US20190138412A1 - Apparatuses and methods for repairing memory devices including a plurality of memory die and an interface - Google Patents
Apparatuses and methods for repairing memory devices including a plurality of memory die and an interface Download PDFInfo
- Publication number
- US20190138412A1 US20190138412A1 US15/808,771 US201715808771A US2019138412A1 US 20190138412 A1 US20190138412 A1 US 20190138412A1 US 201715808771 A US201715808771 A US 201715808771A US 2019138412 A1 US2019138412 A1 US 2019138412A1
- Authority
- US
- United States
- Prior art keywords
- memory
- die
- semiconductor dies
- signal
- circuit
- Prior art date
- Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
- Granted
Links
Images
Classifications
-
- G—PHYSICS
- G11—INFORMATION STORAGE
- G11C—STATIC STORES
- G11C8/00—Arrangements for selecting an address in a digital store
- G11C8/12—Group selection circuits, e.g. for memory block selection, chip selection, array selection
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F11/00—Error detection; Error correction; Monitoring
- G06F11/07—Responding to the occurrence of a fault, e.g. fault tolerance
- G06F11/16—Error detection or correction of the data by redundancy in hardware
- G06F11/20—Error detection or correction of the data by redundancy in hardware using active fault-masking, e.g. by switching out faulty elements or by switching in spare elements
- G06F11/2053—Error detection or correction of the data by redundancy in hardware using active fault-masking, e.g. by switching out faulty elements or by switching in spare elements where persistent mass storage functionality or persistent mass storage control functionality is redundant
- G06F11/2094—Redundant storage or storage space
-
- G—PHYSICS
- G11—INFORMATION STORAGE
- G11C—STATIC STORES
- G11C29/00—Checking stores for correct operation ; Subsequent repair; Testing stores during standby or offline operation
- G11C29/006—Checking stores for correct operation ; Subsequent repair; Testing stores during standby or offline operation at wafer scale level, i.e. wafer scale integration [WSI]
-
- G—PHYSICS
- G11—INFORMATION STORAGE
- G11C—STATIC STORES
- G11C29/00—Checking stores for correct operation ; Subsequent repair; Testing stores during standby or offline operation
- G11C29/02—Detection or location of defective auxiliary circuits, e.g. defective refresh counters
- G11C29/022—Detection or location of defective auxiliary circuits, e.g. defective refresh counters in I/O circuitry
-
- G—PHYSICS
- G11—INFORMATION STORAGE
- G11C—STATIC STORES
- G11C29/00—Checking stores for correct operation ; Subsequent repair; Testing stores during standby or offline operation
- G11C29/02—Detection or location of defective auxiliary circuits, e.g. defective refresh counters
- G11C29/025—Detection or location of defective auxiliary circuits, e.g. defective refresh counters in signal lines
-
- G—PHYSICS
- G11—INFORMATION STORAGE
- G11C—STATIC STORES
- G11C29/00—Checking stores for correct operation ; Subsequent repair; Testing stores during standby or offline operation
- G11C29/04—Detection or location of defective memory elements, e.g. cell constructio details, timing of test signals
- G11C29/08—Functional testing, e.g. testing during refresh, power-on self testing [POST] or distributed testing
- G11C29/12—Built-in arrangements for testing, e.g. built-in self testing [BIST] or interconnection details
- G11C29/38—Response verification devices
-
- G—PHYSICS
- G11—INFORMATION STORAGE
- G11C—STATIC STORES
- G11C29/00—Checking stores for correct operation ; Subsequent repair; Testing stores during standby or offline operation
- G11C29/04—Detection or location of defective memory elements, e.g. cell constructio details, timing of test signals
- G11C29/08—Functional testing, e.g. testing during refresh, power-on self testing [POST] or distributed testing
- G11C29/12—Built-in arrangements for testing, e.g. built-in self testing [BIST] or interconnection details
- G11C29/44—Indication or identification of errors, e.g. for repair
- G11C29/4401—Indication or identification of errors, e.g. for repair for self repair
-
- G—PHYSICS
- G11—INFORMATION STORAGE
- G11C—STATIC STORES
- G11C29/00—Checking stores for correct operation ; Subsequent repair; Testing stores during standby or offline operation
- G11C29/04—Detection or location of defective memory elements, e.g. cell constructio details, timing of test signals
- G11C29/08—Functional testing, e.g. testing during refresh, power-on self testing [POST] or distributed testing
