Tandem command and truncation command associated with computing flag capacity for high speed link (CXL) memory devices

By introducing concatenation and truncation commands into compute high-speed link (CXL) memory devices, the flexibility problem of tag capacity management is solved, enabling more efficient resource allocation and management and improving the performance of the memory system.

CN121116163APending Publication Date: 2025-12-12MICRON TECHNOLOGY INC
View PDF 0 Cites 0 Cited by

Patent Information

Application Number
CN202510741945.7
Authority / Receiving Office
CN · China
Patent Type
Applications(China)
Current Assignee / Owner
Priority Date
2025-05-19
Filing Date
2025-06-05
Publication Date
2025-12-12

AI Technical Summary

Technical Problem

In the prior art, the tag capacity management of compute high-speed link (CXL) memory devices lacks flexibility, and cannot effectively combine or divide stored data, resulting in insufficient flexibility in resource allocation.

Method used

The introduction of concatenation and truncation commands allows the host system to combine or split data in a tag capacity unit. The concatenation command stores data from multiple source tags into a concatenated destination tag, while the truncation command removes a portion of data from a source tag and stores it into a truncated destination tag.

Benefits of technology

It improves the flexibility of tag capacity, enhances the flexibility and efficiency of resource allocation, and supports higher-performance memory management.

✦ Generated by Eureka AI based on patent content.

Smart Images

  • Figure CN121116163A_ABST
    Figure CN121116163A_ABST
Patent Text Reader

Abstract

The application relates to a concatenation command and a truncation command associated with a flag capacity for computing a high speed link (CXL) memory device. A system may include a memory device including a plurality of dynamic capacity devices; and a processing device operably coupled with the memory device to perform operations including receiving a host command to combine first data associated with a first flag associated with a first memory segment of the plurality of dynamic capacity devices and second data associated with a second flag associated with a second memory segment of the plurality of dynamic capacity devices, wherein the second flag is associated with a second memory segment of the plurality of dynamic capacity devices, and wherein the first memory segment is allocated to a first host system to store the first data, and the second memory segment is allocated to the first host system to store the second data.
Need to check novelty before this filing date? Find Prior Art

Description

TECHNICAL FIELD

[0001] Embodiments of the present disclosure generally relate to memory subsystems and, more particularly, to implementing concatenation commands and truncation commands associated with tag capacity in compute express link (CXL) memory devices. BACKGROUND

[0002] A memory subsystem can include one or more memory devices that store data. The memory devices can be, for example, non-volatile memory devices and volatile memory devices. Generally, a host system can utilize a memory subsystem to store data at and retrieve data from the memory devices. SUMMARY

[0003] Embodiments of the present disclosure provide a system comprising: a memory device comprising a plurality of dynamic capacity devices; and a processing device operably coupled with the memory device to perform operations comprising: receiving a host command to combine first data associated with a first tag and second data associated with a second tag, wherein the first tag is associated with a first memory section of the plurality of dynamic capacity devices, wherein the second tag is associated with a second memory section of the plurality of dynamic capacity devices, and wherein the first memory section is allocated to a first host system to store the first data and the second memory section is allocated to the first host system to store the second data; and in response to receiving the host command to combine: determining a third memory section of the plurality of dynamic capacity devices and associating a third tag with the third memory section; combining the first data and the second data into combined data; and storing the combined data in the third memory section.

[0004] Another embodiment of the present disclosure provides a method comprising: receiving, by a processing device, a host command to combine first data associated with a first tag and second data associated with a second tag, wherein the first tag is associated with a first memory section of a plurality of dynamic capacity devices in a memory device, wherein the second tag is associated with a second memory section of the plurality of dynamic capacity devices, and wherein the first memory section is allocated to a first host system to store the first data and the second memory section is allocated to the first host system to store the second data; in response to receiving the host command to combine: determining a third memory section of the plurality of dynamic capacity devices and associating a third tag with the third memory section; combining the first data and the second data into combined data; and storing the combined data in the third memory section.

[0005] Yet another embodiment of the disclosure provides a non-transitory computer- readable storage medium comprising instructions that, when executed by a processing device, cause the processing device to perform operations comprising: receiving a host command to combine first data associated with a first tag and second data associated with a second tag, wherein the first tag is associated with a first memory section of a plurality of dynamic capacity devices in a memory device, wherein the second tag is associated with a second memory section of the plurality of dynamic capacity devices, and wherein the first memory section is allocated to a first host system to store the first data and the second memory section is allocated to the first host system to store the second data; and in response to receiving the host command to combine: determining a third memory section of the plurality of dynamic capacity devices and associating a third tag with the third memory section; combining the first data and the second data into combined data; and storing the combined data in the third memory section. BRIEF DESCRIPTION OF DRAWINGS

[0006] The disclosure will become more fully understood from the detailed description given herein and the accompanying drawings, which are meant to be non-limiting. The drawings are merely schematic representations not intended to portray specific embodiments of the work within a specific scope.

[0007] Figure 1 An example computing system including a memory sub-system according to some embodiments of the disclosure is described.

[0008] Figure 2 is a block diagram of an example system for using concatenation commands associated with tagged capacities in a compute express link (CXL) memory device according to some embodiments of the disclosure.

[0009] Figure 3A Examples of concatenation / truncation commands according to some embodiments of the disclosure are described.

[0010] Figure 3B Examples of tag mapping data structures according to some embodiments of the disclosure are described.

[0011] Figure 4 is a flowchart of an example method for using concatenation commands associated with tagged capacities in a compute express link (CXL) memory device according to some embodiments of the disclosure.

[0012] Figure 5 is a flowchart of an example method for using truncation commands associated with tagged capacities in a compute express link (CXL) memory device according to some embodiments of the disclosure.

[0013] Figure 6is a block diagram of an example computer system in which embodiments of the disclosure can operate. DETAILED DESCRIPTION

[0014] Aspects of the disclosure relate to implementing concatenation commands and truncation commands associated with tag capacity in a compute express link (CXL) memory device. The memory subsystem can be a storage device, a memory module, or a combination of a storage device and a memory module. Examples of storage devices and memory modules are described below in connection with Figure 1 Examples of storage devices and memory modules are described. Generally, a host system can utilize a memory subsystem including one or more components such as a storage device storing data. The host system can provide data to store at the memory subsystem and can request data to retrieve from the memory subsystem.

[0015] A compute express link (CXL) system is an optionally cache coherent interconnect for processors, memory expansion, and accelerators. The CXL system maintains memory coherency between CPU memory space and memory on attached devices, which allows resource sharing to enable higher performance, reduce software stack complexity, and lower overall system cost. Generally, CXL is an interface standard that can support multiple protocols that can run on top of PCIe, including a CXL.io protocol, a CXL.mem protocol, and a CXL.cache protocol. The CXL.io protocol is a PCIe-like protocol that can be viewed as an “enhanced” PCIe protocol capable of partitioning managed memory. CXL.io can be used for initialization, connection, device discovery and enumeration, register access, and can provide an interface for I / O devices. The CXL.mem protocol can enable a host to access memory of an attached device using memory semantics (e.g., load and store commands). This approach can support both volatile and persistent memory architectures. The CXL.cache protocol can define host device interactions to implement efficient caching of host memory with low latency using a request and response approach. Traffic (e.g., NVMe traffic) can run over the CXL.io protocol, and the CXL.mem and CXL.cache protocols can share a common link layer and transaction layer. Thus, CXL protocols can be multiplexed and transported via a PCIe physical layer.

[0016] Memory devices that support the CXL protocol and are attachable to a host via CXL are referred to as CXL memory devices, which can provide additional bandwidth and capacity to a host processor. The CXL memory devices are independent of the host memory. In some implementations, a CXL memory device can partition resources into multiple logical devices, and each logical device can be visible as a memory device. In some implementations, a CXL memory device can support multiple host systems. A fabric manager can configure resource allocation for the multiple host systems across the logical devices. Dynamic capacity (DC) is a feature of a CXL memory device that allows exposed memory capacity to be dynamically allocated and deallocated without requiring a reset of the CXL memory device. Although a CXL memory device is used here as an illustrative example for implementing dynamic capacity, the dynamic capacity feature can be applied to other memory devices.

[0017] In particular, a dynamic capacity device (DCD) is a memory device, such as a CXL memory device, that implements dynamic capacity (DC). The device physical address (DPA) range of the DCD is subdivided into a number of zones (e.g., 1 to 8 zones), and each of these zones can be subdivided into a set of blocks. One or more blocks that can be allocated to a host system and associated with a tag are referred to as a taggable DC unit. A taggable DC unit can represent a management unit that can be tagged, assigned at various capacity sizes, and dynamically allocated to various host systems. A taggable DC unit that has been assigned a tag is referred to as a tagged capacity unit. Each tag is globally unique, so the tags associated with a taggable DC unit can form an aggregate tag space in a memory device, such as a CXL memory device, and each tag in the aggregate tag space is uniquely identifiable. Each tag can be associated with one or more host systems, and can be mapped to one or more DPA ranges (e.g., one or more contiguous physical address ranges or a set of physical address extents list (i.e., non-contiguous address ranges) that identify respective locations on the DCD to store data). Each tag can be shareable or non-shareable.

