Bare die level fault recovery

By switching programming techniques and using ECC engines and XOR data in the storage system, and dynamically adjusting the read reference voltage, the problem of die-level fault recovery is solved, and the reliability and efficiency of the storage system are improved.

CN123125650APending Publication Date: 2026-11-17SANDISK TECHNOLOGIES LLC
View PDF 0 Cites 0 Cited by

Patent Information

Application Number
CN202580029059.X
Authority / Receiving Office
CN · China
Patent Type
Applications(China)
Current Assignee / Owner
Priority Date
2024-09-16
Filing Date
2025-04-30
Publication Date
2026-11-17

AI Technical Summary

Technical Problem

Existing storage systems struggle to effectively recover data from die-level failures, leading to data loss. Furthermore, existing redundancy coding methods either reduce storage capacity or result in excessively long recovery times.

Method used

By switching programming techniques in the storage system, using a first voltage and a second voltage to precharge the NAND channel, and combining ECC engine and XOR data, the read reference voltage is dynamically adjusted to recover from die-level faults.

Benefits of technology

It improves the recovery capability from die-level failures, reduces the risk of data loss, and enhances the reliability and efficiency of the storage system.

✦ Generated by Eureka AI based on patent content.

Smart Images

  • Figure CN123125650A_ABST
    Figure CN123125650A_ABST
Patent Text Reader

Abstract

The programming techniques disclosed herein can be used to recover from potential planar and / or die-level faults. If a first programming technique is used in a target area of ​​a block, the memory system can detect evidence of planar or die-level faults affecting memory operations performed in that target area. For example, the memory system might notice that a large number of blocks have memory operation faults in that target area. In response to detecting evidence of planar or die-level faults, the memory system switches to a second programming technique for programming that target area. After programming the block in that target area using the second programming technique, the block can be tested against reliability criteria. The second programming technique overcomes the planar or die-level faults, allowing the block to continue to be used with the second programming technique.
Need to check novelty before this filing date? Find Prior Art

Description

Cross-references to related applications

[0001] This application claims priority to U.S. non-provisional patent application No. 18 / 886,461, filed on September 16, 2024, entitled “DIE LEVEL FAILURE RECOVERY,” which is incorporated herein by reference. Background Technology

[0002] This disclosure relates to non-volatile memory.

[0003] Semiconductor memories are widely used in a variety of electronic devices, such as cellular phones, digital cameras, personal digital assistants, medical electronics, mobile computing devices, servers, solid-state drives, non-mobile computing devices, and other devices. Semiconductor memories can include non-volatile memory or volatile memory. Non-volatile memory allows information to be stored and retained even when it is not connected to a power source (e.g., a battery).

[0004] Modern memory systems, such as solid-state drives, typically contain several semiconductor dies, each containing memory cells such as NAND strings. Each die may contain one or more planes, each plane containing a large number of blocks. Each block contains a large number of memory cells, such as NAND strings. A NAND string contains memory cell transistors connected in series, a drain-side select gate at one end, and a source-side select gate at the other end. Each NAND string is associated with a bit line. A block typically has a number of word lines that provide voltage to the control gates of the memory cell transistors. In some architectures, each word line is connected to the control gate of a memory cell on each corresponding NAND string in the block. The block is associated with source lines. Source-side select gates are used to connect or disconnect the NAND channel from the source lines.

[0005] Memory cells are programmed one group at a time. Programming cells are typically called pages. Typically, memory cells are programmed to multiple data states. Using a larger number of data states allows more bits to be stored per memory cell. For example, four data states can be used to store two bits per memory cell, eight data states to store three bits per memory cell, 16 data states to store four bits per memory cell, and so on. Some memory cells can be programmed to data states by storing charge in the memory cell. For example, the threshold voltage (Vt) of a NAND memory cell can be set as a target Vt by programming charge into a charge storage region such as a charge trapping layer. The amount of charge stored in the charge trapping layer determines the Vt of the memory cell.

[0006] Once a memory cell in a memory device has been programmed, data can be read from the memory cell by sensing its programming state. However, due to one or more factors, the sensed state can sometimes differ from the written state. Error detection and correction decoding can be used to detect and correct data errors caused by a mismatch between the sensed state and the written state. Typically, user data is encoded into error correction code (ECC) codewords before programming. Therefore, an ECC engine can be used to correct errors in the encoded user data. However, for the ECC algorithm to successfully correct all errors, there are limitations on how many bits in the data read from the memory cell might be erroneous. Therefore, storage systems typically rely on correcting errors not only in the ECC codewords.

[0007] Many memory systems employ one or more techniques to recover from block-level failures, such as failures to decode ECC codewords programmed into the block. One technique for recovering from block-level failures is to dynamically adjust (i.e., calibrate) the read reference voltage and then reread the memory cell using the calibrated read reference voltage. Another technique for recovering from block-level failures is to use XOR data to provide block-level redundancy. In addition to storing ECC codewords in a non-volatile memory system, the system can also compute and store XOR data. One technique is to accumulate XOR data based on successive exclusive OR (XOR) operations of data programmed into the memory cells of the block. For example, an initial bitwise XOR is performed between the first two data cells programmed into the memory cell to form an initial XOR accumulation. Then, as each additional data cell in the block is programmed, a bitwise XOR is performed between the current XOR accumulation and the new data cell being programmed. XOR data can be used to recover user data in cases where the ECC decoder fails to decode one or more ECC codewords in a block, which is referred to in this paper as unrecoverable by ECC decoder error (UECC). However, in the case of UECC, there are limitations on how much user data within a block can be recovered using XOR data.

[0008] Therefore, XOR data can be used for block-level redundancy to recover from block-level failures. Many storage systems can recover from such block-level failures, but not from die-level failures. Some storage systems will preemptively evict bad blocks to reduce the chance of die-level failures. In some cases, storage systems may adopt read-only mode to prevent die-level failures. In rare cases, die-level failures may result in data loss.

[0009] Other storage systems utilize redundancy coding, such as at the plane or die level, to provide recovery from die-level failures. An example of redundancy coding is often referred to as Redundant Array of Independent Disks (RAID). Redundancy coding can be used as an additional level of data integrity protection for data written to non-volatile storage systems. Redundancy coding typically involves storing data across multiple planes, platters, disks, or other areas in the event of a failure in a single plane, die, disk, etc., to increase data integrity. However, die-level redundancy reduces the storage capacity of the storage system. Furthermore, data recovery can be time-consuming because all data on the die needs to be read to recover unreadable data. Attached Figure Description

[0010] Elements with the same number represent common parts in different figures.

[0011] Figure 1 It is a block diagram depicting one implementation of a storage system.

[0012] Figure 2A This is a block diagram of one implementation scheme for a memory die.

[0013] Figure 2B This is a block diagram of one implementation scheme for an integrated memory component.

[0014] Figure 3A and Figure 3B Different implementation schemes for the integrated memory component are described.

[0015] Figure 3C It is a block diagram depicting one embodiment of a column control circuit that includes multiple read / write circuits.

[0016] Figure 4 This is a perspective view of a portion of an example implementation of a monolithic three-dimensional memory architecture.

[0017] Figure 4A This is a block diagram of one implementation of a memory structure with two planes.

[0018] Figure 4B It is a block diagram depicting a top view of a portion of a memory cell block.

[0019] Figure 4C An implementation scheme for stacking is described, showing along... Figure 4B The cross-sectional view of line AA.

[0020] Figure 4D Depicting Figure 4C View of section 445.

[0021] Figure 4E This is a schematic diagram of one implementation of a block, depicting several NAND strings.

[0022] Figure 5A and Figure 5B The threshold voltage distribution was depicted.

[0023] Figure 6 This is a flowchart describing one implementation of the process for programming a memory cell connected to a selected word line.

[0024] Figure 7 This is a flowchart of an implementation of the process of restoring a block that is a candidate to be eliminated.

[0025] Figure 8 It shows how the magnitude of the source line voltage relates to the location of the selected word line, the magnitude of the programming voltage, and whether the block has been identified as a candidate for elimination.

[0026] Figure 9 This is a flowchart of an implementation scheme for programming memory cells.

[0027] Figure 10 This is a block-level diagram showing the components attached to the control die of the memory structure die.

[0028] Figure 11 This is a flowchart of an implementation of the process of using block-level recovery to recover blocks that are candidates for elimination.

[0029] Figure 12 This is a flowchart of an implementation of the process of recovering blocks that are candidates for elimination using bare-chip level recovery. Detailed Implementation

[0030] Techniques for programming NAND memory are disclosed herein. The programming techniques disclosed herein can be used to recover from potential planar and / or die-level faults. Blocks containing NAND strings have different physical characteristics in different regions of the block, which can lead to a preference for different programming techniques in different regions. However, a programming technique preferred at one point in time may become problematic at another point in time due to defects, for example, developing outside the memory cell block. Therefore, defects may potentially lead to planar or die-level faults. Implementations of the memory system use a first programming technique when programming a target region of a memory block, provided that the first programming technique does not cause a large number of block-level faults in the target region. The memory system can detect evidence of planar or die-level faults affecting memory operations performed in the target region of the block. For example, the memory system can detect a large number of blocks with memory operation faults in the target region. Due to, for example, memory operation faults in the target region, some of these blocks can be identified as candidates for termination after using the first programming technique. Candidates for termination can be based on block-level faults, such as faults in decoding data pages read from the block. However, the memory system may attempt to continue using at least some of the blocks that are candidates for termination. For at least some of these blocks that are candidates for termination ("candidate blocks"), the storage system switches to a second programming technique when programming a region of the block. When programming that region of the block, the candidate blocks can be tested against reliability criteria after programming using the second programming technique. Candidate blocks that pass the reliability criteria after programming using the second programming technique can continue to be used, even if they are candidates for termination. For those candidate blocks that continue to be used, the storage system can use the second programming technique when programming that target region of the block. Furthermore, the storage system can also use the second programming technique when programming target regions of other blocks that may also be affected by planar or die-level issues.

[0031] In the implementation, the first programming technique includes applying a first voltage to the source line, and the second programming technique includes applying a second voltage to the source line. The second voltage may be lower than the first voltage. The first voltage can provide a higher NAND channel precharge than the second voltage, thereby potentially improving the prohibition of programming memory cells that will not be programmed. However, the circuitry providing the first voltage may be more susceptible to defects than the circuitry providing the second voltage. This defect can cause errors when the NAND channel is precharged using the first voltage. Such errors may render the block unusable. However, switching to a circuitry system that provides a second voltage for precharging the NAND channel allows continued use of the block.

[0032] Figure 1This is a block diagram of one embodiment of the storage system 100 implementing the techniques described herein. In one embodiment, the storage system 100 is a solid-state drive (“SSD”). The storage system 100 may also be a memory card, a USB drive, or other type of storage system. The proposed techniques are not limited to any one type of storage system. The storage system 100 is connected to a host 102, which may be a computer, server, electronic device (e.g., a smartphone, tablet, or other mobile device), electrical appliance, or another device that uses memory and has data processing capabilities. In some embodiments, the host 102 is separate from but connected to the storage system 100. In other embodiments, the storage system 100 is embedded within the host 102.

[0033] Figure 1 The components of the storage system 100 depicted are electrical circuits. The storage system 100 includes a memory controller 120 (or memory controller) connected to a non-volatile storage device 130 and a local high-speed memory 140 (e.g., DRAM, SRAM, MRAM). The local high-speed memory 140 is a non-transitory memory, which may include volatile or non-volatile memory. The memory controller 120 uses the local high-speed memory 140 to perform certain operations. For example, the local high-speed memory 140 may store a logic-to-physical address translation table (“L2P table”).

