMA) configuration according to host interface protocols (e.g., NVMe, UFS, PCIe, etc.), and can send programming / read operations requested by the host to FTL 102 and FIL 103. Furthermore, HIL 101 can manage internal command queue generation and resource allocation based on information such as the host's queue ID or priority, thereby ensuring high-performance input / output processing and parallelism.

[0069] FTL 102 can perform address mapping for maintaining and updating a mapping table that translates logical addresses to physical addresses, wear leveling for extending memory lifetime by preventing intensive use of specific blocks, collection of useless cells from free blocks by moving valid data and erasing blocks, and defective block management for managing defective blocks and allocating replacement blocks. In this embodiment, FTL 102 can maximize processing efficiency by appropriately rearranging host requests via queue scheduling. Specifically, FTL 102 can determine the optimal physical location based on command information received from HIL 101 according to an internal processing strategy, and then forward the request to FIL 103 to enable actual memory access.

[0070] FIL 103 is the layer that forwards memory access requests received from FTL 102 to the actual non-volatile memory 200c, and can control operations according to the detailed protocols and timing of the non-volatile memory 200c. Specifically, FIL 103 can generate and send commands to the non-volatile memory 200c, and control the timing of control signals such as Command Latch Enable (CLE), Address Latch Enable (ALE), Chip Enable Bar (CEB), Read Enable Bar (REB), Write Enable Bar (WEB), and DQS. As described above, FIL 103 is the lowest layer for directly managing the physical characteristics and timing constraints of the non-volatile memory 200c, and can execute requests from FTL 102 and HIL 101 at the actual hardware level.

[0071] Figure 9 The operation of the controller 100a according to an embodiment is shown.

[0072] Let's refer to each other. Figure 8 and Figure 9 HIL 101 can generate a queue entry table 50a based on multiple commands received from the host. For operations requiring retries, information can be reconstructed from the queue entry table 50a of HIL 101 based on the queue ID QID. FTL 102 can generate a raw resource 40a comprising multiple queues 41 to 48 based on queue entries Q_ENTRY0 to Q_ENTRY7 received from HIL 101. FTL 102 can select some queues based on the physical addresses of the multiple queues 41 to 48, merge at least one selected queue into a representative queue, and release the merged queue from the command queue, thereby ensuring command queue resources.

[0073] HIL 101 can provide the host with information about the secured command queue resources, and then receive additional commands from the host for updating queue entry table 50a with queue entries Q_ENTRY8 and Q_ENTRY9. FTL 102 can generate updated resource 50b by updating the command queues with new queues 46N and 47N based on queue entries Q_ENTRY8 and Q_ENTRY9 in the updated queue entry table 50a.

[0074] FTL 102 performs address mapping functions to translate logical addresses received from the host into physical addresses in the non-volatile memory 200c. For this purpose, FTL 102 may include an address mapping table storing mapping information for planes, blocks, pages, and sectors. Additionally, FTL 102 can analyze the logical address range corresponding to each command (i.e., each queue ID QID) to determine the physical location the command will access, and can provide the results to FIL 103.

[0075] FIL 103 can determine commands for a specific plane or block based on the physical address information provided by FTL 102 and send the commands to non-volatile memory 200c, thereby performing efficient storage operations in conjunction with FTL 102. In this embodiment, FIL 103 can generate a queue list 91a storing queue IDs QIDs corresponding to the first plane P0 to the fourth plane P3 based on the raw resource 40a generated by FTL 102. For example, FIL 103 can place the corresponding queue IDs in a first-in-first-out (FIFO) or linked list for each plane based on the physical addresses provided by FTL 102. For example, FIL 103 may include a scheduler that determines the queue IDs QIDs from which commands will be sent for each plane.

[0076] FIL 103 can determine which commands will be sent to the non-volatile memory 200c based on the ready / busy signals received from the non-volatile memory 200c. Specifically, when the non-volatile memory 200c is busy, FIL 103 can send a queuing command corresponding to the queue to the non-volatile memory 200c and enqueue the command in the internal queue circuit 231. Subsequently, when the non-volatile memory 200c is put into a ready state, FIL 103 can send a queue execution command or a merge command to the non-volatile memory 200c, thereby efficiently initiating the storage operation corresponding to the queue enqueued in the non-volatile memory 200c.