[0018] In particular, the fabric manager controls the allocation of these taggable DC units to one or more host systems (or a set of host systems) and signals the host systems with events when the allocation of these taggable DC units changes. The fabric manager also assigns tags to the allocated taggable DC units by associating tags with taggable DC units represented by one or more physical addresses (e.g., one or more DPA ranges) in a tag mapping data structure. The memory device maps DPA ranges to taggable DC units. Thus, a tag can be referred to as representing a tag capacity unit. The host systems can map these DPA ranges to corresponding host physical address (HPA) ranges within a host address space available to the host systems. In some implementations, the memory device can communicate the status of these tag capacity units through a zone list that describes the starting DPAs and lengths of all blocks accessible to the host systems, where the zone list is managed by the memory device. The host systems (or a fabric manager on behalf of the host systems) can use a set of commands to query and configure the tag capacity units. The set of commands can include a command to allocate a new tag capacity unit (e.g., initiate a dynamic capacity add command), a command to release a tag capacity unit (e.g., initiate a dynamic capacity release command), and a command to get information of a tag capacity unit. The capacity of a shareable tag capacity unit associated with a tag and allocated to a host system is immutable such that additional capacity cannot be added to the tag nor can capacity be removed from the tag. That is, although the contents stored in a tag capacity unit can be modified, the mapping between a tag and the tag capacity unit allocated to a host system cannot be modified during the lifetime of the shareable tag. Thus, a host system needs to request a different capacity of a tag capacity unit or a different tag associated to be reallocated. Thus, although the contents stored in a tag capacity unit associated with a particular tag can be modified, the tag representing the tag capacity unit cannot be concatenated with other tag capacity units nor can it be truncated into one or more segments because the tag represents exactly one tag capacity unit and no more or less.

[0019] Aspects of the present disclosure address the above and other deficiencies by implementing concatenation commands and truncation commands associated with a tag capacity in a compute express link (CXL) memory device. A concatenation command associated with a tag capacity enables a host system to combine data stored in a set of tag capacity units, where each tag capacity unit is associated with a respective source tag, and store the combined data in another tag capacity unit with a concatenation destination tag (“data of the concatenation destination tag”). In some embodiments, the concatenation destination tag can thus act as an additive modification of a set of source tags, where the capacity of each source tag cannot be changed, and the mapping between the source tags and the tag capacity units allocated to the host system cannot be modified. The capacity size of the concatenation destination tag is the sum of the sizes of the set of source tags. Hereinafter, for ease of description, data stored in a tag capacity unit associated with a source tag can be referred to as data of the source tag.

[0020] The concatenation command can specify a set of source tags (i.e., two or more source tags) and specify a host system that can access (e.g., read, write, etc.) data of the concatenation destination tag. The concatenation command can further specify a selection policy that indicates a policy for selecting a taggable DC unit for the concatenation destination tag. The selection policy can include a “free and contiguous” policy that allows a memory controller to select a taggable DC unit that is available and contiguous, or a “prescriptive” policy that allows a fabric manager to select a taggable DC unit such that the selection is globally controlled.

[0021] The concatenation command can further specify an order of the set of source tags to be combined. In some embodiments, the concatenation command can provide a default order by using an order of the set of source tags provided in the concatenation command. In some embodiments, the concatenation command can allow a host system to provide a specific order of the set of source tags.

[0022] Thus, a host system that can access data of the concatenation destination tag (i.e., a concatenation destination host) can write and read data by accessing a specific copy of the combined data of the set of source tags (e.g., by referencing the concatenation destination tag). In some embodiments, one or more source tags can be maintained, and a host system that can access these source tags can still directly read and write data of these source tags. In some embodiments, one or more source tags can be deleted, and a host system that can access these source tags can be redirected to the concatenation destination tag for accessing the combined data of these source tags.

[0023] A truncate command associated with a source tag enables a host system to remove a portion of data from the data stored in the source tag, and store the removed data in another tag capacity unit having a truncate destination tag ("data of the truncate destination tag"). The portion of data can include data stored in one or more subzones of the tag capacity unit, where the tag capacity unit can include a plurality of subzones, and each subzone corresponds to a respective DPA range. Thus, removing the portion of data involves removing data stored in one or more DPA ranges. In some implementations, the truncate destination tag can thus act as a deletion modification of the source tag, where the capacity of the source tag cannot be changed, and the mapping between the source tag and the tag capacity unit assigned to the host system cannot be modified. The capacity size of the truncate destination tag is the total size of the subzones of the source tag minus the size of the removed subzones.

[0024] In one example, the source tag can correspond to X subzones, where a first subzone and a last subzone correspond to two end subzones, and (X-2) intermediate subzones fall in between the two ends. The truncate command can remove data stored in the end subzones and / or data stored in the intermediate subzones from the data of the source tag. The end subzones can correspond to DPA ranges from one end point to a truncate point of the tag capacity unit. The intermediate subzones can correspond to DPA ranges in the middle of the tag capacity unit. The removed data can be a combination of data stored in the remaining subzones.

[0025] Although one tag capacity unit is used as an example above, the subzones can be from multiple tag capacity units. In one example, a first source tag and a second source tag can be involved in a truncate command. The truncate command can remove data stored in a first subzone of the first source tag and data stored in a second subzone of the second source tag from the data stored in the first and second source tags, and the removed data can be a combination of data stored in the remaining subzones of the first and second source tags.

[0026] The truncate command can specify one or more source tags, and a host system that can access (e.g., read, write, etc.) data of a destination tag. The truncate command can further specify a selection policy that indicates a policy for selecting a taggable DC unit for the truncate destination tag. The selection policy can include a "free and contiguous" policy that allows a memory controller to select a taggable DC unit that is available and contiguous, or a "regulation" policy that allows a fabric manager to select a taggable DC unit such that the selection is globally controlled.

[0027] A truncate command can specify one or more sub-regions of a source tag. In some implementations, the specified regions represent sub-regions to be removed. In some implementations, the specified regions represent sub-regions to be kept, and in some cases, when all specified regions are sub-regions to be kept, the truncate command can become a concatenate command, such that the sub-regions to be kept will be combined.

[0028] A truncate command can further specify an order of the sub-regions to be combined. In some implementations, a truncate command can provide a default order by using the order of the sub-regions provided in the truncate command. In some implementations, a truncate command can allow a host system to provide a specific order of the sub-regions.

[0029] Thus, a host system that can access the data of the truncated destination tag (i.e., the truncate destination host) can write and read data by accessing specific copies of the combined data of the truncate sub-regions of the one or more source tags (e.g., by referencing the truncate destination tag). In some implementations, the one or more source tags can be kept, and a host system that can access these source tags can still read and write data directly to these source tags. In some implementations, the one or more source tags can be deleted, and a host system that can access these source tags can be redirected to the truncate destination tag for accessing the combined data of the truncate source tags.

[0030] Advantages of the present disclosure include efficiently modifying the tag capacity units allocated to a host system by using commands regarding tags. Specifically, the system significantly improves the flexibility of using tag capacity.

[0031] Figure 1 An example computing system 100 including a compute express link (CXL) memory device 110 is illustrated in accordance with some embodiments of the present disclosure. The CXL memory device 110 can include media, such as one or more volatile memory devices, one or more non-volatile memory devices, or such combinations.

[0032] The computing system 100 can be a computing device, such as a desktop computer, a laptop computer, a network server, a mobile device, a vehicle (e.g., an airplane, a drone, a train, a car, or other transportation vehicle), an internet-of-things (IoT) enabled device, an embedded computer (e.g., an embedded computer included in a vehicle, industrial equipment, or a networked commercial device), or such computing devices including memory and a processing device.

[0033] The computing system 100 can include one or more host systems 120 coupled to the CXL memory device 110. In some embodiments, the host systems 120 are coupled to multiple CXL memory devices 110 of different types. Figure 1An example of a host system 120 coupled to a CXL memory device 110 is illustrated. As used herein, “coupled to” or “coupled with” generally refers to a connection between components that can be an indirect communicative connection or direct communicative connection (e.g., without intervening components), whether wired or wireless, including connections such as electrical, optical, magnetic, etc.

[0034] The host system 120 can include a processor chipset and a software stack executed by the processor chipset. The processor chipset can include one or more cores, one or more caches, a memory controller (e.g., NVDIMM controller), and a storage protocol controller (e.g., PCIe controller, SATA controller). The host system 120 uses the CXL memory device 110, for example, to write data to the CXL memory device 110 and read data from the CXL memory device 110.