- G11C29/12—Built-in arrangements for testing, e.g. built-in self testing [BIST] or interconnection details
- G11C29/46—Test trigger logic
-
- G—PHYSICS
- G11—INFORMATION STORAGE
- G11C—STATIC STORES
- G11C29/00—Checking stores for correct operation ; Subsequent repair; Testing stores during standby or offline operation
- G11C29/70—Masking faults in memories by using spares or by reconfiguring
- G11C29/78—Masking faults in memories by using spares or by reconfiguring using programmable devices
- G11C29/80—Masking faults in memories by using spares or by reconfiguring using programmable devices with improved layout
- G11C29/816—Masking faults in memories by using spares or by reconfiguring using programmable devices with improved layout for an application-specific layout
-
- G—PHYSICS
- G11—INFORMATION STORAGE
- G11C—STATIC STORES
- G11C29/00—Checking stores for correct operation ; Subsequent repair; Testing stores during standby or offline operation
- G11C29/70—Masking faults in memories by using spares or by reconfiguring
- G11C29/88—Masking faults in memories by using spares or by reconfiguring with partially good memories
- G11C29/883—Masking faults in memories by using spares or by reconfiguring with partially good memories using a single defective memory device with reduced capacity, e.g. half capacity
-
- G—PHYSICS
- G11—INFORMATION STORAGE
- G11C—STATIC STORES
- G11C5/00—Details of stores covered by group G11C11/00
- G11C5/06—Arrangements for interconnecting storage elements electrically, e.g. by wiring
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F2201/00—Indexing scheme relating to error detection, to error correction, and to monitoring
- G06F2201/805—Real-time
-
- G—PHYSICS
- G11—INFORMATION STORAGE
- G11C—STATIC STORES
- G11C29/00—Checking stores for correct operation ; Subsequent repair; Testing stores during standby or offline operation
- G11C29/04—Detection or location of defective memory elements, e.g. cell constructio details, timing of test signals
- G11C29/08—Functional testing, e.g. testing during refresh, power-on self testing [POST] or distributed testing
- G11C29/12—Built-in arrangements for testing, e.g. built-in self testing [BIST] or interconnection details
- G11C2029/4402—Internal storage of test result, quality data, chip identification, repair information
Definitions
- High data reliability, high speed of memory access, lower power consumption and reduced chip size are features that are demanded from semiconductor memory.
- 3D memory devices have been introduced. Some 3D memory devices are formed by stacking memory dies (or “dice”) vertically and interconnecting the memory dies using through-silicon (or through-substrate) vias (TSVs). Benefits of the 3D memory devices include shorter interconnects which reduce signal delays and power consumption, a large number of vertical vias between layers which allow wide bandwidth buses between functional blocks in different layers, and a considerably smaller footprint. Thus, the 3D memory devices contribute to higher memory access speed, lower power consumption and chip size reduction.
- Example 3D memory devices include Hybrid Memory Cube (HMC) and High Bandwidth Memory (HBM).
- HBM High Bandwidth Memory
- HBM High Bandwidth Memory
- DRAM dynamic random access memory
- DRAM dynamic random access memory
- the memory device After assembly of the stacked memory dies and interface die, the memory device cannot be easily disassembled in the event one or more of the stacked memory dies become inoperable.
- a memory die may become inoperable for a variety of reasons, for example, one or more of the circuits of the memory die do not function properly, memory cells of the memory die are defective and cannot correctly and/or adequately store data, etc.
- each die typically includes a limited amount of redundant memory circuits for repairing defective memory cells, the redundant memory circuits may not be capable of repairing all defective memory cells of the memory die, and additionally, inoperable circuits of a memory die may not be repairable.
- the entire memory device is scrapped, even if several of or all other memory dies in the stack and the interface are operable.
- FIG. 1 is a wiring diagram of a memory device and a processor according to an embodiment of the disclosure.
- FIG. 2A is a schematic diagram of a memory device stack including an interface (I/F) die and a plurality of core dies.