[0077] The queue IDs (QIDs) corresponding to the first plane P0 can be stored in the order of QID1 (corresponding to queue 41) and QID3 (corresponding to queue 44). The queue IDs (QIDs) corresponding to the second plane P1 can be stored in the order of QID2 (corresponding to queue 42), QID4 (corresponding to queue 43), and QID7 (corresponding to queue 47). The queue IDs (QIDs) corresponding to the fourth plane P3 can be stored in the order of QID0 (corresponding to queue 44), QID5 (corresponding to queue 46), and QID6 (corresponding to queue 48). In this case, because there are no queues corresponding to the third plane P2 among the first planes P0 to the fourth plane P3, the number of active planes is only three. Although there are a total of four planes, the number of active planes is limited to three, which may lead to a deterioration in overall IOPS performance.

[0078] However, according to an embodiment, HIL 101 can receive additional commands from the host, and FTL 102 can assign the physical address corresponding to the additional command to the third plane P2. For example, FTL 102 can assign the physical address of new queue 46N to the third planes P2, BLK5, PG3, and S0, and assign the physical address of new queue 47N to the third planes P2, BLK3, PG6, and S2. Based on the updated resource 50b in FTL 102, FIL 103 can store the queue IDs (QIDs) corresponding to the third plane P2 in the order of QID8 corresponding to new queue 46N and QID9 corresponding to new queue 47N. Therefore, since all four planes (P0 to P3) can be activated, the number of active planes increases to four, thereby enhancing overall IOPS performance.

[0079] Figure 10 The run queue and enqueue queue according to an embodiment are shown.

[0080] Let's refer to each other. Figure 7 and Figure 10The controller 100 can distinguish between run queues and enqueue queues for the non-volatile memories NVM11 to NVM2n in the storage device 200b, and can merge enqueue queues. For example, four queues can be assigned to the first plane of LUN0 connected to the first channel CH0. For example, Q(4) is a queue EXE that is running in the non-volatile memory, and Q(1), Q(557), and Q(613) can be queues that are enqueued in the non-volatile memory. In this case, the controller 100 can merge Q(1), Q(557), and Q(613) to generate a merged queue mQ. For example, the merged queue mQ can be a queue in which Q(557) and Q(613) are merged into Q(1). Therefore, the number of queues assigned to the first plane of LUN0 connected to the first channel CH0 can be reduced to two, and the controller 100 can ensure two command queue resources.

[0081] Figure 11 A read operation performed using command queuing and merging is illustrated according to an embodiment.

[0082] Let's refer to each other. Figure 1 and Figure 11 In comparative example 110a according to this disclosure, sequential read operations can be performed on the non-volatile memory, and the corresponding queue can be released from the command queue resource after each read operation is completed. For example, when the controller sequentially sends normal read commands 111a to 111c corresponding to queues Q1, Q2, and Q3 respectively for each ready period in the non-volatile memory, the lifetimes of queues Q2 and Q3 may unnecessarily increase. Furthermore, in this configuration, the number of queues that can be operated concurrently per unit time is limited to only one, resulting in limited parallelism.

[0083] However, in embodiment 110b according to this disclosure, the controller 100 can sequentially send queuing commands 112a to 112c corresponding to queues Q1, Q2, and Q3, respectively, to the non-volatile memory 200 during busy periods. Subsequently, the controller 100 can send merge commands 113 corresponding to queues Q1, Q2, and Q3 to the non-volatile memory 200 during ready periods. The non-volatile memory 200 can concurrently perform storage operations corresponding to queues Q1, Q2, and Q3 in response to the merge command 113. Therefore, the controller 100 can immediately release queues Q2 and Q3 from the command queuing circuit 120, thereby significantly shortening the lifetime of queues Q2 and Q3 in the command queuing circuit 120. Furthermore, in this configuration, the number of queues that can be operated concurrently per unit time is increased to three, thereby effectively expanding the active memory.