[0035] The host system 120 can be coupled to the CXL memory device 110 via a Peripheral Component Interconnect Express (PCIe) interface. The PCIe interface is a physical host interface used to transfer data between the host system 120 and the CXL memory device 110 for passing control, address, data, and other signals between the CXL memory device 110 and the host system 120. When the CXL memory device 110 is coupled with the host system 120 through the physical host interface (e.g., PCIe bus), the host system 120 can further utilize a NVM Express (NVMe) interface to access components of the CXL memory device 110. Figure 1 The CXL memory device 110 is illustrated as an example. In general, the host system 120 can access multiple CXL memory devices 110 via the same communication connection, multiple separate communication connections, and / or a combination of communication connections. The interface can be represented by a CXL interface.

[0036] In some embodiments, the host system 120 includes a central processing unit (CPU) 109 connected to a host memory 105, such as DRAM or other main memory. The host system 120 includes a bus 107, such as a memory device interface, that interacts with the host interface 118 via a CXL connection 155.

[0037] The CXL connection 155 can include a set of data transfer lanes (“lanes”) for implementing a CXL protocol, including a CXL.io protocol, a CXL.mem protocol, and a CXL.cache protocol. The CXL connection 155 can include any suitable number of lanes in accordance with the embodiments described herein. For example, the CXL connection 155 can include 16 lanes (i.e., CXL x16).

[0038] The host interface 118 can include media access control (MAC) and physical layer (PHY) components of the CXL memory device 110 for ingress of communications from the host system 120 to the CXL memory device 110 and egress of communications from the CXL memory device 110 to the host system 120. The bus 107 and the host interface 118 operate under a communication protocol, such as a CXL serial communication protocol via PCIe or other suitable communication protocol. Other suitable communication protocols include Ethernet, Serial Attached SCSI (SAS), Serial AT Attachment (SATA), any protocol related to Remote Direct Memory Access (RDMA) (e.g., InfiniBand, iWARP, or RDMA over Converged Ethernet (RoCE)), and other suitable serial communication protocols.

[0039] The computing system 100 can be a cache coherent interconnect for processors, memory expansion, and accelerators. The computing system 100 maintains memory coherency between CPU memory space and memory on attached devices, which allows resource sharing to enable higher performance, reduce software stack complexity, and lower overall system cost. In general, CXL is an interface standard that can support multiple protocols that can run on top of PCIe, including a CXL.io protocol, a CXL.mem protocol, and a CXL.cache protocol. The CXL.io protocol is a PCIe-like protocol that can be viewed as an “enhanced” PCIe protocol capable of partitioning managed memory. CXL.io can be used for initialization, connection, device discovery and enumeration, register access, and can provide an interface for I / O devices. The CXL.mem protocol can enable a host to access memory of an attached device using memory semantics (e.g., load and store commands). This approach can support both volatile and persistent memory architectures. The CXL.cache protocol can define host device interactions to enable efficient caching of host memory with low latency using a request and response approach. Traffic (e.g., NVMe traffic) can run over the CXL.io protocol, and the CXL.mem and CXL.cache protocols can share a common link layer and transaction layer. Thus, CXL protocols can be multiplexed and transported via a PCIe physical layer.

[0040] The CXL memory device 110 is a memory device that allows the host system 120 to use it for memory bandwidth expansion, memory capacity expansion, and persistent memory applications, and as a small-scale resource pooling and large-scale resource pooling and sharing.

[0041] In some implementations, a CXL memory device can be a multi-logical device (MLD), which can partition resources into multiple logical devices and each logical device can be visible as a memory device. One of the multi-logical devices can be reserved for a fabric manager to configure resource allocation across the logical devices, while the other logical devices can be for assignment to a host. In some implementations, a CXL memory device can be a device that supports multiple host systems and can be referred to as a fabric-attached memory (FAM). In the context of these computing environments, the term “fabric” can refer to an interconnected communication path that routes signals on a main component of a chip or between chips of a computing system. This “fabric” can form the architecture of an interconnect between processing or computing nodes within a computing device or between multiple computing devices. In this context, processing nodes and computing nodes refer to processing devices that operate as nodes on an interconnected network. A fabric-attached memory can refer to a memory architecture in which memory is connected to a CPU through a fabric interconnect, rather than being connected directly to the CPU. This allows the memory to be located at a distance from the CPU and can provide benefits such as improved scalability and fault tolerance. For example, in some systems, a fabric includes a collection of buses or connections that connect the processing devices of a system to peripheral devices and other processing devices. In other systems, a fabric can also include a collection of network connections between combinations of respective computing nodes and memory nodes. In various systems, a fabric serves as an interconnect to create a network of interconnected devices that work together as a single entity. This unified framework combines many interconnected devices (i.e., like many threads woven together to create a cohesive whole) via a fabric to provide fast and reliable communication between devices. In this context, an “interconnect” can refer to a device or system that connects multiple devices or subsystems together to allow them to communicate and exchange data.

[0042] The CXL memory device 110 can include a storage device, a memory module, or a combination of a storage device and a memory module. Examples of storage devices include a solid-state drive (SSD), a flash drive, a universal serial bus (USB) flash drive, an embedded Multi-Media Controller (eMMC) drive, a Universal Flash Storage (UFS) drive, a Secure Digital (SD) card, and a hard disk drive (HDD). Examples of memory modules include a dual in-line memory module (DIMM), a small outline DIMM (SO-DIMM), and various types of non-volatile dual in-line memory modules (NVDIMM).

[0043] The CXL memory devices 110 can include any combination of different types of non-volatile memory devices and / or volatile memory devices. Volatile memory devices can be, but are not limited to, random access memory (RAM), such as dynamic random access memory (DRAM) and synchronous dynamic random access memory (SDRAM). Some examples of non-volatile memory devices include negative-and (NAND) type flash memory and in-place write memory, such as a 3D cross-point (“3D cross-point”) memory device that is a cross-point array of non-volatile memory cells. A cross-point array of non-volatile memory cells can perform bit storage based on changes in bulk resistance in conjunction with a stackable cross-grided data access array. Additionally, in contrast to many flash-based memories, cross-point non-volatile memory can perform in-place write operations, where a non-volatile memory cell can be programmed without first erasing the non-volatile memory cell. NAND type flash memory includes, for example, two-dimensional NAND (2D NAND) and three-dimensional NAND (3D NAND).

[0044] The DCDs 130A-130N can include volatile memory devices, including random access memory (RAM), such as dynamic random access memory (DRAM) and synchronous dynamic random access memory (SDRAM), and non-volatile memory devices, including negative-and (NAND) type flash memory and in-place write memory, such as a 3D cross-point memory device that is a cross-point array of non-volatile memory cells, read-only memory (ROM), phase change memory (PCM), self-selecting memory, other chalcogenide-based memory, ferroelectric transistor, random access memory (FeTRAM), ferroelectric random access memory (FeRAM), magnetic random access memory (MRAM), spin-transfer torque (STT)-MRAM, conductive-bridge RAM (CBRAM), resistive random access memory (RRAM), oxide-based RRAM (OxRAM), or negative-or (NOR) flash memory, or electrically erasable programmable read-only memory (EEPROM).

[0045] The CXL memory device controller 115 can communicate with the DCDs 130A-130N to perform operations, such as reading data, writing data, or erasing data at the DCDs 130A-130N, and other such operations. The CXL memory device controller 115 can include hardware, such as one or more integrated circuits and / or discrete components, buffer memory, or a combination thereof. The hardware can include digital circuitry with dedicated (i.e., hard-coded) logic to perform the operations described herein. The CXL memory device controller 115 can be a microcontroller, special purpose logic circuitry (e.g., a field-programmable gate array (FPGA), an application specific integrated circuit (ASIC), etc.), or other suitable processor.

[0046] The CXL memory device controller 115 can include processing devices including one or more processors (e.g., processor 117) configured to execute instructions stored in local memory 119. In the illustrated example, the local memory 119 of the CXL memory device controller 115 includes embedded memory configured to store instructions for performing various processes, operations, logic flows, and routines for controlling operations of the CXL memory device 110, including processing communications between the CXL memory device 110 and the host system 120. The CXL memory device controller 115 can manage operations of the CXL memory device 110, such as writes to and reads from the DCDs 130A-130N. The CXL memory device controller 115 can include one or more processors 117, which can be multi-core processors. The processor 117 can generally process or interact with components of the DCDs 130A-130N through firmware code. The CXL memory device controller 115 can operate under the CXL protocol, although other protocols are applicable.

[0047] The CXL memory device controller 115 executes computer-readable program code (e.g., software or firmware) executable instructions (referred to herein as “instructions”). The instructions can be executed by various components of the CXL memory device controller 115, such as the processor 117, logic gates, switches, application specific integrated circuits (ASICs), programmable logic controllers, embedded microcontrollers, and other components of the CXL memory device controller 115. The instructions that can be executed by the CXL memory device controller 115 for carrying out the embodiments described herein are stored in a non-transitory computer-readable storage medium. In certain embodiments, the instructions are stored in a non-transitory computer-readable storage medium of the CXL memory device 110 (e.g., the DCDs 130A-130N). The instructions stored in the CXL memory device 110 can be executed without the addition of input or direction from the host system 120. In other embodiments, the instructions are transmitted from the host system 120. The CXL memory device controller 115 is configured with hardware and instructions to perform the various functions described herein and illustrated in the figures.