- FIG. 2B is a schematic diagram of a portion of the memory device stack of FIG. 2A .
- FIG. 2C is a schematic diagram of the memory device stack of FIG. 2A including the I/F die and the plurality of core dies.
- FIG. 3A is a schematic diagram of a memory device stack including an interface (I/F) die and a plurality of core dies according to an embodiment of the disclosure.
- FIG. 3B is a schematic diagram of the memory device stack of FIG. 3A including the I/F die and the plurality of core dies according to an embodiment of the disclosure.
- FIG. 4A is a wiring diagram of the memory device stack of FIG. 2A including an I/F die and a plurality of core dies.
- FIG. 4B is a wiring diagram of the memory device stack of FIG. 3A including an I/F die and a plurality of core dies according to an embodiment of the disclosure.
- FIG. 5 is a block diagram of a memory device according to an embodiment of the disclosure.
- FIG. 6 is a block diagram of the memory device of FIG. 5 having two inoperable channels.
- FIG. 7 is a block diagram of the memory device of FIG. 5 having the two inoperable channels of FIG. 6 and showing an operable channel of one stack group replacing an inoperable channel of another stack group according to an embodiment of the disclosure.
- FIG. 8 is a schematic diagram of circuits included in the memory device of FIG. 5 according to an embodiment of the disclosure.
- FIG. 9 is a schematic diagram of a repair circuit according to an embodiment of the disclosure.
- FIG. 1 is a wiring diagram of a memory device 1 and a processor 2 according to an embodiment of the disclosure.
- the memory device 1 is a High Bandwidth Memory (HBM).
- the processor 2 may be a graphical processor unit or a memory controller.
- the memory device 1 may include terminals coupled by balls 3 (e.g., microbumps) to an interposer 5 .
- the processor 2 may include terminals coupled by balls 4 (e.g., microbumps) to the interposer 5 and further to the corresponding terminals of the memory device 1 through the interposer 5 .
- the interposer may be stacked on a packaging substrate (not shown) by balls 6 .
- the interposer 5 may be made of silicon.
- FIG. 2A is a schematic diagram of a memory device stack including an interface (I/F) die 22 and a plurality of DRAM core dies 23 .
- the number of the plurality of core dies 23 in the memory device stack 21 may be four.
- Each of the core dies 23 may be a memory die including a memory array for storing data and further including circuits for performing memory operations.
- FIG. 2B is a schematic diagram of a portion of the memory device stack 21 .
- the I/F die 22 and the plurality of core dies 23 may be coupled by a plurality of conductive vias 27 (e.g., through silicon (substrate) via (TSV)).
- TSV through silicon
- FIG. 2C is a schematic diagram of the memory device stack 21 including the I/F die 22 and the plurality of core dies 23 .
- the memory device stack 21 may have two 128-bit channels per core die for a total of eight input/output channels and a width of 1024 bits in total.
- each core die of the plurality of the core dies 23 may include two channels.
- a channel of a core die represents a separable addressable memory space that may be accessed to read data from the channel and to write data to the channel.
- the core dies 23 a, 23 b, 23 c and 23 d include Channels A and C, Channels B and D, Channels E and G, and Channels F and H, respectively.
- a clock frequency, a command sequence, and data can be independently provided for each channel.
- FIG. 4A is a wiring diagram of the memory device stack 21 including the I/F die 22 and the plurality of core dies 23 .
- the I/F die 22 of the memory device 21 provides interfaces 28 a, 28 b, 28 e and 28 f which provide signals on four input/output channels among the eight input/output channels, which function independently of each other.
- Memory arrays of the Channel A, Channel B, Channel E and Channel F of the core dies 23 a, 23 b, 23 c and 23 d may be coupled to the I/F die 22 via native input/output lines (IOs) 27 a, 27 b, 27 e and 27 f, respectively.
- the native IOs 27 a to 27 f may be implemented as conductive vias.
- each core die 23 may include a command circuit for each channel.
- the core dies 23 a to 23 d may include command circuits 26 a to 26 d for Channel A, Channel B, Channel E and Channel F, respectively.
- clock signals, command signals and data signals for each channel may be transmitted independently and a plurality of data buses and their respective channels can operate individually.