[0034] The memory controller 120 includes a host interface 152 that connects to and communicates with the host 102. In one embodiment, the host interface 152 implements Fast NVM (NVMe) via Fast PCI (PCIe). Other interfaces, such as SCSI, SATA, etc., may also be used. The host interface 152 is also connected to a network on-chip (NOC) 154. The NOC is a communication subsystem on an integrated circuit. The NOC can span synchronous and asynchronous clock domains or use non-clocked asynchronous logic. NOC technology applies networking theory and methods to on-chip communication, resulting in significant improvements over conventional buses and crossbar interconnects. Compared to other designs, the NOC improves the scalability of the system-on-chip (SoC) and the power efficiency of complex SoCs. The wires and links of the NOC are shared by many signals. Because all links in the NOC can operate simultaneously on different data packets, a high level of parallelism is achieved. Therefore, as the complexity of integrated subsystems continues to increase, the NOC provides enhanced performance, such as throughput, and scalability compared to previous communication architectures, such as dedicated point-to-point signal lines, shared buses, or segmented buses with bridges. In other implementations, NOC 154 may be replaced by a bus. Connected to and communicating with NOC 154 are processor 156, ECC engine 158, memory interface 160, and local memory controller 164. Local memory controller 164 is used to operate and communicate with local high-speed memory 140 (e.g., DRAM, SRAM, MRAM).

[0035] ECC engine 158 performs error correction services. For example, ECC engine 158 performs data encoding and decoding. In one embodiment, ECC engine 158 is a software-programmable electrical circuit. For example, ECC engine 158 may be a programmable processor. In other embodiments, ECC engine 158 is a custom-designed dedicated hardware circuit without any software. In yet another embodiment, the functionality of ECC engine 158 is implemented by processor 156.

[0036] Processor 156 performs various controller memory operations, such as programming, erasing, reading, and memory management processes. In one embodiment, processor 156 is programmed by firmware. In other embodiments, processor 156 is a custom-designed dedicated hardware circuit without any software. Processor 156 may also implement a translation module as software / firmware processing or dedicated hardware circuitry. In many systems, non-volatile memory is internally addressed to the memory system using physical addresses associated with one or more memory dies. However, the host system will use logical addresses to address various memory locations. This allows the host to assign data to consecutive logical addresses while the memory system is free to store data as desired between locations on one or more memory dies. To implement this system, memory controller 120 (e.g., translation module) performs address translation between logical addresses used by the host and physical addresses used by the memory dies. An example specific implementation maintains a table identifying the current translation between logical and physical addresses (i.e., the L2P table described above). Entries in the L2P table may include identifiers for logical addresses and their corresponding physical addresses. Although logical address to physical address tables (or L2P tables) include the word "table," they do not need to be tables in the literal sense. Instead, logical address to physical address tables (or L2P tables) can be any type of data structure. In some examples, the memory space of the storage system is so large that local memory 140 cannot hold all the L2P tables. In this case, the entire set of L2P tables is stored in storage device 130, and a subset of the L2P tables is cached (the L2P cache) in local memory 140.

[0037] Memory interface 160 communicates with non-volatile memory device 130. In one embodiment, the memory interface provides a switching mode interface. Other interfaces may also be used. In some example implementations, memory interface 160 (or another part of controller 120) implements a scheduler and buffer for sending data to one or more memory dies and receiving data from one or more memory dies.

[0038] In one embodiment, the non-volatile storage device 130 includes one or more memory dies. Figure 2A This is a functional block diagram of one embodiment of a memory die 200 including a non-volatile memory device 130. Each memory die in one or more memory dies of the non-volatile memory device 130 can be implemented as follows: Figure 2A 200 memory dies. Figure 2AThe components depicted are electrical circuits. Memory die 200 includes a memory structure 202 (e.g., a memory array) that may include non-volatile memory cells (also referred to as non-volatile memory units), as described in more detail below. The array terminal lines of memory structure 202 include various layers of word lines organized into rows and various layers of bit lines organized into columns. However, other orientations may also be implemented. Memory die 200 includes row control circuitry 220, the output of which is connected to a corresponding word line of memory structure 202. Row control circuitry 220 receives a set of M row address signals and one or more different control signals from system control logic circuitry 260, and typically includes circuitry such as a row decoder 222, an array driver 224, and a block select circuitry 226 for both read and write (programming) operations. Row control circuitry 220 may also include read / write circuitry. Memory die 200 also includes column control circuitry 210, which includes read / write circuitry 225. The read / write circuitry 225 may include a sense amplifier and a data latch. The sense amplifier inputs / outputs are connected to corresponding bit lines of the memory structure 202. Although only a single block is shown for structure 202, the memory die may include multiple arrays that can be accessed individually. The column control circuitry 210 receives a set of N column address signals and one or more various control signals from the system control logic unit 260, and typically includes circuitry such as a column decoder 212, array terminal receiver or driver circuitry 214, block selection circuitry 216, and read / write circuitry and I / O multiplexers. The system control logic 260, column control circuitry 210, and / or row control circuitry 220 are configured to control memory operations, such as open block reads, at the die level.

[0039] System control logic unit 260 receives data and commands from memory controller 120 and provides output data and status to the host. In some embodiments, system control logic unit 260 (which includes one or more electrical circuits) includes a state machine 262 that provides die-level control for memory operations. In one embodiment, state machine 262 is programmable by software. In other embodiments, state machine 262 does not use software and is implemented entirely in hardware (e.g., electrical circuits). In another embodiment, state machine 262 is replaced by a microcontroller or microprocessor on or outside the memory chip. System control logic unit 260 may also include a power control module 264 that controls the power and voltage supplied to rows and columns of memory structure 202 during memory operations. System control logic unit 260 includes a storage device 266 (e.g., RAM, registers, latches, etc.) that can be used to store parameters for operating memory structure 202.

[0040] Commands and data are transmitted between memory controller 120 and memory die 200 via memory controller interface 268 (also referred to as the "communication interface"). Memory controller interface 268 is an electrical interface used for communicating with memory controller 120. Examples of memory controller interface 268 include a switching mode interface and an Open NAND Flash Interface (ONFI). Other I / O interfaces may also be used.

[0041] In some embodiments, all elements of memory die 200 (including system control logic unit 260) may be formed as part of a single die. In other embodiments, some or all of the system control logic unit 260 may be formed on a different die than the die containing memory structure 202.

[0042] In one embodiment, memory structure 202 includes a three-dimensional memory array of non-volatile memory cells, wherein multiple memory levels are formed over a single substrate, such as a wafer. The memory structure may include any type of non-volatile memory formed monolithically in memory cells having one or more physical levels of active regions located over a silicon substrate (or other type of substrate). In one example, the non-volatile memory cells include vertical NAND strings with charge trapping layers.

[0043] In another embodiment, memory structure 202 includes a two-dimensional memory array of non-volatile memory cells. In one example, the non-volatile memory cells are NAND flash memory cells utilizing floating gates. Other types of memory cells (e.g., NOR flash memory) may also be used.

[0044] The exact type of memory array architecture or memory cell included in memory structure 202 is not limited to the examples above. Many different types of memory array architectures or memory technologies can be used to form memory structure 202. For the purposes of the embodiments of the new claims presented herein, no specific non-volatile memory technology is required. Other examples of suitable technologies for memory cells of memory structure 202 include ReRAM (Resistive Random Access Memory), magnetoresistive memory (e.g., MRAM, spin-orbit MRAM, spin-track MRAM), FeRAM, phase-change memory (e.g., PCM), etc. Examples of suitable technologies for memory cell architectures of memory structure 202 include two-dimensional arrays, three-dimensional arrays, cross-point arrays, stacked two-dimensional arrays, vertical bitline arrays, etc.

[0045] An example of a ReRAM crosspoint memory includes a reversible resistive switching element arranged as a crosspoint array accessed by X-lines and Y-lines (e.g., word lines and bit lines). In another embodiment, the memory cell may include a conductive bridge memory element. The conductive bridge memory element may also be referred to as a programmable metallized cell. The conductive bridge memory element can be used as a state-changing element based on the physical relocation of ions within a solid electrolyte. In some cases, the conductive bridge memory element may include two solid metal electrodes, one relatively inert (e.g., tungsten) and the other electrochemically active (e.g., silver or copper), with a solid electrolyte film between the two electrodes. As temperature increases, ion mobility also increases, resulting in a decrease in the programming threshold of the conductive bridge memory cell. Therefore, the conductive bridge memory element can have a wide range of programming thresholds that vary with temperature.

[0046] Another example is magnetoresistive random access memory (MRAM) that stores data via magnetic memory elements. This element is formed of two ferromagnetic layers, each of which can remain magnetized, separated by a thin insulating layer. One of these layers is a permanent magnet set to a specific polarity; the magnetization of the other layer can be changed to match the magnetization of an external field, thereby storing memory. Memory devices are constructed from a grid of such memory cells. In one embodiment for programming, each memory cell is located between a pair of write lines arranged perpendicular to each other, parallel to the cell, one above and one below. When current flows through the pair of write lines, an induced magnetic field is generated.

[0047] Phase-change memories (PCMs) utilize the unique behavior of chalcogenides. One embodiment uses a GeTe-Sb₂Te₃ superlattice to achieve a non-thermal phase transition by simply changing the coordination state of germanium atoms with a laser pulse (or a light pulse from another source). Therefore, the programming dose is the laser pulse. A memory cell can be disabled by blocking it from receiving light. In other PCM embodiments, the memory cell is programmed by a current pulse. Note that the use of "pulse" in this document does not require a square pulse, but includes (continuous or discontinuous) vibrations or bursts of sound, current, voltage, light, or other waves. These memory elements within a single selectable memory cell or bit may include another cascaded element as a selector, such as a bidirectional threshold switch or a metallic insulator substrate.

[0048] Those skilled in the art will recognize that the techniques described herein are not limited to a single particular memory structure, memory construction, or material composition, but encompass many related memory structures as described herein and as understood by those skilled in the art.

[0049] Figure 2AThe components can be divided into two parts: (1) memory structure 202 and (2) peripheral circuitry, which includes... Figure 2A All other components depicted herein. A key characteristic of memory circuitry is its capacity, which can be increased by increasing the area of ​​the memory die allocated to memory structure 202 in memory system 100; however, this reduces the area of ​​memory die available for peripheral circuitry. This can impose fairly strict limitations on these components of the peripheral circuitry. For example, the need to fit sense amplifier circuitry within the available area can significantly limit the sense amplifier design architecture. With regard to system control logic component 260, the reduced area availability can limit the available functionality that can be implemented on-chip. Therefore, the fundamental trade-off in the design of the memory die for memory system 100 is the amount of area dedicated to memory structure 202 and the amount of area dedicated to peripheral circuitry.

[0050] Another area where the memory structure 202 and peripheral circuitry are often inconsistent is the processing involved in forming these regions, as these regions typically involve different processing techniques and trade-offs in using different techniques on a single die. For example, when the memory structure 202 is NAND flash memory, it is an NMOS structure, while the peripheral circuitry is typically CMOS-based. For example, elements such as sense amplifier circuitry, charge pumps, logic elements in state machines, and other peripheral circuitry in system control logic unit 260 typically employ PMOS devices. The processing operations used to fabricate CMOS dies will differ in many ways from those optimized for NMOS flash NAND memory or other memory cell technologies. Three-dimensional NAND structures (e.g., see...) Figure 4 They can especially benefit from specialized processing operations.