[0048] The CXL memory device controller 115 can interact with the DCDs 130A-130N for read and write operations. The CXL memory device controller 115 can perform direct memory access (DMA) for data transfer between the host system 120 and the DCDs 130A-130N without the involvement of the CPU 109. The CXL memory device controller 115 can control the data transfer while activating the control path for fetch commands, issuing completion and interrupts, and activating DMA for the actual data transfer between the host system 120 and the DCDs 130A-130N. The CXL memory device controller 115 can have an error correction module to correct data fetched from the memory arrays in the DCDs 130A-130N.

[0049] In some embodiments, the local memory 119 can include memory registers to store memory pointers, fetched data, etc. The local memory 119 can also include read-only memory (ROM) to store microcode. Although Figure 1 In some embodiments, the CXL memory device 110 has been illustrated as including the CXL memory device controller 115, but in another embodiment of the disclosure, the CXL memory device 110 does not include the CXL memory device controller 115, but can rely on external control (e.g., provided by an external host or by a processor or controller separate from the memory sub-system).

[0050] In general, the CXL memory device controller 115 can receive commands or operations from the host system 120 or the fabric manager 140 and can convert the commands or operations into instructions or appropriate commands to achieve the desired access to the DCDs 130A-130N. The CXL memory device controller 115 can be responsible for other operations such as wear leveling operations, garbage collection operations, error detection and error correction code (ECC) operations, encryption operations, cache operations, and address translation between logical addresses (e.g., logical block addresses (LBAs), namespaces) and physical addresses (e.g., physical MU addresses, physical block addresses) associated with the DCDs 130A-130N. The CXL memory device controller 115 can further include host interface circuitry to communicate with the host system 120 via a physical host interface. The host interface circuitry can convert commands received from the host system into command instructions to access the DCDs 130A-130N and convert responses associated with the DCDs 130A-130N into information for the host system 120.

[0051] The CXL memory device 110 can also include additional circuitry or components not illustrated. In some embodiments, the CXL memory device 110 can include a cache or buffer (e.g., DRAM) and address circuitry (e.g., row and column decoders) that can receive addresses from the CXL memory device controller 115 and decode the addresses to access the DCDs 130A-130N.

[0052] In some embodiments, each or some of the DCDs 130A-130N include a local media controller 135 that operates in conjunction with the CXL memory device controller 115 to perform operations on one or more memory units of the DCDs 130A-130N. An external controller (e.g., the CXL memory device controller 115) can externally manage the DCDs 130A-130N (e.g., perform media management operations on the memory device 130). In some embodiments, the CXL memory device 110 is a managed memory device that is a raw DCD 130A-130N with control logic (e.g., the local media controller 135) on the die and a controller (e.g., the CXL memory device controller 115) for media management within the same memory device package. An example of a managed memory device is a managed NAND (MNAND) device.

[0053] In some embodiments, the computing system 100 can include a fabric manager 140. The fabric manager 140 is an external logical process that queries and configures the operational state of the computing system 100, and can include application logic and policies that assign the DCDs 130A-130N to the host system 120 at runtime. In some embodiments, the fabric manager 140 can be software running on the host system 120, firmware embedded within a baseboard management controller (BMC) on another CXL device or CXL switch, or a dedicated device running in the CXL device. The fabric manager 140 can assign (logical) devices (e.g., the DCDs 130A-130N) to the host system 120 by using a command set over a component command interface (CCI). The CCI can be exposed through mailbox registers that provide the ability to issue commands (“mailbox commands”) to devices (e.g., the DCDs 130A-130N). In some implementations, each of the DCDs 130A-130N can include one or more taggable DCD units 136. In some embodiments, the CXL memory device 110 can include a CXL host interface 120A that is a CXL controller that is configured to communicate with the CXL memory device controller 115. Figure 1In the example of FIG. 1, the fabric manager 140 can assign one taggable DC unit to the host system 120 and create a globally unique tag attached to the taggable DC unit as the tag capacity unit 137; the fabric manager 140 can assign another taggable DC unit to the host system 120 and create a globally unique tag attached to the taggable DC unit as the tag capacity unit 138. Although Figure 1 In the example of FIG. 1, the fabric manager 140 can assign one taggable DC unit to the host system 120 and create a globally unique tag attached to the taggable DC unit as the tag capacity unit 137; the fabric manager 140 can assign another taggable DC unit to the host system 120 and create a globally unique tag attached to the taggable DC unit as the tag capacity unit 138. Although Figure 1 In the example of FIG. 1, the fabric manager 140 can assign one taggable DC unit to the host system 120 and create a globally unique tag attached to the taggable DC unit as the tag capacity unit 137; the fabric manager 140 can assign another taggable DC unit to the host system 120 and create a globally unique tag attached to the taggable DC unit as the tag capacity unit 138. Although

[0054] In some embodiments, the CXL memory device 110 includes a concatenate and truncate command component 113 that enables the host system 120 to execute concatenate commands and truncate commands. In some embodiments, the concatenate and truncate commands are each the mailbox command described above. In some embodiments, the CXL memory device controller 115 includes at least a portion of the concatenate and truncate command component 113. In some embodiments, the concatenate and truncate command component 113 is part of the host system 120, an application, or an operating system. In other embodiments, the local media controller 135 includes at least a portion of the concatenate and truncate command component 113 and is configured to perform the functionality described herein. Additional details regarding the operation of the concatenate and truncate command component 113 are described below with reference to Figure 2 to 6 In some embodiments, the concatenate and truncate command component 113 includes the command component 113A and the command component 113B shown in FIG. 1, which can operate together to perform the functionality of the concatenate and truncate command component 113. In some embodiments, some or all of the functionality of the concatenate and truncate command component 113 can be performed by the fabric manager 240, the controller 215, the command component 113A, the command component 113B, and / or combinations thereof, as shown in FIG. 2. Figure 2 Figure 2

[0055] Those skilled in the art will appreciate that additional circuitry and signals can be provided and that the components of FIG. 1 have been simplified for understanding Figure 1 Those skilled in the art will appreciate that additional circuitry and signals can be provided and that the components of FIG. 1 have been simplified for understanding Figure 1 ​​The functionality described for various block components can not necessarily be separated into different components or component parts of the integrated circuit device. For example, a single component or component part of the integrated circuit device can be adapted to perform the functionality of more than one block component described Figure 1 Alternatively, one or more components or component parts of the integrated circuit device can be combined to perform the functionality of a single block component described Figure 1

[0056] Figure 2 is a schematic block diagram of a system 200 to implement a taggable dynamic capacity unit in a compute express link (CXL) memory device. In various embodiments, the system 200 includes one or more host systems 220A-D (e.g., host systems 120), a CXL memory device 210 (e.g., CXL memory device 110) including a controller 215 (e.g., controller 115), a CXL fabric interconnect 245, a fabric manager 240 that can perform operations to manage the CXL fabric interconnect 245, and an orchestrator 250. In some embodiments, aspects of the controller 215 are included in processing logic of DCDs 230A-230D. The CXL memory device 210 can be connected to the host systems 220A-D via a network connection interface (e.g., a compute express link (CXL) fabric interconnect 245) that utilizes a high-speed bus (e.g., a peripheral component interconnect express (PCIe) bus). The compute express link (CXL) fabric interconnect 245 can provide an interface that can support several protocols that can run on top of PCIe, including a CXL.io protocol, a CXL.mem protocol, and a CXL.cache protocol. The CXL fabric interconnect 245 can be a collection of one or more switches, and each switch has a port-based routing (PBR) capability and is interconnected with PBR links. The CXL fabric interconnect 245 can connect one or more host ports to devices within a single coherent host physical address (HPA) space.

[0057] In the example of Figure 2 DCD 230A can include a first zone 236A, DCD 230B can include a second zone 236B, DCD 230C can include a third zone 236C, and DCD 230D can include a fourth zone 236D. As shown in Figure 2 each of the first zone 236A, the second zone 236B, the third zone 236C, and the fourth zone 236D can include one or more taggable dynamic capacity units. Although the zones are illustrated in Figure 2 as having uniform capacity sizes, the zones can have various capacity sizes.