- FIG. 3A is a schematic diagram of a memory device stack 31 that may also be the memory device 1 of FIG. 1 , including an interface (I/F) die 32 and a plurality of memory core dies 33 such as DRAM dies according to an embodiment of the disclosure.
- the core dies 33 may be stacked with one another.
- plurality of core dies 33 in the memory device stack 31 are stacked and may include eight core dies.
- FIG. 3B is a schematic diagram of the memory device stack 31 including the I/F die 32 and the plurality of core dies 33 according to an embodiment of the disclosure.
- the memory device stack 31 may have two 128-bit channels per core die for a total of eight input/output channels and a width of 1024 bits in total.
- each core die of the plurality of the core dies 33 may include two channels.
- the memory device stack 31 may be logically divided into multiple stack groups.
- a stack group 34 a having a stack identifier (SID) “0” includes the core dies 33 a, 33 b, 33 c and 33 d including Channels A and C, Channels B and D, Channels E and G, and Channels F and H, respectively.
- a stack group 34 b having a stack ID (SID) “1” includes the core dies 33 e, 33 f, 33 g and 33 h including Channels A and C, Channels B and D, Channels E and G, and Channels F and respectively.
- a destination dies among a plurality of core dies in each channel (e.g., core dies 33 a and 33 e of Channel A) addressed in a command may be identified by the SID.
- the core dies 33 a, 33 b, 33 c, and 33 d may represent a “stack” and the core dies 33 e, 33 f, 33 g, and 33 h may represent another stack that is stacked on the stack of core dies 33 a, 33 b, 33 c, and 33 d.
- FIG. 4B is a wiring diagram of the memory device stack 31 including the I/F die 32 and the plurality of core dies 33 according to an embodiment of the disclosure.
- the I/F die 32 of the memory device 31 provides interfaces 38 a, 38 b, 38 e and 38 f which provide signals on four input/output channels among the eight input/output channels of two stack groups.
- Memory arrays of Channels A, B, E and F of the stack group 34 a and memory arrays of Channels A, B, E and F of the stack group 34 b may be coupled to the same native input/output lines (IOs) 37 a, 37 b, 37 e and 37 f, respectively.
- IOs native input/output lines
- memory arrays of Channel A of the core die 33 a in the stack group 34 a and memory arrays of Channel A of the core die 33 e in the stack group 34 b may be coupled to the native IO 37 a.
- Each core die 33 may include a command circuit for each channel.
- the core dies 33 a to 33 d in the stack group 34 a may include command circuits 36 a to 36 d for Channel A, Channel B, Channel E and Channel F, respectively.
- the core dies 33 e to 33 h in the stack group 34 b may include command circuits 36 e to 36 h for Channel A, Channel B, Channel E and Channel F, respectively.
- Each command circuit 36 may detect the SID in a command, check whether the SID in the command matches with an SID of the stack group of the core dies 33 including the command circuit 36 , and decode the command if the SID matches and memory access actions responsive to the command may be performed. For example, when the interface 38 a transmits a command on the input/output line 37 a, the command circuit 36 a receives the command and check whether the SID in the command is “0”. The command circuit 36 a processes the command if the SID is “0” and ignores the command if the SID is “1”. The command circuit 36 e also receives the command and check whether the SID in the command is “1”. The command circuit 36 e processes the command if the SID is “1” and ignores the command if the SID is “0”. Thus, clock signals, command signals and data signals for each channel on each die may be transmitted independently.
- FIG. 5 is a block diagram of a memory device 500 according to an embodiment of the disclosure.
- the memory device 500 may be a memory device stack in some embodiments of the disclosure.
- the memory device 500 includes an interface (I/F) die 502 and a plurality of core dies 503 .
- Each of the core dies 503 may be a memory die including a memory array for storing data and further including circuits for performing memory operations.
- the memory device 500 is shown in FIG. 5 as including eight core dies 503 a - 503 h. However, greater or fewer core dies may be included in other embodiments of the disclosure.
- the memory device 500 may have multiple channels.
- each core die 503 of the memory device 500 may include two channels.