[0051] To mitigate these limitations, the implementation scheme described below can... Figure 2AThe components are separated onto separately formed dies, which are then bonded together. More specifically, the memory structure 202 can be formed on a single die (referred to as the memory die), and some or all of the peripheral circuitry elements (including one or more control circuits) can be formed on separate dies (referred to as the control die). For example, the memory die can be formed solely of memory elements, such as flash NAND memory, MRAM memory, PCM memory, ReRAM memory, or other memory cell arrays of other memory types. Some or all of the peripheral circuitry (even including elements such as decoders and sense amplifiers) can then be moved to separate control dies. This allows each memory die in the memory die to be individually optimized according to its technology. For example, a NAND memory die can be optimized for an NMOS-based memory array structure without worrying about CMOS elements that have now been moved to a control die that can be optimized for CMOS processing. This frees up more space for peripheral elements, which can now incorporate additional capabilities that would otherwise be difficult to incorporate due to the limited margin of the same die for the memory cell array. Two dies can then be bonded together in a bonded multi-die memory circuit, with an array on one die connected to peripheral components on the other die. Although the following will focus on a bonded memory circuit with one memory die and one control die, other embodiments may use more dies, such as two memory dies and one control die.

[0052] Figure 2B It shows Figure 2A An alternative arrangement of the arrangement, which can be implemented using wafer-to-wafer bonding to provide bonded die pairs. Figure 2B A functional block diagram of one embodiment of the integrated memory component 207 is depicted. One or more integrated memory components 207 may be used to implement the non-volatile memory device 130 of the memory system 100. The integrated memory component 207 includes two types of semiconductor dies (or more simply, "dies"). The memory structure die 201 includes a memory structure 202. The memory structure 202 includes non-volatile memory cells. The control die 211 includes control circuitry 260, 210, and 220 (as described above). In some embodiments, the control die 211 is configured to be connected to the memory structure 202 within the memory structure die 201. In some embodiments, the memory structure die 201 and the control die 211 are bonded together.

[0053] Figure 2B An example of peripheral circuitry (including control circuitry) is shown, which is formed in the peripheral circuitry or control die 211 of the memory structure 202 formed in the memory structure die 201. General components and Figure 2ASimilarly labeled. System control logic unit 260, row control circuitry 220, and column control circuitry 210 are located in control die 211. In some embodiments, all or part of column control circuitry 210 and all or part of row control circuitry 220 are located on memory structure die 201. In some embodiments, some circuitry in system control logic unit 260 is located on memory structure die 201.

[0054] System control logic unit 260, row control circuitry 220, and column control circuitry 210 can be formed using a common process (e.g., CMOS process), so that additional elements and functionalities (such as ECC) more commonly found on memory controller 120 may require few or no additional process steps (i.e., the same process steps used to manufacture controller 120 can also be used to manufacture system control logic unit 260, row control circuitry 220, and column control circuitry 210). Therefore, while moving such circuitry from a die such as memory structure die 201 can reduce the number of steps required to manufacture such a die, adding such circuitry to a die such as control die 211 may not require many additional process steps. Control die 211 may also be referred to as a CMOS die because some or all of the control circuitry in control circuitry 260, 210, and 220 is implemented using CMOS technology.

[0055] Figure 2B A column control circuit 210, including read / write circuitry 225 on a control die 211, is shown. This column control circuitry is coupled to a memory structure 202 on a memory structure die 201 via an electrical path 206. For example, electrical path 206 may provide electrical connections between a column decoder 212, driver circuitry 214, and block select device 216 and the bit lines of the memory structure 202. The electrical path may extend from the column control circuitry 210 in the control die 211 through pads on the control die 211 that are bonded to corresponding pads on the memory structure die 201 that are connected to the bit lines of the memory structure 202. Each bit line of the memory structure 202 may have a corresponding electrical path in electrical path 206, including a pair of bonded pads connected to the column control circuitry 210. Similarly, a row control circuitry 220, including a row decoder 222, an array driver 224, and a block select device 226, is coupled to the memory structure 202 via an electrical path 208. Each electrical path 208 may correspond to a word line, a dummy word line, or a select gate line. Additional electrical paths may also be provided between the control die 211 and the memory structure die 201.

[0056] For the purposes of this document, the phrase "control circuitry" or "one or more control circuits" may include any or any combination of the following: memory controller 120, system control logic unit 260, row control circuitry 220, column control circuitry 210, read / write circuitry 225, sense amplifier, microcontroller, microprocessor, and / or other similar functional circuitry. Control circuitry may consist of hardware only or a combination of hardware and software (including firmware). For example, a controller programmed by firmware to perform the functions described herein is an example of control circuitry. Control circuitry may include processors, FPGAs, ASICs, integrated circuits, or other types of circuitry.

[0057] For the purposes of this document, the term "device" may include, but is not limited to, one or more of the following: storage system 100, memory controller 120, storage device 130, memory die 200, integrated memory assembly 207, and / or control die 211.

[0058] In some embodiments, the integrated memory component 207 contains more than one control die 211 and more than one memory structure die 201. In some embodiments, the integrated memory component 207 includes a stack of multiple control dies 211 and multiple memory structure dies 201. Figure 3A A side view of one embodiment of an integrated memory assembly 207 (e.g., a stack including control dies 211 and memory structure dies) stacked on a substrate 271 is depicted. The integrated memory assembly 207 has three control dies 211 and three memory structure dies 201. In some embodiments, there are more than three memory structure dies 201 and more than three control dies 211. Figure 3A In the original embodiment, there are an equal number of memory structure dies 201 and control dies 211; however, in one embodiment, there are more memory structure dies 201 than control dies 211. For example, one control die 211 can control multiple memory structure dies 201.

[0059] Each control die 211 is attached (e.g., bonded) to at least one memory structure die 201 within the memory structure die 201. Some of the bonding pads 282 / 284 are depicted. More bonding pads may be present. The space between the two bondsed dies 201, 211 is filled with a solid layer 280, which may be formed of epoxy resin or other resins or polymers. This solid layer 280 protects the electrical connection between the dies 201, 211 and further secures the dies together. Various materials may be used as the solid layer 280.

[0060] The integrated memory component 207 can be stacked, for example, in a stepped offset manner, so that the bonding pads at each level are not covered and can be accessed from above. Wire bonds 270 connected to the bonding pads connect the control die 211 to the substrate 271. This can span the width of each control die 211 (i.e., into...). Figure 3A (On the page) Multiple such lead wire connections are formed.

[0061] Memory die through-silicon vias (TSVs) 276 can be used to route signals through memory structure die 201. Control die through-silicon vias (TSVs) 278 can be used to route signals through control die 211. TSVs 276 and 278 can be formed before, during, or after the formation of integrated circuits in semiconductor dies 201 and 211. TSVs can be formed by etching holes through the wafer. The holes can then be lined with a barrier layer to prevent metal diffusion. The barrier layer can then be lined with a seed layer, and the seed layer can be plated with an electrical conductor such as copper, but other suitable materials such as aluminum, tin, nickel, gold, doped polysilicon, and alloys or combinations thereof can also be used.

[0062] Solder balls 272 may optionally be attached to contact pads 274 on the lower surface of substrate 271. Solder balls 272 can be used to electrically and mechanically couple the integrated memory assembly 207 to a host device such as a printed circuit board. Solder balls 272 may be omitted if the integrated memory assembly 207 is to be used as an LGA package. Solder balls 272 may form part of the interface between the integrated memory assembly 207 and memory controller 120.

[0063] Figure 3B A side view depicting another embodiment of an integrated memory assembly 207 stacked on a substrate 271. Figure 3B The integrated memory component 207 has three control dies 211 and three memory structure dies 201. In some embodiments, there are far more than three memory structure dies 201 and far more than three control dies 211. In this example, each control die 211 is coupled to at least one memory structure die 201. Optionally, the control die 211 may be coupled to two or more memory structure dies 201.

[0064] Some of the bonding pads 282 and 284 are shown. More bonding pads may be present. The space between the two bonded dies 201 and 211 is filled with a solid layer 280, which may be formed of epoxy resin or other resins or polymers. Figure 3A The example in the middle is the opposite. Figure 3BThe integrated memory component 207 in the memory structure does not have a stepped offset. A memory die through-silicon via (TSV) 276 can be used to route signals through the memory structure die 201. A control die through-silicon via (TSV) 278 can be used to route signals through the control die 211.

[0065] Solder balls 272 may optionally be attached to contact pads 274 on the lower surface of substrate 271. Solder balls 272 can be used to electrically and mechanically couple integrated memory assembly 207 to a host device such as a printed circuit board. Solder balls 272 may be omitted if integrated memory assembly 207 is to be used as an LGA package.

[0066] As briefly discussed above, the control die 211 and the memory structure die 201 can be bonded together. Bonding pads on each die 201, 211 can be used to bond the two dies together. In some embodiments, in a so-called Cu-Cu bonding process, the bonding pads are directly bonded to each other without solder or other additive materials. In a Cu-Cu bonding process, the bonding pads are controlled to be highly flat and formed in a highly controlled environment largely free of environmental particles that could otherwise deposit on the bonding pads and prevent tight bonding. Under such properly controlled conditions, the bonding pads are aligned and pressed together to form a bond based on surface tension. Such bonding can be formed at room temperature, although heat can also be applied. In embodiments using Cu-Cu bonding, the bonding pads can be approximately 5µm square and spaced apart from each other at a pitch of 5µm to 5µm. Although the process is referred to herein as Cu-Cu bonding, the term can also be applied when the bonding pads are formed of materials other than Cu.

[0067] When the area of ​​the bonding pads is small, it can be difficult to bond semiconductor dies together. The size and spacing of the bonding pads can be further reduced by providing a thin film layer on the surface of the semiconductor die, including the bonding pads. The thin film layer is disposed around the bonding pads. When the dies are placed together, the bonding pads can bond to each other, and the thin film layers on the corresponding dies can bond to each other. This type of bonding technique can be called hybrid bonding. In embodiments using hybrid bonding, the bonding pads can be approximately 5µm squares and spaced apart from each other at a pitch of 1µm to 5µm. Bonding techniques can be used to provide bonding pads with even smaller sizes and pitches.

[0068] Some embodiments may include a film on the surfaces of the bare wafers 201, 211. If no such film is initially provided, the space between the wafers may be underfilled with epoxy resin or other resins or polymers. The underfill material may be applied as a liquid, which then hardens into a solid layer. This underfilling step protects the electrical connections between the wafers 201, 211 and further secures the wafers together. Various materials may be used as underfill materials.

[0069] Figure 3C This is a block diagram depicting one embodiment of a column control circuitry 210 comprising multiple read / write circuits 225. Each read / write circuit 225 is divided into a sense amplifier 325 and a data latch 340. Management circuitry 330 controls the read / write circuits 225. Management circuitry 330 can communicate with state machine 262. In one embodiment, each sense amplifier 325 is connected to a corresponding bit line. Each bit line can be connected to one of a large number of different NAND strings at a given time. Select gates on the NAND strings can be used to connect NAND string channels to the bit lines.

[0070] Each sense amplifier 325 operates to provide voltage to one of these bit lines (see BL0, BL1, BL2, BL3) during programming, verification, erase, and read operations. The sense amplifiers are also used to sense the condition (e.g., data status) of memory cells in the NAND string connected to the respective sense amplifier.

[0071] Each sense amplifier 325 may have a sense node. During sensing, the sense node is charged to an initial voltage Vsense_init, such as 3V. The sense node is then connected to the bit line for the sensing time, and the amount of decay of the sense node is used to determine whether the memory cell is in an on or off state. The amount of decay of the sense node also indicates whether the current Icell in the memory cell exceeds a reference current Iref. A larger decay corresponds to a larger current. If Icell <= Iref, the memory cell is in an off state, and if Icell > Iref, the memory cell is in an on state. In an embodiment, the sense node has a capacitor that is pre-charged and then discharged for the sensing time.