[0058] ​In some embodiments, orchestrator 250 can control the accessibility of each tag to host systems 220A-D. Orchestrator 250 can make global control and management decisions about the cluster of host systems 220A-D. Orchestrator 250 can be responsible for maintaining a desired state of host systems 220A-D (i.e., the state that a client expects when running the cluster), such as which applications are running and which container images they use, which resources should be available to them, and other configuration details. In some embodiments, orchestrator 250 can be a container orchestration system, such as Kubernetes. In some embodiments, orchestrator 250 can be used to provide a containerized computing service platform, such as a platform-as-a-service (PaaS) system. A PaaS system provides resources and services (e.g., microservices) for developing and executing applications owned or managed by multiple users. A PaaS system provides a platform and environment that allows users to build applications and services in a cluster computing environment (“the cloud”). Orchestrator 250 can include nodes for executing applications and / or processes associated with the applications. A “node” that provides computing functionality can provide an execution environment for an application. In some embodiments, a “node” can include a virtual machine hosted on a physical machine, such as host systems 220A-220D implemented as part of a cloud. In some embodiments, a node can additionally or alternatively include a group of virtual machines, containers, or a group of containers to execute functionality of a PaaS application. When nodes are implemented as virtual machines, they can be executed by an operating system (OS) on each host system 220A-220D. Although embodiments of the present disclosure are described in terms of a certain type of system, this should not be taken as limiting the scope or usefulness of the features of the present disclosure. For example, the features and techniques described herein can be used with other types of multi-tenant systems and / or containerized computing service platforms.

[0059] Host systems 220A-D (e.g., through nodes running on host systems 220A-D) can request allocation of tag capacity in DCDs 230A-230D. For example, host systems 220A-D through nodes (e.g., applications, virtual machines) running on host systems 220A-D can request allocation of tag capacity in DCDs 230A-230D, where the request can specify a capacity size.

[0060] For initial allocation of a tagged capacity, the controller 215 and / or the fabric manager 240 can determine a portion of the DCDs 230A-230D for allocation. In some implementations, the controller 215 can determine an available portion of the DCDs 230A-230D to allocate to the host system 220A in a requested capacity size, and request the fabric manager 240 to provide a tag. The controller 215 can receive the tag from the fabric manager 240, and assign the tag to the allocated portion, referred to as a tagged capacity unit (e.g., tagged capacity unit 231A or 232A, or tagged capacity unit 231C). In some implementations, the fabric manager 240 can determine an available portion of the DCDs 230A-230D to allocate to the host system 220A in a requested capacity size, and assign the tag to the allocated portion, referred to as a tagged capacity unit (e.g., tagged capacity unit 231A or 232A, or tagged capacity unit 231C). In various implementations, the tag is created by the fabric manager 240 such that the tag is globally unique. The controller 215 can store the tag, a DPA range of the allocated portion of the DCDs 230A-230D, and a host identifier (or host group identifier) defining the host system that can access the tag in the tag mapping data structure 217.

[0061] At the time the tagged capacity unit is initially allocated, the host system 220A-220D can write data to the tagged capacity unit. Using the host system 220A and the DCD 230A as an example, at the time the tagged capacity unit 231A is initially allocated to the host system 220A, the controller 215 can receive data from the host system 220A created by an application running on the host system 220A. The data can include content reflecting the state of the application (e.g., the data can include information representing values of variables, memory layout, location of instruction pointers, and other details about the state of the application). The controller 215 can store the data in the tagged capacity unit 231A or 232A. The controller 215 can write the data to the tagged capacity unit 231A or 232A and map one or more DPA ranges identifying respective locations of the data on the CXL memory device 210 with corresponding virtual address ranges in a virtual address space available to the host system 220A (i.e., a virtual / logical address space assigned by the host system to the host application that created the data). Thus, the controller 215 can record the data at respective locations identified by a corresponding set of address ranges (e.g., a contiguous physical address range or a list of extents of non-contiguous physical address ranges indicating locations of the data on the CXL memory device 210). An example of using the tagged capacity unit 231A or 232A as a source tag for a concatenation command will be explained below.

[0062] Using host system 220C and DCD 230C as another example, when a tagged capacity unit 231C is initially allocated to host system 220C, controller 215 can receive data created by an application running on host system 220C from host system 220C. The data can include content reflecting the state of the application (e.g., the data can include information representing values of variables, memory layout, locations of instruction pointers, and other details about the state of the application). Controller 215 can store the data in tagged capacity unit 231C. Controller 215 can write the data to tagged capacity unit 231C and map one or more DPA ranges identifying respective locations containing the data on CXL memory device 210 to corresponding virtual address ranges in a virtual address space available to host system 220C (i.e., a virtual / logical address space assigned by the host system to the host application that created the data). Thus, controller 215 can record the data at respective locations identified by a corresponding set of address ranges (e.g., a contiguous physical address range or a list of extents of non-contiguous physical address ranges indicating locations of the data on CXL memory device 210). An example of using tagged capacity unit 231C as a source tag for a truncate command will be explained below.

[0063] Figure 3AExamples of commands 300A that can be used as concatenate commands or truncate commands are illustrated. A concatenate command can include bytes designated to represent parameters including two or more source tags, a host identifier (or host group identifier), a zone number, a selection policy, etc. In some implementations, a concatenate command can further include a destination tag (not shown). Each parameter of a concatenate command can have a specific length in bytes and be placed at a specific byte offset. The source tags of a concatenate command can identify two or more tags, where each tag is associated with data to be combined by the concatenate command. The host identifier (or host group identifier) of a concatenate command can identify a host system to which the destination tag of the concatenate command is to be assigned. The zone number of a concatenate command can identify a zone of a DCD in which the tag capacity unit associated with the destination tag resides, where the identified zone has various attributes including whether the zone is writable and / or readable, whether the zone has a latest copy from a cache line consistent in hardware is a hardware-managed consistency, or whether the zone is shareable by one or more host systems. The selection policy of a concatenate command can specify a policy for selecting a taggable DC unit of a DCD as the tag capacity unit associated with the destination tag. For example, a "free and contiguous" policy can allow a memory controller (e.g., controller 215) to select free and contiguous space in a DCD, while a "provisioned" policy can allow a fabric manager (e.g., fabric manager 240) to select provisioned space in a DCD. A concatenate command can optionally include bytes designated to represent parameters including a capacity size of a taggable DC unit associated with a specified tag, a number of taggable DC units associated with a specified tag, etc. Although not shown in Figure 3A , a concatenate command can include bytes designated to represent parameters including an order of source tags to be combined.

[0064] The truncate command can include bytes designated to represent parameters including one or more source tags, a host identifier (or host group identifier), a zone number, a selection policy, etc. In some implementations, the truncate command can further include a destination tag (not shown). Each parameter of the truncate command can have a specific length in bytes and be placed at a specific byte offset. The source tags of the truncate command can identify one or more tags, where each tag is associated with data to be removed and combined by the truncate command. The host identifier (or host group identifier) of the truncate command can identify a host system to which the destination tag of the truncate command is to be assigned. The zone number of the truncate command can identify a zone of the DCD in which the tag capacity unit associated with the destination tag resides, where the identified zone has various attributes including whether the zone is writable and / or readable, whether the zone has a latest copy from a cache line is consistent in hardware is hardware-managed consistency, or whether the zone is shareable by one or more host systems. The selection policy of the truncate command can specify a policy for selecting a taggable DC unit of the DCD as the tag capacity unit associated with the destination tag. For example, a "free and contiguous" policy can allow a memory controller (e.g., controller 215) to select free and contiguous space in the DCD, while a "provisioned" policy can allow a fabric manager (e.g., fabric manager 240) to select provisioned space in the DCD. The truncate command can optionally include bytes designated to represent parameters including a capacity size of the taggable DC unit associated with the destination tag, a number of taggable DC units associated with the new tag, etc.

[0065] Although not shown in Figure 3A , the truncate command can include bytes designated to represent parameters including one or more subzones of the one or more source tags. The subzones of the truncate command can identify one or more subzones, where each subzone is associated with data to be removed by the truncate command. The subzones can be identified by using a DPA range. In some implementations, the truncate command can include bytes designated to represent parameters including an order of the subzones.

[0066] Figure 3BAn example tag mapping data structure 300B (e.g., tag mapping data structure 217) that can be used to implement concatenation commands and truncation commands is illustrated. The tag mapping data structure 300B can include an item “DPA Range,” an item “Tag,” and an item “Host ID.” The item “DPA Range” indicates a location on a CXL memory device where data is stored (i.e., one or more physical address ranges of a tag capacity unit). A physical address range that identifies a corresponding location on a CXL memory device where data is stored can be referred to as a “physical address range of a tag capacity unit that contains data.” The item “Tag” indicates a tag associated with a tag capacity unit. The item “Host ID” indicates a host system from which a tag capacity unit associated with a tag can be accessed. The mapping data structure 300B can include multiple records, and each record can correspond to a tag and include multiple items as described above.

[0067] In view of the item “DPA Range,” the item “Tag,” and the item “Host ID,” the tag mapping data structure 300B can be used to map a DPA range to a host system by mapping a physical address range of a tag to a corresponding virtual address range in a virtual address space of a host system (i.e., a virtual / logical address space assigned by a host system to a host application that is permitted to access data).