[0084] Figure 12 The illustration shows a programming operation that utilizes command queuing and merging for execution according to an embodiment.

[0085] Let's refer to each other. Figure 1 and Figure 12 In comparative example 120a according to this disclosure, sequential programming operations can be performed on the non-volatile memory, and the corresponding queue can be released from the command queue resource after each programming operation is completed. However, in embodiment 120b according to this disclosure, the controller 100 can sequentially send queuing commands 121 and 122 corresponding to queues Q1 and Q2, respectively, to the non-volatile memory 200 during busy periods. For example, queuing command 121 may correspond to a set of commands including a first command 121a, data input, and a second command 121b. For example, queuing command 122 may correspond to a set of commands including a first command 122a, data input, and a second command 122b.

[0086] Subsequently, during the ready period of the non-volatile memory 200, the controller 100 may send a merge command or an execution command 123 corresponding to queues Q1 and Q2 to the non-volatile memory 200. The non-volatile memory 200 may then perform a storage operation corresponding to queues Q1 and Q2 in response to the merge command or execution command 123. Therefore, the controller 100 can immediately release queue Q2 from the command queue circuit 120, thereby significantly shortening the lifetime of queue Q2 in the command queue circuit 120.

[0087] Figure 13 This is a flowchart illustrating the operation between the host 20 and the storage device 10 according to an embodiment.

[0088] refer to Figure 13 In operation S110, an initialization operation is performed for the host 20 and the storage device 10. During the initialization process when the initialization operation is performed, the maximum queue depth of the command queue of the storage device 10 can be identified. For example, during the initialization process, the host 20 can request the maximum queue depth information of the command queue circuit 120 from the storage device 10, and in response, the storage device 10 can send the maximum queue depth information to the host 20. For example, the maximum queue depth can correspond to 8 or 16, and the maximum queue depth can vary depending on the embodiment.

[0089] In operation S120, host 20 sends a first request to storage device 10. In operation S130, storage device 10 enqueues a first command CMD1 corresponding to the first request into command queue circuit 120. In operation S140, host 20 sends a second request to storage device 10. In operation S150, storage device 10 enqueues a second command CMD2 corresponding to the second request into command queue circuit 120. Operations S120 to S150 can be repeated. Host 20 can sequentially send multiple requests to storage device 10 according to the maximum queue depth of command queue circuit 120.

[0090] In operation S160, storage device 10 merges the second command CMD2 into the first command CMD1 based on the physical addresses corresponding to the first command CMD1 and the second command CMD2. For example, storage device 10 can utilize the command queuing and merging function of non-volatile memory 200 to select the first command CMD1 and the second command CMD2 with concurrently operable physical addresses. For example, storage device 10 can set the queue corresponding to the first command CMD1 as a representative queue and merge the queue corresponding to the second command CMD2 into the representative queue.

[0091] In operation S170, storage device 10 releases the second command CMD2 from command queue circuit 120. For example, queue manager circuit 130 can shorten the lifetime of the queue corresponding to the second command CMD2 in command queue circuit 120 by releasing the second command CMD2 corresponding to the queue merged into the representative queue. Therefore, queue manager circuit 130 can ensure command queue resources by releasing the second command CMD2 from command queue circuit 120.

[0092] In operation S180, storage device 10 sends a response including queuing resource information to host 20. For example, storage device 10 may send a response to host 20 including the number of command queue resources secured. In operation S185, host 20 sends a new request to storage device 10. For example, host 20 may send additional commands to storage device 10 in a number equal to the number of command queue resources received from storage device 10. In operation S190, storage device 10 enqueues a new command corresponding to the new request into command queue circuit 120. Therefore, storage device 10 can efficiently manage command queue resources by merging commands enqueued in command queue circuit 120 and removing merged commands.

[0093] Figure 14 This is a flowchart illustrating the operation between the controller 100 and the non-volatile memory 200 according to an embodiment.