- the core die 503 a includes Channels 0 and 2
- the core die 503 b includes Channels 1 and 3
- the core die 503 c includes Channels 4 and 6
- the core die 503 d includes Channels 5 and 7 .
- the core die 503 e includes Channels 0 and 2
- the core die 503 f includes Channels 1 and 3
- the core die 503 g includes Channels 4 and 6
- the core die 503 h includes Channels 5 and 7 .
- the Channels 0 - 7 are coupled to the interface die 502 through signal lines 527 .
- signal lines 527 may include conductive vias, for example, through-silicon (or through-substrate) vias (TSVs).
- the core dies 503 are logically arranged into two stack groups.
- a stack group 504 a having a stack identifier (SID) “ 0 ” includes the core dies 503 a, 503 b, 503 c, and 503 d, and includes Channels 0 and 2 , 1 and 3 , 4 and 6 , and 5 and 7 , respectively.
- the stack group 504 a may be referred to as a “master” stack group.
- a stack group 504 b having a stack ID (SID) “ 1 ” includes the core dies core dies 503 e, 503 f, 503 g, and 503 h, and includes Channels 0 and 2 , 1 and 3 , 4 and 6 , and 5 and 7 , respectively.
- the stack group 504 b may be referred to as a “slave” stack group.
- Each stack group 504 a and 504 b includes eight channels (e.g., Channels 0 - 7 ), Memory locations associated with one of the eight channels included in each of the stack groups 504 a and 504 b may be accessed by identifying the stack group with the SID along with an address for the memory locations.
- memory commands and addresses are provided to the interface die 502 from a (graphic) processor or a memory controller (see FIG. 1 ).
- the commands and addresses are then provided to the core dies 503 over the signal lines.
- the core dies 503 provide data over the signal lines 527 to the interface die 502 for read commands and the core dies 503 are provided with data over the signal lines 527 by the interface die 503 for write commands.
- One or more channels of the memory device 500 may be or become inoperable (e.g., defective).
- the entire memory device 500 may be inoperable because of the one or more inoperable channels.
- FIG. 6 is a block diagram of the memory device 500 having two inoperable channels.
- a first inoperable channel is Channel 0 of core die 503 a, which is included in stack group 504 a and a second inoperable is Channel 3 of core die 503 f, which is included in stack group 504 b.
- Both inoperable channels are identified in FIG. 6 by the “X” next to the respective channel. All the remaining channels are operable.
- the memory device 500 includes circuits for replacing an inoperable channel of one stack group with an operable channel of another stack group.
- the inoperable channel of one stack group may be replaced by accessing the operable channel of the other stack group instead of accessing the inoperable channel.
- the resulting memory device 500 may be used as a single stack group memory device having eight operable channels. While the memory capacity of the resulting memory device 500 is less than if all the channels for both stack groups 504 a and 504 b are operable, the resulting memory device may nonetheless be operable as a reduced memory capacity memory device.
- FIG. 7 is a block diagram of the memory device 500 having the two inoperable channels of FIG. 6 and showing an operable channel of one stack group replacing an inoperable channel of another stack group according to an embodiment of the disclosure.
- the inoperable Channel 0 of stack group 504 a is replaced by the operable Channel 0 of stack group 504 b.
- the operable Channel 0 of stack group 504 b is accessed. All the other (i.e., operable) channels of stack group 504 a are accessed normally.
- the memory device 500 may be used as an eight channel memory device with a single stack group.
- FIG. 8 is a schematic diagram of circuits included in the interface die 502 and core dies 503 a - 503 h of the memory device 500 according to an embodiment of the disclosure.
- Each of the core dies 503 a - 503 h may be a memory die including memory for storing data and further including circuits for performing memory operations.
- the core dies 503 a - 503 h are arranged in two stack groups 504 a and 504 b.
- Stack group 504 a includes core dies 503 a - 503 d and stack group 504 b includes core dies 503 e - 503 h.
- each of the other core dies 503 b - 503 d and 503 f - 503 h of the memory device 500 may include the same or similar circuits as shown in FIG. 8 for core dies 503 a and 503 e.
- the interface die 502 includes receiver circuits 810 , 812 , and 814 .