[0068] Now referring again to Figure 2 , embodiments of concatenation commands are illustrated using examples of host systems 220A and 220B and DCDs 230A and 230B, while embodiments of truncation commands are illustrated using examples of host systems 220C and 220D and DCDs 230C and 230D. It should be noted that each of the concatenation commands and the truncation commands can be implemented by one or more host systems, one or more nodes, and one or more DCDs. For example, with respect to concatenation commands, a host system associated with a source tag can be the same or different than a host system associated with a destination tag; with respect to truncation commands, a host system associated with a source tag can be the same or different than a host system associated with a destination tag. In some embodiments, the command component 113A included in the fabric manager 240 and the command component 113B included in the controller 215 can perform the functions of the concatenation and truncation command components 113 of Figure 1 and operate together to implement concatenation commands and / or truncation commands.

[0069] To implement a concatenate command, host system 220B can send a concatenate command to CXL memory device 210 through a node (e.g., an application, a virtual machine) running on host system 220B. The concatenate command can enable host system 220B to combine two or more source-tagged data and store the combined data in a tagged capacity unit with a destination tag such that the combined data is separate from the original data and can be given full access to host system 220B.

[0070] For example, host system 220B can send a concatenate command, where the concatenate command includes first source tag 231 A and second source tag 232A, a host identifier (or host group identifier) 220B, and a selection policy of “free and contiguous.” For example, first source tag 231 A and second source tag 232A can be associated with host system 220A, and thus host system 220A can be a source host system and host system 220B can be a destination host system. In response to determining that the selection policy is “free and contiguous,” command component 113B can determine a taggable DC unit in region 236A (e.g., the same region as the source tags) and send a request for a tag to command component 113A. In response to the request for a tag, command component 113A can send a tag to command component 113B. Command component 113B can assign the tag to the determined taggable DC unit as tagged capacity unit 233A. Command component 113B can combine the data of tagged capacity units 231 A and 232A and store the combined data to tagged capacity unit 233A. In one example, the concatenate command can specify an order for combining the source tags, e.g., first source tag 231 A before second source tag 232A. Command component 113B can update tag mapping data structure 217 to include a record of the destination tag associated with tagged capacity unit 233A. Command component 113B can notify host system 220B about the completion of the concatenate command.

[0071] As another example, host system 220B can send a concatenate command, where the concatenate command includes first source tag 231 A and second source tag 232A, a host identifier (or host group identifier) 220B, a zone number 236B, and a selection policy that is “prescriptive.” In response to determining that the selection policy is “prescriptive,” command component 113A can determine the taggable DC units in zone 236B, and send a tag with information of the determined taggable DC units to command component 113B. Command component 113B can assign the tag to the determined taggable DC units as tag capacity unit 231B. Command component 113B can combine data of tag capacity units 231 A and 232A, and store the combined data in tag capacity unit 231B. In one example, the concatenate command can specify an order for combining the source tags, e.g., first source tag 231 A after second source tag 232A. Command component 113B can update tag mapping data structure 217 to include a record of the destination tag associated with tag capacity unit 231B. Command component 113B can notify host system 220B about the completion of the concatenate command.

[0072] In some embodiments, source host system 220A can exclusively access data of source tags (e.g., tag capacity unit 231 A and tag capacity unit 232A), and destination host system 220B can exclusively access data of destination tags (tag capacity unit 233A or tag capacity unit 231B). In some embodiments, the concatenate command enables controller 215 to delete source tags (e.g., tag capacity unit 231 A and tag capacity unit 232A) by freeing up the tag capacity units associated with the source tags, and in such cases, source host system 220A can no longer access data of the source tags.

[0073] To implement the truncate command, host system 220C can send a truncate command to CXL memory device 210 through a node (e.g., an application, a virtual machine) running on host system 220C. The truncate command can enable host system 220C to remove a portion of data from data of one or more source tags, and store the removed data in another tag capacity unit with a truncate destination tag, such that the removed data is separated from the original data and can be given full access to host system 220C.

[0074] For example, host system 220C can send a truncate command, where the truncate command includes source tag 231C, subregion 232C, host identifier (or host group identifier) 220C, and a selection policy of "free and contiguous." In response to determining that the selection policy is "free and contiguous," command component 113B can determine a taggable DC unit in region 235C (e.g., the same region as the source tag) and send a request for a tag to command component 113A. In response to the request for a tag, command component 113A can send a tag to command component 113B. Command component 113B can assign the tag to the determined taggable DC unit as a tag capacity unit 235C. Command component 113B can remove the data of subregion 232C from the data of source tag 231C and store the removed data (i.e., the data of subregions 233C and 234C) in tag capacity unit 235C. In one example, the truncate command can specify an order for combining the removed data, e.g., an order of subregion 233C before subregion 234C. Command component 113B can update tag mapping data structure 217 to include a record of the destination tag associated with tag capacity unit 235C. Command component 113B can notify host system 220C about the completion of the truncate command.

[0075] As another example, host system 220C can send a truncate command, where the truncate command includes source tag 231C, subregion 233C, host identifier (or host group identifier) 220C, region number 236D, and a selection policy of "specification." In response to determining that the selection policy is "specification," command component 113A can determine a taggable DC unit in region 236D and send a tag with information of the determined taggable DC unit to command component 113B. Command component 113B can assign the tag to the determined taggable DC unit as a tag capacity unit 231D. Command component 113B can remove the data of subregion 233C from the data of source tag 231C and store the removed data (i.e., the data of subregions 232C and 234C) in tag capacity unit 231D. In one example, the truncate command can specify an order for combining the removed data, e.g., an order of subregion 232C after subregion 234C. Command component 113B can update tag mapping data structure 217 to include a record of the destination tag associated with tag capacity unit 231D. Command component 113B can notify host system 220C about the completion of the truncate command.

[0076] Although the subregions associated with the truncate command explained above are subregions to be removed, a subregion can represent a subregion to be kept, which is to be kept such that the remaining subregions are removed from the tag capacity unit. Thus, in some embodiments, a truncate command can specify whether a subregion is a subregion to be removed or a subregion to be kept.

[0077] Although not illustrated in Figure 2 In another example, the node can send a truncate command that includes a subregion (not shown) in source tag 231 A, a subregion 234C in source tag 231C, and a host identifier (or host group identifier) 220D. Command component 113B can remove data stored in the subregion in source tag 231 A from the data of source tag 231 A, remove data stored in subregion 234C from the data of source tag 231C, combine the data stored in the removed subregion of source tag 231 A and the data stored in subregions 232C and 233C in source tag 231C, and store the combined data in another tag capacity unit (e.g., 231B) associated with a new destination tag. The node can further send a concatenate command that includes a first source tag 231B and a second source tag 232A and a host identifier (or host group identifier) 220D. Command component 113B can combine the data of source tag 231B and the data of source tag 232A and store the combined data in another tag capacity unit (e.g., 233A) associated with a new destination tag.

[0078] Figure 4 is an example method 400 for using concatenate commands associated with tag capacities in a compute express link (CXL) memory device, according to some embodiments of the present disclosure. The method 400 can be performed by processing logic that can include hardware (e.g., processing device, circuitry, dedicated logic, programmable logic, microcode, hardware of a device, integrated circuit, etc.), software (e.g., instructions run or executed on a processing device), or a combination thereof. In some embodiments, the method 400 is performed by the concatenate and truncate command component 113 of Figure 1 or the controller 215 (including command component 113B) and the structure manager 240 (including command component 113A) of Figure 2 of the present disclosure. Although shown in a particular sequence or order, unless otherwise specified, the order of the processes can be modified. Thus, it is to be understood that the illustrated embodiments are merely examples, and the illustrated processes can be performed in different orders, and some processes can be performed in parallel. Additionally, one or more processes can be omitted in various embodiments. Thus, not all processes are required in every embodiment. Other process flows are also possible.

[0079] At operation 410, the processing logic can receive a host command to combine first data associated with a first tag and second data associated with a second tag, where the first tag is associated with a first memory section (e.g., a memory section corresponding to 231A) of a plurality of dynamic capacity devices (e.g., DCDs 230A-230D), where the second tag is associated with a second memory section (e.g., a memory section corresponding to 232A) of the plurality of dynamic capacity devices (e.g., DCDs 230A-230D), and where the first memory section is allocated to a first host system to store the first data and the second memory section is allocated to the first host system to store the second data. In some embodiments, each of the plurality of dynamic capacity devices includes a plurality of memory sections, where each of the plurality of memory sections is associated with a respective one of a plurality of tags, and where each of the plurality of tags is unique.

[0080] In some embodiments, the host command to combine specifies the first tag and the second tag. In some embodiments, the host command to combine specifies at least one of: a zone identifier of the memory device or a selection policy for selecting a memory section. In some embodiments, the memory device comprises a compute express link (CXL) enabled memory device.

[0081] In some embodiments, the processing logic can receive an allocation request from the first host system, where the allocation request specifies the first host system and the capacity, allocate the first memory section to the first host system, and associate the first tag with the first memory section. In some embodiments, the processing logic can receive the first data from the host system and store the first data in the first memory section. In some embodiments, the processing logic can create the first tag in response to receiving the allocation request by a node in an orchestrator cluster, where the node runs on the first host system.