[0072] Specifically, the comparator circuit 320 determines the attenuation amount by comparing the sense node voltage with the trigger voltage after the sensing time. If the sense node voltage attenuates below the trigger voltage Vtrip, the memory cell is in the ON state and its Vth is at or below the verification voltage. If the sense node voltage does not attenuate below Vtrip, the memory cell is in the OFF state and its Vth is above the verification voltage. For example, the comparator circuit 320 sets the sense node latch 322 to 0 or 1 based on whether the memory cell is in the ON or OFF state. The bits in the sense node latch 322 can also be used in a lock scan to determine whether to set the bit line voltage to a suppress or program enable level in the next programming cycle. The bits in the sense node latch 322 can also be used in a lock mode to determine whether to set the bit line voltage to the sense voltage or the lock voltage in a read operation.

[0073] Data latch 340 is coupled to sense amplifier 325 via local data bus 346. In this example, data latch 340 includes three latches (ADL, BDL, CDL) for each sense amplifier 325. More or fewer latches may be included in data latch 340. In one embodiment, each data latch 340 is used to store a bit to be stored into a memory cell for programming, and each data latch 340 is used to store a bit read from the memory cell for reading. In the three-bit-per-memory-cell embodiment, ADL stores bits for the lower data page, BDL stores bits for the middle data page, and CDL stores bits for the upper data page. Each read / write circuit 225 is connected to XDL latch 348 via XDL bus 352. In this example, transistor 336 connects local data bus 346 to XDL bus 352. I / O interface 332 is connected to XDL latch 348. The XDL latch 348 associated with the specific read / write circuit 225 acts as an interface latch for storing / latching data from the memory controller.

[0074] Management circuitry 330 performs calculations to determine the data to be stored in the sensed memory cell and stores the determined data in a set of data latches. Each set of data latches 340 stores data bits determined by management circuitry 330 during a read operation and data bits imported from data bus 334 during a programming operation, representing write data intended to be programmed into memory. I / O interface 332 provides an interface between XDL latches 348 and data bus 334.

[0075] During the read operation, the system operates under the control of state machine 262, which supplies different control gate voltages to the addressed memory cell. As the sensing circuit iterates through various predefined control gate voltages corresponding to different memory states supported by the memory in a step-by-step manner, the sensing circuit can be triggered at one of these voltages and provides a corresponding output from the sensing amplifier to the management circuit 330. At this time, the management circuit 330 determines the resulting memory state by considering the triggering event of the sensing circuit and information about the applied control gate voltage from the state machine. It then calculates the binary code of the memory state and stores the resulting data bits in the data latch 340.

[0076] During a memory cell programming or verification operation, data to be programmed (written data) is stored from the data bus 334 into a set of data latches 340 via XDL latch 348. The programming operation, under the control of state machine 262, applies a series of programming voltage pulses to the control gate of the addressed memory cell. In a process known as incremental step-pulse programming, the amplitude of each voltage pulse can be stepped from the previous programming pulse. In one embodiment, each programming voltage is followed by a verification operation to determine whether the memory cell has been programmed to the desired memory state. In some cases, management circuitry 330 monitors the readback memory state relative to the desired memory state. When both are consistent, management circuitry 330 sets the bit line in a programming suppression mode, such as by updating its latches. This prevents further programming of the memory cell coupled to the bit line, even if additional programming pulses are applied to its control gate.

[0077] Figure 4 This is a perspective view as part of an example embodiment of a monolithic three-dimensional memory array / structure that may include memory structure 202, which includes multiple non-volatile memory cells arranged as vertical NAND strings. For example, Figure 4 A portion 400 of a block of memory is shown. The depicted structure includes a set of bit lines BL positioned above a stack 401 of alternating dielectric and conductive layers. For illustrative purposes, one of the dielectric layers is labeled D. The conductive layers are labeled one of the following: SGD, WL, or SGS. The SGD conductive layer acts as a drain-side select line. The WL conductive layer acts as a word line. The SGS conductive layer acts as a source-side select line. For ease of illustration, the number of each of these conductive layers is finite. The number of alternating dielectric and conductive layers may vary based on specific implementation requirements. Below the alternating dielectric and word line layers is a source line layer SL. Memory vias are formed in the stack of alternating dielectric and conductive layers. For example, one of the memory vias is labeled MH. Note that in Figure 4In the illustration, the dielectric layers are depicted as transparent, allowing the reader to see the memory holes positioned within the stack of alternating dielectric and conductive layers. In one embodiment, NAND strings are formed by filling the memory holes with a material including a charge-trapping material to create vertical columns of memory cells. Each memory cell can store one or more bits of data. Further details of a three-dimensional monolithic memory array including memory structure 202 are provided below.

[0078] In one implementation, the block is operated as a plurality of “sub-blocks”. Each of these “sub-blocks” has a plurality of NAND strings. In the implementation, an isolation region (IR) divides the SGD layer into a plurality of SGD selection lines, each of which is used to select strings (e.g., a set of NAND strings). Figure 4 An example with one IR region and thus two strings is depicted. However, more than one IR region can exist, and therefore more than two strings. Optionally, the IR region can extend downward through all the alternating dielectric and conductive layers.

[0079] Figure 4A This is a block diagram illustrating an example organization of memory structure 202, which is divided into two planes 403 and 405. Each plane is then divided into M physical blocks. In one example, each plane has approximately 2000 physical blocks (or more briefly, "blocks"). However, different numbers of blocks and planes can also be used. In a "full block" implementation, a block of memory cells is the erase unit. That is, all memory cells in the block are erased together. In a "sub-block mode" implementation, a block is divided into sub-blocks, and these sub-blocks are the erase units. In the implementation, a block contains multiple word lines, where each sub-block contains a unique set of data word lines. Memory cells can also be grouped into blocks for other reasons, such as to organize the memory structure as enable signaling and select circuitry. In some implementations, a block represents a group of connected memory cells because the memory cells in a block share a common set of word lines. For example, all word lines for a block are connected to all vertical NAND strings for that block. Although Figure 4A Two planes 403 / 405 are shown, but more or fewer planes may be implemented. In some embodiments, memory structure 202 includes four planes. In some embodiments, memory structure 202 includes eight planes. In some embodiments, programming can be performed in parallel in a first selected block in plane 403 and a second selected block in plane 405.

[0080] Figures 4B to 4E Depicting and Figure 4 The structure corresponds to and can be used for implementation Figure 2A and Figure 2B Example three-dimensional (“3D”) NAND structure of memory structure 202. Figure 4B This is a top view depicting a portion 407 of block 2. (As shown from...) Figure 4B visible, Figure 4B The physical blocks depicted extend in the direction of arrow 433. In one implementation, the memory array has multiple layers; however, Figure 4B Only the top layer is shown.

[0081] Figure 4B Multiple circles representing vertical columns are depicted. Each vertical column includes multiple selection transistors (also referred to as select gates) and multiple memory cells. In one implementation, each vertical column implements a NAND string. For example, Figure 4B Vertical columns 422, 432, 442, and 452 are depicted. Vertical column 422 implements NAND string 482. Vertical column 432 implements NAND string 484. Vertical column 442 implements NAND string 486. Vertical column 452 implements NAND string 488. More details about the vertical columns are provided below. Due to... Figure 4B The physical block depicted extends in the direction of arrow 433, so the physical block includes more than... Figure 4B More vertical columns are depicted in the text.

[0082] Figure 4B It also depicts the set of bit lines 415, including bit lines 411, 412, 413, 414, ..., 419. Figure 4B Twenty-four bit lines are shown because only a portion of the physical block is depicted. It is conceivable that more than twenty-four bit lines are connected to the vertical columns of the physical block. Each circle representing a vertical column has an "x" to indicate its connection to a bit line. For example, bit line 414 is connected to vertical columns 422, 432, 442, and 452.

[0083] Figure 4B The depicted physical block comprises a set of isolation regions 402, 404, 406, 408, and 410, which are formed of SiO2; however, other dielectric materials may also be used. Isolation regions 402, 404, 406, 408, and 410 are used to divide the top layer of the physical block into four regions; for example, Figure 4BThe depicted top layer is divided into regions 420, 430, 440, and 450, which are referred to herein as “sub-blocks.” Each sub-block contains a large number of NAND strings. In one embodiment, isolation regions 402 and 410 separate physical block 407 from adjacent physical blocks. Therefore, isolation regions 402 and 410 can extend down to the substrate. In one embodiment, isolation regions 404, 406, and 408 divide only the layers used to implement the select gate, so that NAND strings in different strings can be selected independently. Return to Reference Figure 4 The IR region may correspond to any of the isolation regions 404, 406, or 408. In one example implementation, the bit line is connected to only one vertical column / NAND string in each of the regions (sub-blocks) 420, 430, 440, and 450. In this implementation, each physical block has sixteen rows of active columns, and each bit line is connected to four NAND strings in each block. In one implementation, all vertical column / NAND strings connected to the common bit line are connected to the same word line (or set of word lines); therefore, the system uses a drain-side select line (SGD) to select one (or another subset) of the four vertical column / NAND strings that will undergo memory operations (programming, verification, reading, and / or erasing).

[0084] although Figure 4B The diagram shows each zone (420, 430, 440, 450) with four vertical columns, and a block with four zones (420, 430, 440, 450) and sixteen vertical columns, but those exact numbers are for example implementation. Other implementations may include more or fewer zones (420, 430, 440, 450) per block, more or fewer vertical columns per zone, and more or fewer vertical columns per block. Figure 4B The diagram also shows staggered vertical columns. In other embodiments, different staggered patterns may be used. In some embodiments, the vertical columns are not staggered.

[0085] Figure 4C An example of stacking 435 is depicted, showing along... Figure 4B A cross-sectional view of line AA is shown. The SGD layers include SGDT0, SGDT1, SGD0, and SGD1. There may be more or fewer than four SGD layers. The SGS layers include SGSB0, SGSB1, SGS0, and SGS1. There may be more or fewer than four SGS layers. In addition to the data word line layers WL0 to WL111, six dummy word line layers DD0, DD1, WLIFDU, WLIDDL, DS1, and DS0 are provided. There may be more or fewer than 112 data word line layers and more or fewer than four dummy word line layers. Each NAND string has a drain-side selected gate at the SGD layer. Each NAND string has a source-side selected gate at the SGS layer. Dielectric layers DL0 to DL124 are also depicted.

[0086] Columns 432 and 434 of memory cells are depicted in a multilayer stack. The stack includes a substrate 457, an insulating film 454 on the substrate, and a portion of the source line SL. A portion of the bit line 414 is also depicted. Note that a NAND string 484 is connected to the bit line 414. The NAND string 484 has a source terminal at the bottom of the stack and a drain terminal at the top of the stack. The source terminal is connected to the source line SL. A conductive via 417 connects the drain terminal of the NAND string 484 to the bit line 414.

[0087] In one implementation, the memory cells are arranged as NAND strings. Word line layers WL0-WL111 are connected to the memory cells (also referred to as data memory cells). Dummy word line layers DD0, DD1, DS0, and DS1 are connected to the dummy memory cells. The dummy memory cells do not store and are not eligible to store host data (data provided from the host, such as data from a user of the host), while the data memory cells are eligible to store host data. In some implementations, the data memory cells and dummy memory cells may have the same structure. The drain-side select layer SGD is used to electrically connect and disconnect (or cut off) the channel of the corresponding NAND string from the bit line. The source-side select layer SGS is used to electrically connect and disconnect (or cut off) the channel of the corresponding NAND string from the source line SL.