- the receiver circuit 810 receives a clock signal ck_t and provides the ck_t signal to a transmitter circuit 820 .
- the ck_t signal may be used for timing various memory operations, for example memory access operations.
- the receiver circuit 812 receives control signals CTL,
- the CTL signals may include various commands for performing various memory operations.
- the CTL signals may include memory access commands such as read commands, write commands, etc. to be performed by the core dies 503 a - 503 h.
- the CTL signals may also include commands, as represented by AFCTL, for programming storage circuit elements included in storage circuit 830 (which may also be referred to as a defective information store circuit).
- the storage circuit elements of the storage circuit 830 may be programmed in response to the AFCTL command to store, for example, information related to the operability of each channel of the core dies 503 a - 503 h (e.g., defective information for the channel of the core dies 503 a - 503 h ).
- the storage circuit elements of the storage circuit 830 may include non-volatile memory that will continue to store the information even when power is not provided.
- the storage circuit elements of the storage circuit 830 may be antifuse circuits in sonic embodiments of the disclosure.
- the information related to the operability of each channel of the core dies 503 a - 503 h may be determined by product testing during the manufacture of the memory device 500 , and the information programmed in response to the AFCTL command, which may be provided by a memory tester.
- the storage circuit 830 provides signals ChEn (which may also be referred to as defective signals) to a repair circuit 832 .
- the ChEn signals may be based on the information stored in the storage circuit 830 .
- the ChEn signals provided by the storage circuit 830 may indicate which channels of the core dies 503 a - 503 h are operable (or inoperable).
- a first logic level of the ChEn signal may indicate that the respective channel is operable and a second logic level of the ChEn signal may indicate that the respective channel is inoperable.
- the ChEn signals include a respective signal for each channel of the memory device 500 .
- each of the 8 channels for both of the two stack groups has a respective ChEn signal.
- Each of the ChEn signals may indicate whether the respective channel is operable.
- the receiver circuit 814 receives data DQ and provides the data DQ to an input-output (IO) driver circuit 824 .
- the data DQ received by the receiver circuit 814 may be write data to be stored in memory of the core dies 503 a - 503 h associated with a write operation.
- the transmitter circuit 816 provides data DQ from the 1 ( )driver circuit 824 .
- the data provided by the transmitter circuit 816 may be read data from the memory of the core dies 503 a - 503 h associated with a read operation.
- the CTL signals may also include, for example, stack identification (SID) information related to a memory access operation.
- SID stack identification
- the SID information may indicate the stack group and channel associated with the memory access operation.
- the SID information may be provided to the repair circuit 832 .
- the SID information may be compared by the repair circuit 832 to the information stored in the storage circuit 830 in order to determine whether to access a channel of the stack group corresponding to the SID information, or to access a corresponding channel of an alternative stack group, such as when the channel of the stack group corresponding to the SID information is inoperable.
- the repair circuit 832 provides channel select signals SelectChn based on the comparison of the SID information and the information stored in the storage circuit 830 .
- the channel select signals SelectChn include a respective channel select signal for each of the channels.
- An active channel select signal selects the corresponding channel for a memory operation.
- the repair circuit 832 provides eight select signals to the core dies 503 a - 503 h.
- a logic level of the respective select signal will select the corresponding channel of the core dies 503 a - 503 h for a memory operation.
- a first logic level of the respective channel select signal may select a first stack group and a second logic level of the respective select signal may select a second stack group.
- the channel select signals SelectChn are provided by the repair circuit 832 through transmitter circuit 818 to the core dies 503 a - 503 h.
- the core dies 503 a - 503 h are coupled to the interface die 502 through signal lines.
- the signal lines may include conductive vias (or TSVs).
- the conductive vias may have a spiral structure.
- Each of the core dies 503 a - 503 h includes a receiver circuit 840 that receives a clock signal ck_t and further includes a receiver circuit 842 that receives the control signals CTL.
- the receiver circuits 840 and 842 provide the ck_t signal and the CTL signals to a transmitter control circuit 854 .
- the transmitter control circuit 854 provides the ck_t and CTL signals to other core dies of the core dies 503 a - 503 h when activated and does not provide the ck_t and CTL signals to other core dies of the core dies 503 a - 503 h when not activated.