[0082] In some embodiments, the processing logic can receive an allocation request from the first host system, where the allocation request specifies the first host system and the capacity, allocate the second memory section to the first host system, and associate the second tag with the second memory section. In some embodiments, the processing logic can receive the second data from the host system and store the second data in the second memory section. In some embodiments, the processing logic can create the second tag in response to receiving the allocation request by a node in an orchestrator cluster, where the node runs on the first host system.

[0083] In some embodiments, the processing logic can map the first tag to the first memory segment and the first host system. In some embodiments, the first tag is shared by the first host system and another host system. In some embodiments, a capacity of the first memory segment allocated to the first host system and associated with the first tag is immutable.

[0084] In some embodiments, the processing logic can map the second tag to the second memory segment and the first host system. In some embodiments, the second tag is shared by the first host system and another host system. In some embodiments, a capacity of the second memory segment allocated to the first host system and associated with the second tag is immutable.

[0085] At operation 420, in response to receiving the host command to combine, the processing logic can determine a third memory segment (e.g., a memory segment corresponding to 231B) of the plurality of dynamic capacity devices and associate a third tag with the third memory segment.

[0086] In some embodiments, the host command to combine is received from the second host system, and wherein the host command specifies the second host system. In some embodiments, the processing logic can map the third tag to the third memory segment and the second host system. In some embodiments, the third tag is shared by the second host system and another host system. In some embodiments, a capacity of the third memory segment allocated to the second host system and associated with the third tag is immutable.

[0087] At operation 430, the processing logic can combine the first data and the second data and store the combined data in the third memory segment (e.g., a memory segment corresponding to 231B). In some embodiments, the first data and the second data are combined according to an order provided in the host command to combine. In some embodiments, the first memory segment and the second memory segment are accessible by the first host system and the third memory segment is accessible by the second host system. In some embodiments, the first memory segment and the second memory segment are exclusively accessible by the first host system and the third memory segment is exclusively accessible by the second host system. In some embodiments, at least one of the first tag and the second tag is deleted in the tag mapping data structure. In some embodiments, the processing logic can delete a mapping of the first tag, the first memory segment, and an identifier of the first host system and delete a mapping of the second tag, the second memory segment, and the identifier of the first host system.

[0088] Figure 5is a flowchart of an example method 500 for using a truncate command associated with a tag capacity in a compute express link (CXL) memory device, according to some embodiments of the present disclosure. The method 500 can be performed by processing logic that can include hardware (e.g., processing device, circuitry, dedicated logic, programmable logic, microcode, hardware of a device, integrated circuit, etc.), software (e.g., instructions run or executed on a processing device), or a combination thereof. In some embodiments, the method 500 is performed by the Figure 1 of the concatenation and truncate command component 113 or Figure 2 of the controller 215 (including the command component 113B) and the fabric manager 240 (including the command component 113A). Although shown in a particular sequence or order, unless otherwise specified, the order of the processes can be modified. Thus, it is to be understood that the illustrated embodiments are merely examples, and the illustrated processes can be performed in different orders, and some processes can be performed in parallel. Additionally, one or more processes can be omitted in various embodiments. Thus, not all processes are required in every embodiment. Other process flows are also possible.

[0089] At operation 510, the processing logic can receive a host command to remove first data stored in a first subregion (e.g., a memory subregion corresponding to 233C) of a plurality of subregions associated with a first tag from second data stored in the plurality of subregions associated with the first tag, where the first tag is associated with a first memory section (e.g., a memory section corresponding to 231C) of a plurality of dynamic capacity devices (e.g., DCDs 230A-230D), and where the first memory section is allocated to store the second data. In some implementations, each of the plurality of dynamic capacity devices includes a plurality of memory sections, where each of the plurality of memory sections is associated with a respective one of a plurality of tags, and where each of the plurality of tags is unique.

[0090] In some implementations, the host command to remove specifies the first tag and the first subregion. In some implementations, the host command to remove specifies at least one of: a region identifier of the memory device or a selection policy for selecting the memory section. In some implementations, the memory device comprises a compute express link (CXL) enabled memory device.

[0091] In some embodiments, the processing logic can receive an allocation request from the first host system, where the allocation request specifies the first host system and a capacity, allocate the first memory section to the first host system, and associate the first tag with the first memory section. In some embodiments, the processing logic can receive second data from the host system and store the second data in the first memory section. In some embodiments, the processing logic can create the first tag in response to receiving the allocation request by a node in the orchestrator cluster, where the node runs on the first host system.

[0092] In some embodiments, the processing logic can map the first tag to the first memory section and the first host system. In some embodiments, the first tag is shared by the first host system and another host system. In some embodiments, the capacity of the first memory section allocated to the first host system and associated with the first tag is immutable.

[0093] At operation 520, in response to receiving the host command to remove, the processing logic can determine a second memory section (e.g., a memory section corresponding to 231D) of the plurality of dynamic capacity devices and associate a second tag with the second memory section.

[0094] In some embodiments, the host command to remove is received from the second host system, and where the host command specifies the second host system. In some embodiments, the processing logic can map the second tag to the second memory section and the second host system. In some embodiments, the second tag is shared by the second host system and another host system. In some embodiments, the capacity of the second memory section allocated to the second host system and associated with the second tag is immutable.

[0095] At operation 530, the processing logic can remove the first data from the second data to generate post-removal data and store the post-removal data in the second memory section (e.g., a memory section corresponding to 231D). In some embodiments, the post-removal data includes a plurality of data sets, and the processing logic can combine the plurality of data sets according to an order provided in the host command. In some embodiments, the first memory section is accessible by the first host system, and the second memory section is accessible by the second host system. In some embodiments, the first memory section is exclusively accessible by the first host system, and the second memory section is exclusively accessible by the second host system. In some embodiments, the first tag is deleted in the tag mapping data structure. In some embodiments, the processing logic can delete the mapping of the first tag, the first memory section, and an identifier of the first host system.

[0096] Figure 6An example machine, the computer system 600, within which a set of instructions, for causing the machine to perform any one or more of the methodologies discussed herein, can be executed, is described. In some embodiments, the computer system 600 can correspond to a host system (e.g., the host system 120) that includes, is coupled to, or utilizes a memory sub-system (e.g., the CXL memory device 110) or can be used to perform operations of a controller (e.g., execute an operating system to perform operations related to the Figure 1 Figure 1 Figure 1 Figure 2 In alternative embodiments, the machine can be connected (e.g., networked) to other machines in a LAN, an intranet, an extranet, and / or the Internet. The machine can operate in the capacity of a server or a client machine in client-server network environment, as a peer machine in a peer-to-peer (or distributed) network environment, or as a server or a client machine in a cloud computing infrastructure or environment.

[0097] The machine can be a personal computer (PC), a tablet PC, a set-top box (STB), a personal digital assistant (PDA), a cellular telephone, a web appliance, a server, a network router, a switch or bridge, or any machine capable of executing a set of instructions (sequential or otherwise) that specify actions to be taken by that machine. Further, while a single machine is illustrated, the term “machine” shall also be taken to include any collection of machines that individually or jointly execute a set (or multiple sets) of instructions to perform any one or more of the methodologies discussed herein.

[0098] The example computer system 600 includes a processing device 602, a main memory 604 (e.g., read-only memory (ROM), flash memory, dynamic random access memory (DRAM) such as synchronous DRAM (SDRAM) or RDRAM, etc.), a static memory 606 (e.g., flash memory, static random access memory (SRAM), etc.), and a data storage system 618, which communicate with each other via a bus 630.

[0099] ​​​Processing device 602 represents one or more general-purpose processing devices such as a microprocessor, a central processing unit, or the like. More particularly, the processing device can be a complex instruction set computing (CISC) microprocessor, reduced instruction set computing (RISC) microprocessor, very long instruction word (VLIW) microprocessor, or a processor implementing other instruction sets, or processors implementing a combination of instruction sets. Processing device 602 can also be one or more special-purpose processing devices such as an application specific integrated circuit (ASIC), a field programmable gate array (FPGA), a digital signal processor (DSP), network processor, or the like. The processing device 602 is configured to execute instructions 626 for performing the operations and steps discussed herein. The computer system 600 can further include a network interface device 608 to communicate over the network 620.

[0100] The data storage system 618 can include a machine-readable storage medium 624 (also known as a computer-readable medium) on which is stored one or more sets of instructions 626 or software embodying any one or more of the methodologies or functions described herein. The instructions 626 can also reside, completely or at least partially, within the main memory 604 and / or within the processing device 602 during execution thereof by the computer system 600, the main memory 604 and the processing device 602 also constituting machine-readable storage media. The machine-readable storage medium 624, data storage system 618, and / or main memory 604 can be the same as or different from one another. Both the machine-readable storage medium 624 and the processing device 602 are Figure 1 The CXL memory device 110 can correspond to the CXL memory device 110 of FIG. 1.