[0088] Figure 4C An example of a stack 435 with two sub-blocks (SBs) is depicted. This two-SB stack includes SB0 and SB1. Dual SBs or other multi-SB stacks can be used to form relatively tall stacks while maintaining relatively narrow memory via widths (or diameters). Memory via portions are formed in the lower SB after the layers are formed. Subsequently, memory via portions are formed in the upper SB after the layers are formed, aligning with the memory via portions in the lower SB to form continuous memory vias from the bottom to the top of the stack. The resulting memory vias are narrower than if the vias were etched from the top to the bottom of the stack rather than individually in each SB. An interface (IF) region is created, connecting the two SBs at the IF region. The IF region is typically thicker than other dielectric layers. Due to the presence of the IF region, adjacent word line layers are affected by edge effects, such as programming or erasing difficulties. These adjacent word line layers can therefore be set as dummy word lines (WLIFDL, WLIFDU).

[0089] Figure 4D Depicting Figure 4CA view of region 445. Data memory cell transistors 520, 521, 522, 523, and 524 are indicated by dashed lines. Multiple layers may be deposited along the sidewalls (SW) of memory via 432 and / or within each word line layer, for example, using atomic layer deposition. For example, each column (e.g., a pillar formed by material within the memory via) may include a blocking oxide / blocking high-k material 470, a charge trapping layer or film 463 (such as SiN or other nitrides), a tunneling layer 464, a polysilicon body or channel 465, and a dielectric core 466. The word line layer may include a conductive metal 462 (such as tungsten) as a control gate. For example, control gates 490, 491, 492, 493, and 494 are provided. In this example, all layers except the metal are disposed within the memory via. In other methods, some layers may be within the control gate layer. Similarly, additional pillars are formed in different memory vias. The pillars may form columnar active regions (AA) of the NAND string.

[0090] When a data memory cell transistor is programmed, electrons are stored in the portion of the charge trapping layer associated with the data memory cell transistor. These electrons are attracted from the channel into the charge trapping layer and then pass through the tunneling layer. The Vth of the data memory cell transistor increases proportionally to the amount of charge stored. During an erase operation, the electrons return to the channel.

[0091] Each of these memory vias may be filled with multiple annular layers (also referred to as memory films), which include a blocking oxide layer, a charge trapping layer, a tunneling layer, and a channel layer. The core region of each of these memory vias is filled with the host material, and the multiple annular layers are located between the core region and the WLLs in each of these memory vias. In some cases, the tunneling layer 464 may include multiple layers, such as those configured in an oxide-nitride-oxide configuration.

[0092] Figure 4E This is a schematic diagram of a portion of memory array 202. Figure 4E The physical data word lines WL0 to WL111, which extend across the entire block, are shown. Figure 4E The structure corresponds to Figure 4A Section 407 of block 2 includes bit lines 411. In one implementation, within the physical block, each bit line is connected to four NAND strings. Therefore, Figure 4E Bit lines 411 connected to NAND strings NS0, NS1, NS2 and NS3 are shown.

[0093] In one implementation, four sets of drain-side select lines exist in the physical block. For example, the set of drain-side select lines connected to NS0 includes SGDT0-s0, SGDT1-s0, SGD0-s0, and SGD1-s0. The set of drain-side select lines connected to NS1 includes SGDT0-s1, SGDT1-s1, SGD0-s1, and SGD1-s1. The set of drain-side select lines connected to NS2 includes SGDT0-s2, SGDT1-s2, SGD0-s2, and SGD1-s2. The set of drain-side select lines connected to NS3 includes SGDT0-s3, SGDT1-s3, SGD0-s3, and SGD1-s3. Throughout this document, the term "SGD" may be used as a general term to refer to any one or more lines in the drain-side select line sets. In some implementations, the same operating voltage is applied to SGDT0 and SGDT1. In some implementations, the same operating voltage is applied to SGD0 and SGD1. In some erase implementations, a different operating voltage is applied to SGDT0 / SGDT1 instead of SGD0 / SGD1. Note that SGDT0 / SGDT1 is adjacent to the bit line. In some erase implementations, the voltage applied to SGDT0 / SGDT1, combined with the bit line voltage, can be used to generate gate-induced gate leakage (GIDL) current. Such voltages applied to SGDT0 / SGDT1 may be referred to herein as GIDL voltages.

[0094] In the implementation, each line in a given set can operate independently of the other lines in that set to allow different voltages to the gates of the four drain-side select transistors on the NAND string. Furthermore, each set of drain-side select lines can be selected independently of the other sets. Each set of drain-side select lines is connected to a group of NAND strings in the block. Figure 4E The diagram depicts only one NAND string for each group. These four sets of drain-side select lines correspond to four "sub-blocks." The first sub-block corresponds to the vertical NAND strings controlled by SGDT0-s0, SGDT1-s0, SGD0-s0, and SGD1-s0. The second sub-block corresponds to the vertical NAND strings controlled by SGDT0-s1, SGDT1-s1, SGD0-s1, and SGD1-s1. The third sub-block corresponds to the vertical NAND strings controlled by SGDT0-s2, SGDT1-s2, SGD0-s2, and SGD1-s2. The fourth sub-block corresponds to the vertical NAND strings controlled by SGDT0-s3, SGDT1-s3, SGD0-s3, and SGD1-s3. As described, Figure 4E Only the NAND strings connected to bit line 411 are shown. However, a full schematic of the block would show each bit line and the four vertical NAND strings connected to each bit line.

[0095] In one implementation, all memory cells on the NAND string within a physical block are erased as cells. However, in some implementations, the physical block is operated as a multi-level operation, where each level contains a contiguous set of word lines. For example, memory cells connected to WL0 through WL55 may be in one level, and memory cells connected to WL56 through WL111 may be in another level. A physical block may operate in more than two levels. Erasing cells may be formed based on other partitions of the physical block.

[0096] although Figures 4 to 4E The example memory is a three-dimensional memory structure comprising vertical NAND strings with charge-trapping material, but other 3D memory structures can also be used with the techniques described herein.

[0097] The memory system discussed above can be erased, programmed, and read. At the end of a successful programming process, the threshold voltage of the memory cell should be appropriately within one or more distributions of the threshold voltages of the programmed memory cell or within the distribution of the threshold voltages of the erased memory cell. Figure 5A This is a graph showing the threshold voltage versus the number of memory cells, and illustrates an example threshold voltage distribution for a memory array when each memory cell stores one data bit per memory cell. A memory cell that stores one data bit per memory cell is called a single-level cell (“SLC”). The data stored in an SLC memory cell is called SLC data; therefore, SLC data consists of one bit per memory cell. Data stored as one bit per memory cell is SLC data. Figure 5A Two threshold voltage distributions are shown: E and P. Threshold voltage distribution E corresponds to the erase data state. Threshold voltage distribution P corresponds to the program data state. Memory cells with threshold voltages in threshold voltage distribution E are therefore in the erase data state (e.g., they are erased). Memory cells with threshold voltages in threshold voltage distribution P are therefore in the program data state (e.g., they are programmed). In one embodiment, erased memory cells store data "1" and programmed memory cells store data "0". Figure 5A The reference voltage Vr is described. By testing (e.g., performing one or more sensing operations) whether the threshold voltage of a given memory cell is higher or lower than Vr, the system can determine whether the memory cell is erased (state E) or programmed (state P). Figure 5A The programmable verification reference voltage Vv is also described. In some implementations, when memory cells are programmed to data state P, the system will test whether those memory cells have a threshold voltage greater than or equal to Vv.

[0098] A memory cell that stores multiple bits of data per memory cell is called a Multi-Level Cell (“MLC”). The data stored in an MLC memory cell is called MLC data; therefore, MLC data comprises multiple bits per memory cell. Data stored as multiple data bits per memory cell is MLC data. Figure 5B In the example implementation, each memory cell stores three data bits. Other implementations may use other data capacities per memory cell (e.g., two, four, or five data bits per memory cell).

[0099] Figure 5B Eight threshold voltage distributions corresponding to eight data states are shown. The first threshold voltage distribution (data state) Er represents an erased memory cell. The other seven threshold voltage distributions (data states) A ​​through G represent programmed memory cells and are therefore also called programming states. Each threshold voltage distribution (data state) corresponds to a predetermined value for that set of data bits. The specific relationship between the data programmed into the memory cell and the cell's threshold voltage level depends on the data encoding scheme used for the cell. In one embodiment, Gray code allocation is used to assign data values ​​to a range of threshold voltages such that if the memory's threshold voltage is erroneously shifted to its adjacent physical state, only one bit will be affected. In this embodiment, the number of memory cells in each state is approximately the same.

[0100] Figure 5B Seven read reference voltages, VrA, VrB, VrC, VrD, VrE, VrF, and VrG, are shown for reading data from memory cells. By testing (e.g., performing a sensing operation) whether the threshold voltage of a given memory cell is higher or lower than the seven read reference voltages, the system can determine what data state the memory cell is in (i.e., A, B, C, D, ...). Figure 5B Several verification reference voltages are also shown. These verification reference voltages are VvA, VvB, VvC, VvD, VvE, VvF, and VvG. In some implementations, when memory cells are programmed to data state A, the system tests whether those memory cells have a threshold voltage greater than or equal to VvA. If a memory cell has a threshold voltage greater than or equal to VvA, the memory cell is locked and cannot be further programmed. Similar reasoning applies to other data states.

[0101] Figure 6This is a flowchart describing one embodiment of the process for programming memory cells connected to a selected word line. Programming a memory cell connected to a word line is referred to herein as a programming word line. For the purposes of this document, the term programming is synonymous with writing. The process includes multiple loops, each of which includes a programming phase and a verification phase. Typically, a programming voltage applied to the control gate (via the selected data word line) during a programming operation is applied as a series of programming pulses (e.g., voltage pulses). Between the programming pulses is a set of verification pulses (e.g., voltage pulses) used to perform verification. In many specific embodiments, the magnitude of the programming pulses increases by a predetermined step with each successive pulse. Figure 6 In step 601, the programming voltage signal (Vpgm) is initialized to an initial amplitude (e.g., approximately 12V to 16V or another suitable level), and the programming counter PC maintained by state machine 262 is initialized to 1. In one embodiment, a group of memory cells selected for programming (referred to herein as selected memory cells) is programmed concurrently and all connected to the same word line (selected word line). There may be other memory cells not selected for programming (unselected memory cells), which are also connected to the selected word line. That is, the selected word line will also be connected to memory cells that should be prevented from programming. Additionally, further programming of these memory cells is prevented when they reach their intended target data state. Those NAND strings (e.g., unselected NAND strings) connected to the selected word line that are to be prevented from programming have their channels boosted to prevent programming. When the channel has a boosted voltage, the voltage difference between the channel and the word line is not large enough to induce programming. Step 602 includes determining the source line voltage. The magnitude of the source line voltage may depend on which region of the block (e.g., WL) is being programmed, the magnitude of the programming voltage, and / or whether evidence of planar or die-level defects has been detected in the memory system, necessitating the use of different magnitudes of the source line voltage. The magnitude of the source line voltage can affect the pre-charge of the NAND channel of an unselected NAND string. To aid in the boost, in step 604, the system pre-charges the channel of the NAND string, including the memory cells connected to the selected word lines to be disabled for programming. Step 604 may include applying a voltage to the source line while simultaneously turning on a source-side selection transistor to deliver the source line voltage to the channel of the NAND string. Reference Figure 4E The diagram shows how to apply appropriate voltages to SGS1, SGS0, SGSB1, and SGSB0 to turn on the source-side selection transistors.