- the receiver circuit 840 also provides the ck_t signal and the CTL signals to a command circuit 856 .
- a control circuit 858 controls various memory circuits to perform operations for a memory command when activated by the command circuit 856 .
- the control circuit 858 may also receive and provide data DQ, for example, receiving write data to be written to memory array 860 , or providing read data from the memory array 860 .
- the data DQ may be received from the IO driver 824 of the interface die 502 , or from the control circuit 858 of other core dies of the core dies 503 a - 503 h.
- the control circuit 858 may be controlled by the command circuit 856 .
- the command circuit 856 may control the control circuit 858 based on the ck_t signal from the receiver circuit 840 and on the CTL signals from the receiver circuit 842 a.
- the core dies 503 a - 503 h further include a receiver circuit 844 that receives a respective channel select signal SelectChn from the interface die 502 and provides the channel select signal to a transmitter circuit 846 and a comparison circuit 852 .
- the transmitter circuit 846 provides the channel select signal to other core dies of the core dies 503 a - 503 h that include a corresponding channel.
- the comparison circuit 852 compares the channel select signal with a respective chip identification (CID) count, or portion of a multibit CID count. In some embodiments of the disclosure, for example, as shown in FIG. 8 , the most significant bit of the CID count (e.g., CID ⁇ 2 > of a three bit CID count) may be compared by the comparison circuit 852 with the channel select signal.
- CID chip identification
- the comparison circuit 852 of a core chip compares the respective SelectChn signal with CID ⁇ 2 > to detect selection of that core chip for a memory access operation.
- the comparison circuit 852 provides an active SIDChnEn signal (e.g., active high logic level) when the core chip is selected and provides an inactive SIDChnEn signal (e.g., inactive low logic level) when the core chip is not selected.
- the SIDChnEn signal is provided to the transmitter control circuit 854 and the command circuit 856 .
- An active SIDChnEn signal activates the command circuit 856 to perform memory operations and does not activate the transmitter control circuit 854
- an inactive SIDChnEn signal does not activate the command circuit 856 but activates the transmitter control circuit 854 .
- memory operations are not performed by the core die, although the clock signal ck_t and CTL signals are provided to the other core die.
- the CID count (or portion of the CID count) is provided by a CID counter circuit 850 .
- the CID counter circuit 850 of a core die provides a CID count that is associated with that core die.
- the CID counter circuit of all the core dies 503 a - 503 h may be reset to an initial value when reset (e.g., power up, reset operation, etc.) and decremented (or incremented in some embodiments) for each of the core dies 503 a - 503 h in the memory device 500 .
- the CID counter circuits 850 of the core dies 503 a - 503 h may be reset to the highest count (or the lowest count).
- the highest count may be 111 for a three-bit count CID ⁇ 2 : 0 > (the lowest count may be 000 for a three-bit count CID ⁇ 2 : 0 >).
- the CID count is decremented (or incremented) by one for each succeeding lower core die in the memory device 500 .
- the most significant bit of the CID ⁇ 2 : 0 > count for a core die, that is, CID ⁇ 2 > has a value that represents the SID of the stack group in which that core die is included.
- a memory command is included in the CTL signal provided to the interface die 502 .
- SID information identifying the stack group and channel to which the memory command is directed.
- the SID associated with a read command identifies the stack group from which data is read
- the SID associated with a write command identifies the stack group to which data is written.
- the repair circuit 832 compares the SID information with the ChEn signals from the storage circuit 830 .
- the ChEn signals may indicate which channels of the core dies 503 a - 503 h are operable.
- the repair circuit 832 provides respective channel select signals SelectChn to the core dies 503 a - 503 h to select the channel of the core die identified by the SID information.
- the repair circuit 832 Based on the ChEn signals, the repair circuit 832 provides the SelectChn signals to select the channel of the core die of the stack group actually corresponding to the SID information, or to select the same channel of a core die of another stack group, such as when the channel of the core die of the stack group that actually corresponds to the SID information is inoperable.
- Channel 0 of core die 503 a is operable, as indicated by information stored in the storage circuit 830 for Channel 0 of core die 503 a, and that a memory command is directed to memory of Channel 0 of stack group 504 a (i.e., core die 503 a including Channel 0 will be accessed), as indicated by the SID information.