[0101] In one embodiment, the instructions 626 include instructions to implement functionality corresponding to the concatenation and truncation command component 113 of FIG. 1 or the command component 113B of FIG. 2. Figure 1 In one embodiment, the instructions 626 include instructions to implement functionality corresponding to the concatenation and truncation command component 113 of FIG. 1 or the command component 113B of FIG. 2. Figure 2 Although the machine-readable storage medium 624 is shown in an example embodiment to be a single medium, the term “machine-readable storage medium” should be taken to include a single medium or multiple media that store one or more sets of instructions. The term “machine-readable storage medium” shall also be taken to include any medium that is capable of storing or encoding a set of instructions for execution by the machine and that cause the machine to perform any one or more of the methodologies of the present disclosure. The term “machine-readable storage medium” shall accordingly be taken to include, but not be limited to, solid-state memories, optical media, and magnetic media.

[0102] Some portions of the preceding detailed descriptions have been presented in terms of algorithms and symbolic representations of operations on data bits within a computer memory. These algorithmic descriptions and representations are the means used by those skilled in the data processing arts to most effectively convey the substance of their work to others skilled in the art. In this context, and for purposes of

[0103] It should be borne in mind, however, that all of these and similar terms are to be associated with the appropriate physical quantities and are merely convenient labels applied to these quantities. The present disclosure can refer to the action and processes of a computer system, or similar electronic computing device, that manipulates and transforms data represented as physical (electronic) quantities within the computer system's registers and memories into other data similarly represented as physical quantities within the computer system memories or registers or other such information storage systems.

[0104] The present disclosure also relates to an apparatus for performing the operations herein. This apparatus can be specially constructed for the required purposes, or it can comprise a general purpose computer selectively activated or reconfigured by a computer program stored in the computer. Such a computer program can be stored in a computer readable storage medium, such as any type of disk including floppy disks, optical disks, CD-ROMs, and magnetic-optical disks, read-only memories (ROMs), random access memories (RAMs), EPROMs, EEPROMs, magnetic or optical cards, or any type of media suitable for storing electronic instructions, each coupled to a computer system bus.

[0105] The algorithms and displays presented herein are not inherently related to any particular computer or other apparatus. Various general purpose systems can be used with programs in accordance with the teachings herein, or it can prove convenient to construct a more specialized apparatus to perform the method. The structure for a variety of these systems will appear as described in the description below. In addition, the present disclosure is not described with reference to any particular programming language. It will be appreciated that a variety of programming languages can be used to implement the teachings of the disclosure as described herein.

[0106] The disclosure can be provided as a computer program product, or software, that can include a machine-readable medium having stored thereon instructions that can be used to program a computer system (or other electronic devices) to perform a process according to the disclosure. A machine-readable medium includes any mechanism for storing information in a form accessible by a machine (e.g., a computer). In some embodiments, a machine- readable (e.g., computer-readable) medium includes a machine (e.g., a computer) readable storage medium, such as read only memory ("ROM"), random access memory ("RAM"), magnetic disk storage media, optical storage media, flash memory components, etc.

[0107] In the foregoing specification, embodiments of the disclosure have been described with reference to specific examples thereof. It will be evident that various modifications can be made to the disclosure without departing from the broader spirit and scope of the embodiments of the disclosure as set forth in the technical disclosure. The specification and drawings are, accordingly, to be regarded in an illustrative sense rather than a restrictive sense.

Claims

1. A system comprising: a memory device comprising a plurality of dynamic capacity devices; and a processing device operably coupled with the memory device to perform operations comprising: receiving a host command to combine first data associated with a first tag and second data associated with a second tag, wherein the first tag is associated with a first memory section of the plurality of dynamic capacity devices, wherein the second tag is associated with a second memory section of the plurality of dynamic capacity devices, and wherein the first memory section is allocated to a first host system to store the first data and the second memory section is allocated to the first host system to store the second data; and in response to receiving the host command to combine: determining a third memory section of the plurality of dynamic capacity devices and associating a third tag with the third memory section; combining the first data and the second data into combined data; and storing the combined data in the third memory section.

2. The system of claim 1, wherein each of the plurality of dynamic capacity devices comprises a plurality of memory sections, wherein each of the plurality of memory sections is associated with a respective one of a plurality of tags, and wherein each of the plurality of tags is unique.

3. The system of claim 1, wherein the host command specifies the first tag and the second tag.

4. The system of claim 1, wherein the host command is received from a second host system, and wherein the host command specifies the second host system.

5. The system of claim 4, wherein the first memory section and the second memory section are accessible by the first host system, and wherein the third memory section is accessible by the second host system.

6. The system of claim 1, wherein a capacity of the first memory section allocated to the first host system and associated with the first tag is invariant, and wherein a capacity of the second memory section allocated to the first host system and associated with the second tag is invariant.

7. The system of claim 1, wherein the host command specifies at least one of: an area of the plurality of dynamic capacity devices for the third memory section, or a selection policy for selecting the third memory section.

8. The system of claim 1, wherein the memory device comprises a compute express link (CXL) enabled memory device.

9. The system of claim 1, wherein the processing device is to perform operations further comprising: in response to receiving a first allocation request from the first host system, determining the first memory section and associating the first tag with the first memory section; storing the first data in the first memory section; in response to receiving a second allocation request from the first host system, determining the second memory section and associating the second tag with the second memory section; and ​ storing the second data in the second memory section.

10. The system of claim 9, wherein the processing device is to perform operations further comprising: creating the first tag in response to receiving the first allocation request by a first node in an orchestrator cluster, wherein the first node runs on the first host system; and creating the second tag in response to receiving the second allocation request by a second node in the orchestrator cluster, wherein the second node runs on the first host system.

11. The system of claim 1, wherein the processing device is to perform operations further comprising: deleting the first tag, the first memory section, and a mapping of an identifier of the first host system; and deleting the second tag, the second memory section, and a mapping of the identifier of the first host system.

12. The system of claim 1, wherein the first data and the second data are combined according to an order provided in the host command.

13. A method comprising: receiving, by a processing device, a host command to combine first data associated with a first tag and second data associated with a second tag, wherein the first tag is associated with a first memory section of a plurality of dynamic capacity devices in a memory device, wherein the second tag is associated with a second memory section of the plurality of dynamic capacity devices, and wherein the first memory section is allocated to a first host system to store the first data and the second memory section is allocated to the first host system to store the second data; in response to receiving the host command to combine: determining a third memory section of the plurality of dynamic capacity devices and associating a third tag with the third memory section; combining the first data and the second data into combined data; and storing the combined data in the third memory section.

14. The method of claim 13, wherein each of the plurality of dynamic capacity devices comprises a plurality of memory sections, wherein each of the plurality of memory sections is associated with a respective one of a plurality of tags, and wherein each of the plurality of tags is unique.

15. The method of claim 13, wherein the host command specifies the first tag and the second tag.

16. The method of claim 13, wherein the host command is received from a second host system, and wherein the host command specifies the second host system.

17. The method of claim 13, wherein a capacity of the first memory section allocated to the first host system and associated with the first tag is unchangeable, and wherein a capacity of the second memory section allocated to the first host system and associated with the second tag is unchangeable.

18. The method of claim 13, wherein the host command specifies at least one of: a zone of the plurality of dynamic capacity devices for the second memory section, or a selection policy for selecting the second memory section.

19. The method of claim 13, wherein the first data and the second data are combined according to an order provided in the host command.

20. A non-transitory computer-readable storage medium comprising instructions that, when executed by a processing device, cause the processing device to perform operations comprising: receiving a host command to combine first data associated with a first tag and second data associated with a second tag, wherein the first tag is associated with a first memory section of a plurality of dynamic capacity devices in a memory device, wherein the second tag is associated with a second memory section of the plurality of dynamic capacity devices, and wherein the first memory section is allocated to a first host system to store the first data and the second memory section is allocated to the first host system to store the second data; and in response to receiving the host command to combine: determining a third memory section of the plurality of dynamic capacity devices and associating a third tag with the third memory section; combining the first data and the second data into combined data; and storing the combined data in the third memory section.

21. The method of claim 20, wherein the first data and the second data are combined according to an order provided in the host command.

22. The method of claim 20, wherein the first data and the second data are combined according to a selection policy provided in the host command.

23. The method of claim 20, wherein the first data and the second data are combined according to a selection policy provided in the host command.

24. The method of claim 20, wherein the first data and the second data are combined according to a selection policy provided in the host command.

25. The method of claim 20, wherein the first data and the second data are combined according to a selection policy provided in the host command.

26. The method of claim 20, wherein the first data and the second data are combined according to a selection policy provided in the host command.

27. The method of claim 20, wherein the first data and the second data are combined according to a selection policy provided in the host command.

28. The method of claim 20, wherein the first data and the second data are combined according to a selection policy provided in the host command.

29. The method of claim 20, wherein the first data and the second data are combined according to a selection policy provided in the host command.

30. The method of claim 20, wherein the first data and the second data