[0094] refer to Figure 14In operation S210, controller 100 sends a first queuing command corresponding to a first queue with respect to a first physical address to non-volatile memory 200. In operation S220, non-volatile memory 200 enqueues a first command CMD1 corresponding to the first queuing command into internal queue circuit 231. In operation S230, controller 100 sends a second queuing command corresponding to a second queue with respect to a second physical address to non-volatile memory 200. In operation S240, non-volatile memory 200 enqueues a second command CMD2 corresponding to the second queuing command into internal queue circuit 231.

[0095] Operations S210 to S240 can be performed during busy periods of the non-volatile memory 200. According to an embodiment, operations S210 to S240 can be performed repeatedly. For example, the first physical address and the second physical address can correspond to the same memory bank. For example, the memory bank can include a chip, a die, a LUN, or a plane. For example, the first physical address and the second physical address can correspond to the same plane, the same block, and the same page. For example, the first physical address and the second physical address can correspond to concurrently operable commands.

[0096] In operation S250, the controller 100 sends a merge command corresponding to the first queue and the second queue to the non-volatile memory 200. Here, the merge command may correspond to a queue execution command that instructs the execution of a storage operation corresponding to the first queue and the second queue. Operation S250 may be executed during the ready period of the non-volatile memory 200. In operation S260, in response to the merge command, the non-volatile memory 200 concurrently executes a first storage operation and a second storage operation corresponding to the first command and the second command at the first physical address and the second physical address, respectively.

[0097] In operation S270, the non-volatile memory 200 sends data or a response to the controller 100 based on the execution result of the storage operation. For example, when the storage operation is a read operation, the non-volatile memory 200 can send the read data to the controller 100. For example, when the storage operation is a programming operation, the non-volatile memory 200 can send a response to the controller 100 indicating that the operation is complete.

[0098] Figure 15 This is a flowchart illustrating a method of operating the storage device 10 according to an embodiment.

[0099] Let's refer to each other. Figure 1 and Figure 15In operation S300, storage device 10 sequentially receives multiple commands from host 20. In operation S310, storage device 10 determines whether a memory conflict has occurred. When a memory conflict is determined to have occurred, in operation S320, storage device 10 merges the first command and the second command by utilizing the command queuing and merging function of non-volatile memory 200.

[0100] In operation S330, storage device 10 secures command queue resources by releasing a second command from command queue circuit 120. In operation S340, storage device 10 sends the secured command queue resource information to host 20. In operation S350, storage device 10 receives an additional command queue from host 20.

[0101] On the other hand, when it is determined that no memory conflict has occurred, in operation S360, the storage device 10 executes a first storage operation corresponding to the first command. In operation S370, the storage device 10 sends a response indicating the storage operation is complete or data to the host 20, including data indicating that the storage operation is complete. In operation S380, the storage device 10 executes a second storage operation corresponding to the second command. In operation S390, the storage device 10 sends a response indicating the storage operation is complete or data to the host 20, including data indicating that the storage operation is complete. In operation S400, the storage device 10 receives a third command from the host 20.

[0102] Figure 16 This is a block diagram illustrating system 160 according to an embodiment.

[0103] refer to Figure 16 System 160 may include a processor 161 and a plurality of high-bandwidth memories 162a to 162c. In some embodiments, system 160 may correspond to a single package, and the processor 161 and the plurality of high-bandwidth memories 162a to 162c may be mounted on an interposer layer within the single package. In some embodiments, system 160 may correspond to a single module, and the packages, each including the processor 161 and the plurality of high-bandwidth memories 162a to 162c, may be mounted on a board.

[0104] Processor 161 can be any component that processes data. For example, processor 161 can be a processor for executing software code, fixed intellectual property (IP) (such as a hardware accelerator), or reconfigurable components (such as a field-programmable gate array (FPGA)). Alternatively, processor 161 can be a component designed for parallel processing, such as a graphics processing unit (GPU) and a neural processing unit (NPU).