[0102] In step 606, the NAND string of memory cells connected to the selected word lines, which is to be disabled for programming, has its channel boosted to disable programming. Such NAND strings are referred to herein as "unselected NAND strings". In one embodiment, at least some of the unselected word lines receive one or more boost voltages (e.g., about 7 to 11 volts) to perform a boost scheme. A programming disable voltage is applied to the bit lines coupled to the unselected NAND string.

[0103] In step 608, a programming voltage pulse of the programming voltage signal Vpgm is applied to the selected word line (the word line selected for programming). If a memory cell on the NAND string should be programmed, the corresponding bit line is biased to the programming enable voltage. In step 608, programming pulses are concurrently applied to all memory cells connected to the selected word line, such that all memory cells connected to the selected word line are programmed concurrently (unless programming is disabled). That is, they are programmed at the same time or during an overlap period (both are considered concurrent). In this way, all memory cells connected to the selected word line will concurrently change their threshold voltage unless programming is disabled.

[0104] In step 610, programming verification is performed, and the die is locked to a memory cell that has reached its target state, preventing further programming. Step 610 includes performing programming verification by sensing at one or more verification reference levels. In one embodiment, the verification process is performed by testing whether a threshold voltage of the selected memory cell for programming has reached an appropriate verification reference voltage. In step 610, the memory cell may be locked after it has been verified (by the test of Vt) that it has reached its target state. For example, if the memory cell reaches the verification reference voltage, it can be locked. In one embodiment, one or more data latches in management circuitry 330 are used to indicate whether the memory cell is locked or will receive full programming.

[0105] If it is determined in step 612 that all memory cells have reached their target threshold voltage (pass), the programming process is complete and successful because all selected memory cells have been programmed and verified to their target state. The status "pass" is reported in step 614. Otherwise, if it is determined in step 612 that not all memory cells have reached their target threshold voltage (failure), the programming process continues to step 616.

[0106] In step 616, the number of memory cells that have not yet reached their corresponding target threshold voltage distribution is counted. That is, the number of memory cells that have failed to reach their target state so far is counted. This counting can be performed by state machine 262, memory controller 120, or another circuit. In one embodiment, there is a total count that reflects the total number of currently programmed memory cells that have failed the previous verification step. In another embodiment, a separate count is maintained for each data state.

[0107] In step 618, it is determined whether the count from step 616 is less than or equal to a predetermined limit. In one embodiment, the predetermined limit is the number of bits that can be corrected by an error correction code (ECC) during a page read operation targeting a memory cell. If the number of failed cells is less than or equal to the predetermined limit, the programming process can stop and a "pass" status is reported in step 614. In this case, enough memory cells are correctly programmed so that ECC can be used during the read operation to correct the few remaining memory cells that have not yet been fully programmed. In some embodiments, the predetermined limit used in step 618 is lower than the number of bits that can be corrected by an error correction code (ECC) during the read operation to allow for future / additional errors. The predetermined limit may be a portion (proportional or non-proportional) of the number of bits that can be corrected by ECC during a page read operation targeting a memory cell. In some embodiments, the limit is not predetermined. Instead, the limit varies based on the number of errors counted against the page, the number of program-erase cycles performed, or other criteria.

[0108] If the number of failed memory cells is not less than the predetermined limit, the programming process continues at step 620 and the programming counter PC is checked against the programming limit value (PL). Examples of programming limit values ​​include 6, 12, 16, 19, 20, and 30; however, other values ​​may be used. If the programming counter PC is not less than the programming limit value PL, the programming process is considered to have failed, and the failure status is reported in step 624. If the programming counter PC is less than the programming limit value PL, the process continues at step 626, during which the programming counter PC increments by 1 and the programming voltage signal Vpgm gradually increases to the next amplitude. For example, the next pulse will have an amplitude with a larger step size ∆Vpgm than the previous pulse (e.g., a step size of 0.1 volts to 1.0 volts). After step 626, the process loops back to step 604, and another programming pulse is applied to the selected word line, causing execution to... Figure 6 Another iteration of the programming process (steps 604 to 626).

[0109] Figure 7This is a flowchart of one embodiment of a process 700 for changing programming techniques in response to evidence of planar or die-level defects. Process 700 may be executed by one or more control circuits, such as memory controller 120, system control logic 260, row control circuitry 220, and / or column control circuitry 210. Process 700 can be used to operate NAND memory cells in a NAND memory architecture, such as, but not limited to… Figures 4 to 4E Those depicted herein. Step 702 includes programming the block using a first programming technique within a target region of the block. The target region of the block may be certain word lines within the block. In one embodiment, the target region is a set of word lines adjacent to the source lines. Figure 4C In this context, the set of word lines adjacent to the source line begins with WL0. The set of word lines adjacent to the source line may comprise a contiguous set of word lines (e.g., WL0 to WL20). However, this set may include more or fewer than 21 word lines. A first programming technique may include applying a specific voltage to the source line when programming a memory cell of one of the word lines connected to a target region. In some embodiments, this specific voltage is higher than a nominal voltage and is applied only on higher-numbered programming cycles. In other words, in one embodiment, this voltage higher than nominal is applied only when the programming voltage is higher than a threshold.

[0110] Between steps 702 and 704, a block can be read once or multiple times. Additionally, the block can be erased. Step 704 is to determine evidence of the presence of planar or die-level defects. The memory system may also determine whether potential planar or die-level defects could affect memory operation in the target area. In one embodiment, the memory system determines whether a threshold number of blocks have exhibited memory operation failures in the target area, which could be evidence of planar or die-level defects. Planar or die-level defects can be defects outside the block of memory cells, in which case the defect could affect any or all blocks in the planar or die. Defects can affect the delivery of voltage or other signals to the block of memory cells. For example, a defect may affect the delivery of voltage to source lines, one or more word lines, SGD lines, SGS lines, bit lines, etc.

[0111] Step 704 may include determining whether a block is a candidate for termination. A block may be a candidate for termination because it fails to meet reliability criteria regarding programming, reading, and / or erasing operations within the block. Reliability criteria are used to determine whether a block can continue to be reliably programmed and read from. For example, a block may fail to meet reliability criteria due to a read failure within the block. An exemplary read failure is the failure to decode an ECC codeword using an ECC decoder. Reliability criteria may allow for some failures in decoding ECC codewords within the block. Another example of a reliability criterion is a post-programmed read test, which reads an ECC codeword exactly after programming. Another way to express the decision in step 704 is whether the block meets failure criteria. Failure criteria may include, but are not limited to, a certain number of failures in decoding ECC codewords within the block. Even if some blocks are candidates for termination, this does not necessarily indicate evidence of a plane-level or die-level defect affecting memory operations in the target area. Therefore, the memory system may identify a subset of candidate blocks exhibiting failures in the target area. For example, the failure might be a failure to decode data programmed within the target area.

[0112] Step 706 includes programming a subset of blocks within the target area using a second programming technique. The second programming technique may include applying a lower voltage to the source lines than used in the first programming technique. In an embodiment, this lower source line voltage is used throughout all programming cycles. In an embodiment, for the second programming technique, the memory system selects a different source to provide voltage to control lines in the memory structure. For example, the memory system may select a voltage generator different from the one used in the first programming technique to provide voltage to specific control lines used for the second programming technique (for at least some programming cycles when programming the target area). Step 706 may be used to attempt to recover blocks that are candidates for elimination. Here, "recovering a block" means continuing to program data in the block even if it is a candidate for elimination. The memory system may wait until a certain number of blocks meet the conditions for attempting to recover blocks before attempting recovery. For example, recovery may be attempted in response to the detection of evidence of planar / die-level defects.

[0113] Step 708 includes determining whether the planar / die-level defect has been resolved. Step 708 may include determining whether memory operations performed within the block after programming with the second programming technique have passed. These memory operations may include programming, reading, and / or erasing. If the planar / die-level defect has been resolved, the memory system may continue to use the second programming technique while programming the target area (step 710). Note that the second programming technique can be used for all blocks in the planar or die to overcome the planar / die-level defect.

[0114] If the planar / die-level defects have not yet been resolved, then planar / die-level inspection and disposal is performed in step 712. Step 712 may include determining whether there are a sufficient number of good blocks to continue using the planar or die. If not, one option is to remove the data from the planar or die. Another option is to enter read-only mode.

[0115] In some implementations, the amount of voltage applied to the source line during programming depends on the location of the selected word line, the amount of the programming voltage, and whether the block has been identified as a candidate to be eliminated. Figure 8 This diagram illustrates how the magnitude of the source line voltage relates to these factors. The X-axis represents the programming voltage magnitude. The Y-axis represents the word line location. The graph is divided into four regions. Lower word lines (closer to the source line) may be more susceptible to programming interference, especially when the programming voltage is higher. Using a higher source line voltage can result in better pre-charging of the NAND channel, thus reducing programming interference. However, using a higher source line voltage will generally consume more power than using a lower source line voltage. For regions with lower WL and lower programming voltage, use a lower source line voltage. For regions with higher WL and lower programming voltage, use a lower source line voltage. For regions with higher WL and higher programming voltage, use a lower source line voltage. For regions with lower WL and higher programming voltage, depending on the state of the block, either a lower or higher source line voltage can be used. A higher source line voltage can be used if fewer than a threshold number of blocks in the plane or die exhibit problems in the target region. Using a higher source line voltage when using a higher programming voltage can improve programming by, for example, providing better pre-charging of the channel of the unselected NAND string, thus reducing programming interference. However, if a large number of blocks are exhibiting memory operational failures in the target area, a smaller source line voltage can be used. Using a smaller source line voltage has the potential to improve programming. One possible reason for improved programming is the existence of a defect that allows a smaller voltage to be delivered to the source line more efficiently than a larger voltage. In some implementations, a voltage source providing a smaller voltage has better DC drive capability than a voltage source providing a larger voltage. However, other factors may affect voltage delivery.

[0116] Figure 9 This is a flowchart of one implementation of the process for programming memory cells. This process can be used for NAND memory (such as, but not limited to, [other types of] memory) Figures 4 to 4E The process involves programming the NAND memory depicted in the diagram. This process is based on... Figure 8 The source line voltage is depicted in the diagram. This process describes one loop of the programming process. (Example) Figure 6 As indicated during the programming process, the programming voltage can be gradually increased in each programming cycle. Step 902 includes determining whether the selected word line is within the target region. Figure 8 In the example, the target region is the lower word line (e.g., the word line closer to the source line). If the selected word line is not in the target region, a lower source line voltage is used during programming in step 904. If the selected word line is in the target region, step 906 includes determining whether the programming voltage is greater than or equal to a threshold voltage. Here, the threshold voltage refers to a programming voltage value for which it has been determined (online or offline) that a larger source line voltage is advantageous. If the programming voltage is below the threshold voltage, a lower source line voltage is used during programming in step 904. If the programming voltage is at or above the threshold voltage, it is determined in step 908 whether to use a lower or higher source line voltage. In an embodiment, a higher source line voltage will be used provided that the memory system has not yet detected evidence of planar or die-level defects that could affect memory operation in the target region. For example, a higher source line voltage can be used if the memory system has not yet identified the number of faulty blocks in the target region. In an embodiment, a lower source line voltage can be used as an attempt to recover blocks in response to the number of faulty blocks in the target region. Furthermore, lower source line voltages can be used to reduce programming interference to any block within a block in a plane or die. It should be noted that the memory system will not always use lower source line voltages when block-level faults are present. Determining whether to use lower source line voltages can be based on which regions (e.g., word lines) have faults that lead to block-level faults. Following the determination in step 908, word lines are programmed depending on the source line voltage, as in steps 904 or 910. It should also be noted that once a block has been recovered using lower source line voltages (step 904), the memory system can continue to use lower source line voltages thereafter.