[0105] Processor 161 can access a plurality of high-bandwidth memories 162a to 162c. Processor 161 can store data in the plurality of high-bandwidth memories 162a to 162c, or read data stored in the plurality of high-bandwidth memories 162a to 162c. In some embodiments, each of the plurality of high-bandwidth memories 162a to 162c may correspond to a reference memory. Figures 1 to 15 The described storage device or memory apparatus, and when flash memory is included, high-bandwidth memory may be referred to as high-bandwidth flash memory (HBF). The memory apparatus in the plurality of high-bandwidth memories 162a to 162c can not only have wide input / output but also provide multiple channels. Furthermore, the memory apparatus can support various modes. Therefore, the plurality of high-bandwidth memories 162a to 162c can provide high bandwidth and flexibility to the processor 161, thereby enhancing the performance of the system 160.

[0106] In some embodiments, the storage devices 10, 70, and 80 described above may include at least one of an SSD, embedded memory, and removable external memory. When the storage devices 10, 70, and 80 are SSDs, they may comply with the NVMe standard. When the storage devices 10, 70, and 80 are embedded memory or external memory, they may comply with Universal Flash Storage (UFS) or Embedded Multimedia Card (eMMC) standards.

[0107] In some embodiments, the storage devices 10, 70, and 80 described above may further include computing devices that perform data processing on data received from a host, thereby performing tasks assigned by the host. For example, the computing devices may perform various neural network-based tasks, such as convolutional neural networks (CNNs), region-with-convolutional-neural networks (R-CNNs), region proposal networks (RPNs), recurrent neural networks (RNNs), stacked deep neural networks (S-DNNs), state-space dynamic neural networks (S-SDNNs), deconvolutional networks, deep belief networks (DBNs), restricted Boltzmann machines (RBMs), fully convolutional networks, long short-term memory (LSTM) networks, and classification networks.

[0108] For example, the computing device can be implemented as a GPU, a digital signal processor (DSP), a field-programmable gate array (FPGA), an application-specific integrated circuit (ASIC), or an NPU. However, this disclosure is not limited thereto, and the computing device can be implemented as various types of accelerators for performing data processing (such as data operations (e.g., arithmetic operations, convolution operations, pooling operations, etc.)) required to perform the assigned task in parallel.

[0109] Although this disclosure has been specifically shown and described with reference to embodiments thereof, it will be understood that various changes in form and detail may be made therein without departing from the spirit and scope of this disclosure.

Claims

1. A storage device, the storage device comprising: Non-volatile memory; as well as A controller, configured to control the non-volatile memory, The controller includes: A command queue circuit, configured to store multiple queues corresponding to multiple commands received from a host. A queue manager circuit configured to, based on the physical addresses of the plurality of queues,: select a first queue corresponding to a first physical address and a second queue corresponding to a second physical address; and merge the second queue into the first queue; and A non-volatile memory interface circuit is configured to sequentially send to the non-volatile memory: a first queuing command corresponding to the first queue; a second queuing command corresponding to the second queue; and a merge command corresponding to the first queue and the second queue. The non-volatile memory is configured to concurrently execute a first storage operation corresponding to the first physical address and a second storage operation corresponding to the second physical address in response to the merge command.

2. The storage device according to claim 1, wherein, The queue manager circuit is configured to release the second queue from the command queue circuit based on a merge operation that merges the second queue into the first queue.

3. The storage device according to claim 2, wherein, The controller further includes a host interface circuit configured to send a request for additional commands to the host based on the controller releasing the second queue from the command queue circuit.

4. The storage device according to claim 1, wherein, The queue manager circuit is configured to set the first queue as a representative queue of the first queue and the second queue by merging the second queue into the first queue, and to set a merge flag corresponding to the representative queue to a first value.

5. The storage device according to claim 4, wherein, The representative queue having the merge flag set to the first value includes information about the second queue, and The plurality of queues also includes a general queue having a merge flag set to a second value different from the first value, and the general queue includes a physical address corresponding to the general queue.

6. The storage device according to claim 1, wherein, The non-volatile memory includes multiple memory banks, and The first physical address and the second physical address correspond to the first memory bank among the plurality of memory banks.

7. The storage device according to claim 6, wherein, The plurality of memory modules each include a plurality of planes, a plurality of dies, or a plurality of chips.

8. The storage device according to claim 1, wherein, The non-volatile memory includes internal queue circuitry, and The non-volatile memory is configured as follows: In response to the first queuing command, the first command corresponding to the first queue is stored in the internal queue circuit; In response to the second queuing command, a second command corresponding to the second queue is stored in the internal queue circuit; as well as In response to the merge command, the first storage operation corresponding to the first command and the second storage operation corresponding to the second command are executed concurrently.

9. The storage device according to claim 8, wherein, The first command and the second command correspond to read commands, and The non-volatile memory is configured to perform read operations on the first physical address and the second physical address in response to the merge command.

10. The storage device according to claim 8, wherein, The first command and the second command correspond to programming commands, and The non-volatile memory is configured to perform programming operations on the first physical address and the second physical address in response to the merge command.

11. The storage device according to claim 1, wherein, The first physical address and the second physical address correspond to the same plane, the same block, and the same page of the non-volatile memory.

12. A controller, the controller comprising: A command queue circuit, configured to store multiple queues corresponding to multiple commands received from a host. A queue manager circuit configured to: merge a first queue and a second queue based on the physical addresses of the plurality of queues, and set the first queue as a representative queue; and release the second queue from the command queue circuit; as well as A host interface circuit, configured to: send a request for an additional command to the host based on a release operation that releases the second queue from the command queue circuit; and subsequently receive the additional command from the host. The command queue circuit is configured to further store the additional commands received from the host.

13. The controller of claim 12, further comprising a non-volatile memory interface circuit configured to sequentially send to the non-volatile memory: The first queuing command corresponding to the first queue; The second queuing command corresponding to the second queue; as well as Merge command corresponding to the first queue and the second queue.

14. The controller according to claim 13, wherein, The first queue corresponds to a first physical address of the non-volatile memory, and the second queue corresponds to a second physical address of the non-volatile memory. The queue manager circuit is configured to merge the first queue and the second queue, and The first physical address and the second physical address correspond to the same plane, the same block, and the same page of the non-volatile memory.

15. The controller according to claim 12, wherein, The queue manager circuit is configured to set the first queue as a representative queue by merging the second queue into the first queue, and to set a merge flag corresponding to the representative queue to a first value.

16. The controller according to claim 15, wherein, The representative queue having the merge flag set to the first value includes information about the second queue, and The plurality of queues also includes a general queue having a merge flag set to a second value different from the first value, and the general queue includes a physical address corresponding to the general queue.

17. A non-volatile storage device, the non-volatile storage device comprising: A storage cell array, the storage cell array comprising multiple planes, each of the multiple planes comprising multiple storage cells; Input / output circuitry, configured to sequentially receive from the controller: The first queuing command corresponding to the first queue for the first physical address; The second queuing command corresponding to the second queue for the second physical address; as well as Merge commands corresponding to the first queue and the second queue; as well as The control logic circuit includes an internal queue circuit configured to store a first command and a second command, wherein the first command corresponds to a first queued command and the second command corresponds to a second queued command. The control logic circuit is configured to control the memory cell array in response to the merge command, such that the non-volatile memory device concurrently performs: a first storage operation corresponding to the first command on the first physical address; and a second storage operation corresponding to the second command on the second physical address.

18. The non-volatile storage device according to claim 17, wherein, The first command and the second command correspond to read commands, and The control logic circuit is configured to perform a read operation on the first physical address and the second physical address in response to the merge command.

19. The non-volatile storage device according to claim 17, wherein, The first command and the second command correspond to programming commands, and The control logic circuit is configured to perform programming operations on the first physical address and the second physical address in response to the merge command.

20. The non-volatile storage device according to claim 17, wherein, The first physical address and the second physical address correspond to the same plane, the same block, and the same page of the non-volatile memory device.

Citation Information

Patent Citations

  • Three-Dimensional Semiconductor Memory Devices And Methods Of Fabricating The Same

    US20110233648A1

  • Vertical-type non-volatile memory devices

    US7679133B2

  • Non-volatile memory device, erasing method thereof, and memory system including the same

    US8553466B2

  • Nonvolatile memory device, operating method thereof and memory system including the same

    US8559235B2

  • Nonvolatile memory devices, channel boosting methods thereof, programming methods thereof, and memory systems including the same

    US8654587B2