[0117] As mentioned above, circuits that provide smaller and larger source line voltages may differ. Figure 10This is a block-level diagram showing the components attached to a control die 211 of a memory structure die 201. The memory structure die 201 has a memory structure 202 that includes NAND strings 1002 extending through alternative conductive and insulating layers. The control die 211 and the memory structure die 201 are bonded to a first set of bonding pads 1006 on the memory structure die 201 and a second set of bonding pads 1007 on the control die. The control die 211 has a first voltage generator (voltage generator A) 1010A and a second voltage generator (voltage generator B) 1010B. Voltage generator A 1010A is configured to provide a smaller source line voltage, and voltage generator B 1010B is configured to provide a larger source line voltage. Each voltage generator 1010A, 1010B is connected to a driver 1012 such that one of the voltages can be provided to the source line (SL) on the memory structure die 201. Peripheral circuitry 1014 controls voltage generators 1010A and 1010B and driver 1012. Voltage from driver 1012 travels through metal layers 1008 on control die 211, bonding pads 1007 and 1006, and metal layers 1004 on memory structure die 201. It should be noted that voltage generator B 1010B is not merely a copy or redundant version of voltage generator A 1010A. Additionally, it should be noted that in some cases, both voltage generators 1010A and 1010B are used for a single programming operation, where voltage generator B 1010B is used to reduce programming interference during a portion of the programming operation, such as when a large programming voltage is applied.

[0118] Assuming no defects, the voltage from voltage generators 1010A and 1010B can be supplied to the source line without any problems. However, over time, a defect may exist or develop that prevents voltage from being supplied from one of the voltage generators 1010A and 1010B. This defect may occur within driver 1012, within one of voltage generators 1010A and 1010B, or somewhere along the path to the source line. One possible defect is a short circuit (Vss) along the path to ground. For example, driver 1012 may be shorted to ground. In an embodiment, the drive strength of the two voltage generators 1010A and 1010B is not the same. In an embodiment, voltage generator A 1010A (providing a lower voltage) has a greater drive strength than voltage generator B 1010B (providing a higher voltage). When voltage generator A 1010A is still able to provide a lower voltage to the source line, defects such as short circuits may prevent voltage generator B 1010B from providing a sufficiently higher voltage to the source line. Insufficiently high voltage from voltage generator B 1010B to the source line may result in weaker channel precharge during programming, leading to greater programming interference. However, switching to voltage generator A 1010A (providing a lower voltage) still provides sufficient precharge during programming, resulting in less programming interference than under these conditions with voltage generator B 1010B. It should be noted that continued use of voltage generator B 1010B under these defective conditions may lead to die-level failure. However, switching to voltage generator A 1010A allows for block recovery as described above and also prevents die-level failure. Although defects are described with respect to two dies 211 and 201, similar problems and defects can occur in a single die architecture (see [link to documentation]). Figure 2A (Memory die 200 in the middle).

[0119] Figure 11 This is a flowchart of one embodiment of a process 1100 for recovering blocks as candidates for elimination using block-level recovery. Process 1100 provides further details of one embodiment of process 700. Process 1100 can be performed by one or more control circuits (e.g., one or more of memory controller 120, system control logic 260, column control circuitry 220, and / or row control circuitry 210). Process 1100 can be used to operate NAND memory cells in a NAND memory architecture, such as, but not limited to… Figures 4 to 4E Those described in it.

[0120] Step 1102 includes performing a memory access operation. The memory access operation may include, for example, programming, reading, or erasing. If the memory operation is a programming operation, a first programming technique is used in the target region. The first programming technique may include applying a higher source line voltage to a higher programming voltage when programming in the target region. The first programming technique may include applying a higher-than-nominal precharge to the NAND channel to a higher programming voltage when programming in the target region. The higher-than-nominal precharge and the higher source line voltage are relative to a nominal precharge that can be used outside the target region and a lower source line voltage (and a lower programming voltage within the region). If the memory operation is a read, the first programming technique is used in the target region for the purposes of discussion. If the memory operation is an erase, the first programming technique is used in the target region for the purposes of discussion.

[0121] Step 1104 includes determining whether the memory access operation passed or failed. This determination can be performed on a page-by-page basis. For a read operation, the failure might be the inability to successfully decode an ECC codeword. For a programming operation, the failure might be related to checking data integrity immediately after the programming operation. The storage system can read the data immediately after the programming operation to confirm that the data was correctly programmed. Therefore, in step 1104, the storage system can determine whether any page in the programmed or read page has a fault. If the memory access operation passed, another memory access operation can be performed (step 1102). If the memory access operation failed (e.g., at least one page had a fault), block-level data recovery is performed in step 1106. Examples of block-level data recovery include, but are not limited to, dynamically adjusting the read reference level and / or block-level XOR recovery. Step 1108 includes a branch depending on whether the data was recovered. If the data was recovered, preemptive block elimination can be performed in step 1110. Preemptive block elimination can place blocks into a pool of newly grown bad blocks (GBBs). In an implementation, only those blocks that fail to meet reliability criteria are placed into the GBB pool. Reliability criteria may include, for example, a certain number of failures in decoding the ECC codewords in the block. As explained below, this may depend on the test in step 1122 to remove the block from the GBB. An alternative description of step 1110 is to identify the block as a candidate for removal, awaiting an attempt to recover it using a new programming technique in step 1120. Since the block may be recovered later in process 1100, it may be referred to as a candidate for removal only at this stage. If no data is recovered in step 1106 (no in step 1108), uncorrectable data loss occurs (step 1112). Step 1112 may also include placing the block into the GBB pool or identifying the block as a candidate for removal.

[0122] Following step 1110 or 1112, the process continues at step 1114, which includes determining whether a fault exists in the target region of the block. For example, the storage system determines whether a fault exists that could program or read pages within the target region of the block. If the way the block is programmed in the target region is changed, the block can be recovered. However, if the storage system determines that no significant fault exists in the target region, the storage system does not attempt to recover the block. It should be noted that the absence of a fault in the target region can be interpreted as the absence of evidence of a plane / die-level defect affecting memory operation in the target region. Therefore, the block remains in the GBB pool (or alternatively, moves from a candidate for obsolescence to an actual obsolescence). The process then returns to step 1102 for another memory access operation in a different block.

[0123] However, if a fault exists within the target area, in step 1116, the storage system increments the fault count. The fault count is the count of such blocks that have exhibited faults within the target area. A sufficiently high number of faulty blocks within the target area can be evidence of planar or die-level defects. Step 1118 includes determining whether the fault count in the target area is greater than or equal to a target limit. The target limit can be set to any desired limit (e.g., 10 blocks, 20 blocks, etc.). The target limit can be as low as one block or as high as 20 blocks. The target limit can be set to a value that suggests a planar or die-level fault may be occurring. For example, drive 1012 (see...) Figure 10 A defect at the plane or die level can affect all blocks in the plane or die. Therefore, a target limit can be set to capture such a plane or die-level fault before too many blocks are affected. If the target count is not met, the block remains on the GBB list. However, note that this block can be recovered later using the steps described below.

[0124] If the target count is met, in step 1120, the memory system is changed to an alternative programming technique. For example, an alternative programming precharge technique is used. The alternative programming precharge technique may be to use a nominal precharge across the entire block and for all programming voltages (as opposed to using a precharge higher than nominal for the target area when using a higher programming voltage). The precharge technique may be related to the source line voltage, where a higher precharge is achieved with a larger source line voltage, and a lower precharge is achieved with a lower source line voltage. In some embodiments, voltage generator A1110A is used for the nominal precharge (smaller source line voltage), and voltage generator B1110B is used for a precharge greater than nominal (larger source line voltage). In embodiments, this alternative programming precharge technique is applied to all blocks in the plane or die because there may be defects (e.g., short circuits) that could affect all blocks in the plane or die. Step 1120 may include applying the alternative programming technique (e.g., the alternative precharge technique) to any block previously identified as a candidate for recovery. For example, alternative programming techniques can be used to test any block on the GBB list that also has a fault in the target region (see step 1114).

[0125] Step 1122 includes testing whether the block has recovered after using the alternative programming precharge technique. This test may perform programming, reading, and / or erasing operations and determine whether such operations pass or fail. Step 1122 can be performed for all blocks that have undergone the alternative programming precharge technique. Step 1124 includes moving all blocks recovered from the GBB list to the good block pool. Alternatively, step 1124 can be stated as recovering plane-level or die-level faults. Furthermore, the recovery of blocks when using the second programming technique provides evidence that the second programming technique can be used for other blocks in the plane or die. Therefore, the second programming technique is now available for all good blocks in the plane or die.

[0126] Despite alternative programming precharge techniques, blocks may not be recovered (step 1122 is no). In this case, such blocks remain on the GBB list. Step 1126 determines if there are enough good blocks. If not, in step 1128, a plane-level or die-level fault action is taken. For example, readable data from the die can be moved to another die. Another option is to use the die in read-only mode. If there are enough good blocks, the process returns to step 1102 to perform another memory access operation on the good blocks.

[0127] Figure 12This is a flowchart of one embodiment of process 1200 for recovering blocks as candidates for termination using die-level recovery. Some steps in process 1200 are similar to their corresponding steps in process 1100 and will therefore not be described in detail. Step 1202 includes performing a memory access operation. Step 1204 includes determining whether the memory access operation is successful or unsuccessful (e.g., any faulty pages). Step 1206 includes block-level data recovery (e.g., dynamic read level, block-level XOR). Step 1210 includes preemptive block termination, which may include adding the block to the GBB list. Step 1214 includes determining whether a fault exists in the target region of the block. For example, step 1214 includes determining whether a page cannot be read or written in a word line near the source line. Step 1216 increments the target region fault count, which may be a count of blocks with faults in the target region. Step 1218 includes determining whether the target region fault count exceeds a target limit. This target limit may be similar to the target limit discussed in step 1128 (see [link to step 1218]). Figure 11 Step 1220 involves modifying the alternative programming technique, which can be similar to step 1120 (see step 1120). Figure 11 Step 1222 determines whether the block has been recovered due to the use of an alternative programming technique. Step 1224 includes moving the block from the GBB list to the good block list in response to the block being recovered. Alternatively, step 1224 can be stated as recovering a plane or die-level fault. Step 1226 includes testing whether the GBB count is greater than the preemptive die retirement (PDR) limit. If the GBB count is not greater than the PDR limit, the process returns to step 1202 for another memory access operation. However, if the GBB count is greater than the PDR limit, preemptive die retirement is performed in step 1228. Step 1230 may include die-level (or plane-level) data recovery (e.g., XOR, RAID, etc.). The data recovered from the faulty plane / die can then be moved to another plane / die. The plane or die can then be retired from further use.

[0128] In view of the foregoing, an embodiment includes an apparatus comprising one or more control circuits configured to be connected to a three-dimensional memory structure having blocks of memory cells in a plane on a die. The blocks are associated with source lines. The one or more control circuits are configured to program a target region within the block in the plane using a first programming technique. The one or more control circuits are configured to, in response to detecting evidence of a plane-level or die-level fault causing a memory operation failure in the target region of the block, program the target region within at least a subset of the block in the plane using a second programming technique.

[0129] In another embodiment of the device, the one or more control circuits are also configured to determine evidence of a planar or die-level fault in response to a threshold number of blocks failing to meet reliability criteria for memory operation failures within a target area.

[0130] In another embodiment of the device, the one or more control circuits are further configured to select a first circuit for providing a first voltage to the control lines in the block for a first programming technique, and a second circuit for providing a second voltage to the control lines in the block for a second programming technique.

[0131] In another embodiment of the device, the one or more control circuits are further configured to test a subset of the blocks against a reliability criterion after programming a target region within a subset of the blocks using the second programming technique. Furthermore, the one or more control circuits are also configured to continue using the second programming technique in the target region for blocks within the subset that pass the reliability criterion.

[0132] In another embodiment of the device, the one or more control circuits are further configured to program the target region within all good blocks in the plane using the second programming technique in response to overcoming the planar or die-level fault in the target region.

[0133] In another embodiment of the device, the first programming technique includes applying a first voltage to the source line for a set of programming voltages, and the second programming technique includes applying a second voltage to the source line for the set of programming voltages. The second voltage is less than the first voltage.

[0134] In another embodiment of the device, the one or more control circuits further include a first voltage generator configured to output a first voltage. The first voltage generator has a first DC drive strength. The one or more control circuits further include a second voltage generator configured to output a second voltage. The second voltage generator has a second DC drive strength greater than the first DC drive strength. The one or more control circuits are also configured to select the first voltage generator prior to evidence of a planar or die-level fault, and to select the second voltage generator in response to evidence of a planar or die-level fault.

[0135] In another embodiment of the device, the first programming technique includes applying a first precharge voltage to the NAND channel in the block, and the second programming technique includes applying a second precharge voltage to the NAND channel in the block, the second precharge voltage being less than the first precharge voltage.

[0136] In another embodiment of the device, the one or more control circuits further include a first voltage generator configured to provide a first electronic precharge to the NAND channel in the block of the plane. The first voltage generator has a first DC drive strength. The one or more control circuits further include a second voltage generator configured to provide a second electronic precharge to the NAND channel in the block of the plane, the second electronic precharge being less than the first electronic precharge. The second voltage generator has a second DC drive strength greater than the first DC drive strength. The one or more control circuits are configured to use the first voltage generator to provide the first electronic precharge and to use the second voltage generator to provide the second electronic precharge.

[0137] In another embodiment of the device, the target region of the block includes a set of word lines adjacent to the source line.

[0138] One embodiment includes a method for operating a NAND memory. The method includes applying a first voltage to a source line in the NAND memory while programming a set of word lines in a block within a plane of the NAND memory. The method includes identifying blocks that have block-level faults after programming the set of word lines while applying the first voltage to the source line. The method includes identifying a subset of blocks with block-level faults that meets fault criteria regarding memory operations performed within the set of word lines. The method includes applying a second voltage to the source line while programming the set of word lines in the subset of blocks, in response to the subset of blocks comprising a threshold number of blocks.

[0139] One embodiment includes a non-volatile memory system. The system includes a three-dimensional NAND memory structure having blocks. Each block includes a NAND string above a source line. Each NAND string has a NAND channel. The system includes: a first voltage generator configured to output a first voltage; and a second voltage generator configured to output a second voltage. The system includes one or more control circuitry communicating with the three-dimensional NAND memory structure, the first voltage generator, and the second voltage generator. The one or more control circuitry is configured to pre-charge the NAND channel of an unselected NAND string in the block using the first voltage generator for a programming voltage set when programming a set of word lines in the block. The one or more control circuitry is configured to identify blocks as candidates to be retired after programming, the programming including using the first voltage generator for the programming voltage set when programming the set of word lines in the block. The one or more control circuitry is configured to identify a subset of the block as candidates to be retired, the subset not meeting reliability criteria within the set of word lines. The one or more control circuits are configured to, when programming the set of word lines that is a subset of candidates to be eliminated, precharge the NAND channel of the unselected NAND string that is a subset of candidates to be eliminated using the second voltage from the second voltage generator for the set of programming voltages.

[0140] For the purposes of this document, references to “implementation scheme,” “one implementation scheme,” “some implementation schemes,” or “another implementation scheme” in the specification may be used to describe different implementation schemes or the same implementation scheme.

[0141] For the purposes of this document, a connection can be a direct connection or an indirect connection (e.g., via one or more other parts). In some cases, when an element is referred to as being connected or coupled to another element, the element may be directly connected to that other element or indirectly connected to it via one or more intermediate elements. When an element is referred to as being directly connected to another element, there are no intermediate elements between the element and the other element. If two devices are directly or indirectly connected, the two devices are "in communication," allowing electronic signals to be transmitted between the two devices.

[0142] For the purposes of this document, the term "based on" may be understood as "at least partially based on".

[0143] For the purposes of this document, the use of numerical terms such as “first” object, “second” object, and “third” object without additional context may not imply an order of objects, but may be used for identification purposes to distinguish different objects.

[0144] For the purposes of this document, the term "set of objects" may refer to a collection of one or more objects.

[0145] The foregoing detailed description has been given for purposes of illustration and description. It is not intended to be exhaustive or limiting to the exact forms disclosed. Many modifications and variations are possible in accordance with the foregoing teachings. The described embodiments were chosen to best explain the principles of the proposed technology and its practical application, thereby enabling others skilled in the art to fully utilize the technology in the various embodiments and with various modifications suitable for the specific intended use. The scope is intended to be defined by the claims appended herein.

Claims

1. An apparatus, the apparatus comprising: One or more control circuits configured to be connected to a three-dimensional memory structure having blocks of memory cells in a plane on a die, the blocks being associated with source lines, the one or more control circuits being configured to: The target region within the block in the plane is programmed using a first programming technique; as well as In response to the detection of evidence of a plane-level or die-level fault causing a memory operation failure in the target region of the block, the target region within at least a subset of the block in the plane is programmed using a second programming technique.

2. The apparatus of claim 1, wherein the one or more control circuits are further configured to: Evidence of a plane- or die-level fault is determined in response to a threshold number of blocks failing to meet reliability criteria regarding memory operation failures within the target region.

3. The apparatus of claim 1, wherein the one or more control circuits are further configured to: For the first programming technique, a first circuit is selected for providing a first voltage to the control lines in the block; and A second circuit is selected for providing a second voltage to the control line in the block, in accordance with the second programming technique.

4. The apparatus of claim 1, wherein the one or more control circuits are further configured to: After programming the target region within the subset of the block using the second programming technique, the subset of the block is tested against reliability standards; and For blocks in the subset that pass the reliability criteria, the second programming technique continues to be used in the target area.

5. The apparatus of claim 4, wherein the one or more control circuits are further configured to: In response to overcoming the planar or die-level faults using the second programming technique within the target area, the target area within all good blocks in the planar area is programmed using the second programming technique.

6. The apparatus according to claim 1, wherein: The first programming technique includes applying a first voltage to the source line for a set of programming voltages; and The second programming technique includes applying a second voltage to the source line for the set of programming voltages, the second voltage being less than the first voltage.

7. The apparatus according to claim 6, wherein: The one or more control circuits further include: A first voltage generator, configured to output the first voltage, the first voltage generator having a first DC drive strength; A second voltage generator, configured to output the second voltage, having a second DC drive strength greater than the first DC drive strength; and The one or more control circuits are also configured to select the first voltage generator prior to the evidence of the planar or die-level fault, and to select the second voltage generator in response to the evidence of the planar or die-level fault.

8. The apparatus according to claim 1, wherein: The first programming technique includes applying a first pre-charge voltage to the NAND channel in the block; and The second programming technique includes applying a second precharge voltage to the NAND channel in the block, the second precharge voltage being less than the first precharge voltage.

9. The apparatus according to claim 8, wherein: The one or more control circuits further include: A first voltage generator, configured to provide a first electronic precharge to the NAND channel in the block in the plane, the first voltage generator having a first DC drive strength; A second voltage generator, configured to provide a second electronic precharge less than the first electronic precharge to the NAND channel in the block in the plane, the second voltage generator having a second DC drive strength greater than the first DC drive strength; and The one or more control circuits are configured to use the first voltage generator to provide the first electronic precharge and to use the second voltage generator to provide the second electronic precharge.

10. The apparatus of claim 1, wherein the target region of the block comprises a set of word lines adjacent to the source line.

11. A method for operating a NAND memory, the method comprising: While programming the set of word lines in a block in the plane of the NAND memory, a first voltage is applied to the source lines in the NAND memory; While applying the first voltage to the source line, identify blocks with block-level faults after programming the word line set; Identify a subset of the blocks that have the block-level faults, the subset satisfying the fault criteria for memory operations performed within the word line set; as well as In response to a subset of the block comprising a threshold number of blocks, a second voltage is applied to the source line while programming the set of word lines in the subset of the block.

12. The method according to claim 11, further comprising: While applying the second voltage to the source line in response to the subset of the block, the subset of the block is tested against reliability criteria after the word line set has been programmed; as well as When programming the word line set while applying the second voltage to the source line, blocks that meet the reliability criteria in the subset of the block continue to be used despite the presence of the block-level fault.

13. The method according to claim 12, further comprising: In response to the subset of the block satisfying the reliability criterion after programming the word line set while applying the second voltage to the source line, the second voltage is applied to the source line while programming the word line set in all the blocks in the plane.

14. The method according to claim 11, further comprising: Before the block of the NAND memory has the block-level fault, the second voltage is applied to the source line of the NAND memory while programming word lines other than the word line set in the block; as well as In response to a subset of the blocks of the NAND memory having the block-level fault and the set of word lines in the block meeting the fault criteria, the second voltage is applied to the source lines in the NAND memory when programming word lines other than the set of word lines in the block.

15. The method according to claim 14, wherein: Applying the first voltage to the source lines in the NAND memory while programming the set of word lines in the block of the NAND memory includes providing the first voltage from a first voltage source having a first drive strength; and Applying the second voltage to the source lines in the NAND memory while programming the word line set in the subset of the block includes providing the second voltage from a second voltage source having a second drive strength greater than the first drive strength.

16. A non-volatile storage system, the system comprising: A three-dimensional NAND memory structure having blocks, each block including NAND strings above a source line, each NAND string having a NAND channel; A first voltage generator, configured to output a first voltage; A second voltage generator, configured to output a second voltage; and One or more control circuits, the one or more control circuits communicating with the three-dimensional NAND memory structure, the first voltage generator and the second voltage generator, wherein the one or more control circuits are configured to: When programming the set of word lines in the block, the NAND channels of the unselected NAND strings in the block are pre-charged with the first voltage from the first voltage generator for the set of programming voltages. Identify blocks as candidates to be eliminated after programming, the programming including, when programming the set of word lines in the block, using the first voltage from the first voltage generator for the set of programming voltages; A subset of the blocks are identified as candidates to be eliminated, and the subset does not meet the reliability criteria within the word line set; as well as When programming the set of word lines that is a subset of candidates to be eliminated, the NAND channels of the unselected NAND strings that are a subset of candidates to be eliminated are precharged with respect to the set of programming voltages using the second voltage from the second voltage generator.

17. The non-volatile memory system of claim 16, wherein the one or more control circuits are further configured to: The subset of candidates to be eliminated, after being programmed with the word line set using the second voltage, are placed into a pool by the reliability criteria. The pool remains eligible for use, provided that the second voltage from the second voltage generator continues to be used to program the word line set.

18. The non-volatile memory system of claim 16, wherein the one or more control circuits are further configured to: Blocks from the subset of candidates to be eliminated that failed to pass the reliability criteria after the word line set was programmed using the second voltage will be placed in a pool of unqualified blocks.

19. The non-volatile memory system according to claim 16, wherein: The first voltage generator has a first DC drive strength; and The second voltage has a second DC drive strength that is greater than the first DC drive strength.

20. The non-volatile memory system according to claim 16, wherein: The first voltage generator provides a first electronic precharge to the NAND channel of the unselected NAND string; and The second voltage generator provides a second electronic precharge to the NAND channel of the unselected NAND string, the second electronic precharge being less than the first electronic precharge.