- the SelectCh 0 signal is provided by the transmitter circuit 818 to core die 503 a of stack group 504 a and to core die 503 e of stack group 504 b.
- signals may be provided to corresponding core dies of different stack groups through conductive vias having a spiral structure, for example, as previously described with reference to FIG. 4B .
- the receiver circuit 844 a of the core die 503 a provides the SelCh 0 _ 0 signal as a SelCh 0 _ 1 signal to the comparison circuit 852 a and to the transmitter circuit 846 a.
- the transmitter circuit 846 a provides the SelCh 0 _ 1 signal to the core die 503 e.
- the core die 503 e includes Channel 0 for stack group 504 b.
- a receiver circuit 844 e provides the SelCh 0 _ 2 to the comparison circuit 852 e and to the transmitter circuit 846 e.
- the SelCh 0 _ 1 and SelCh 0 _ 2 signals have the same logic level as the SelCh 0 _ 0 signal provided by the transmitter circuit 818 of the interface die 502 .
- the comparison circuit 852 a compares the SelCh 0 _ 1 signal to the CID ⁇ 2 > value from the CID counter circuit 850 a and the comparison circuit 852 e (of core die 503 e ) compares the SelCh 0 _ 2 signal to the CID ⁇ 2 > value from the CID counter circuit 850 e.
- the comparison circuit 852 a identifies a match between SelCh 0 _ 1 signal and its respective CID ⁇ 2 > value (both are 0) and the comparison circuit 852 e does not identify a match between SelCh 0 _ 2 and its respective CID ⁇ 2 > (the SelCh 0 _ 2 signal is 0 and the CID ⁇ 2 > value for core die 503 e is 1).
- the comparison circuit 852 a provides an active SIDCh 0 En_ 1 signal to activate the command circuit 856 a to respond to the memory command included in the CTL signals and to the clock signal ck_t.
- the active SIDCh 0 En_ 1 signal does not activate the transmitter control circuit 854 a to prevent the ck_t signal and the CTL signals from being provided to the core die 503 e.
- the activated command circuit 856 a provides signals to the circuits of the core die 503 a to perform operations for the memory command, for example, the control circuit 858 a may be activated by the command circuit 856 a to control execution of the operation.
- the comparison circuit 852 e provides an inactive SIDCh 0 En_ 2 signal so that the command circuit 856 e remains inactive and the core die 503 e is not accessed.
- Channel 0 of core die 503 a is inoperable but Channel 0 of core die 503 e is operable, as indicated by the information stored in the storage circuit 830 for Channel 0 of core dies 503 a and 503 e, and it is further assumed that a memory command is directed to memory of Channel 0 of stack group 504 , as indicated by the SID information.
- the SID information is provided to the repair circuit 832 and compared to the information from the storage circuit 830 as represented by the ChEn signal.
- the memory device 500 may nonetheless be operable as a memory device with lesser memory capacity.
- the SelectCh 0 signal is provided by the transmitter circuit 818 to core die 503 a of stack group 504 a and to core die 503 e of stack group 504 b.
- the receiver circuit 844 a of the core die 503 a provides the SelCh 0 _ 0 signal as a SelCh 0 _ 1 signal to the comparison circuit 852 a and to the transmitter circuit 846 a.
- the transmitter circuit 846 a provides the SelCh 0 _ 1 signal to the core die 503 e.
- the core die 503 e includes Channel 0 for stack group 504 b.
- a receiver circuit 844 e provides the SelCh 0 _ 2 to the comparison circuit 852 e and to the transmitter circuit 846 e.
- the SelCh 0 _ 1 and SelCh 0 _ 2 signals have the same logic level as the SelCh 0 _ 0 signal provided by the transmitter circuit 818 of the interface die 502 .
- the comparison circuit 852 a compares the SelCh 0 _ 1 signal to the CID ⁇ 2 > value from the CID counter circuit 850 a and the comparison circuit 852 e (of core die 503 e ) compares the SelCh 0 _ 2 signal to the CID ⁇ 2 > value from the CID counter circuit 850 e.
- the comparison circuit 852 a does not identify a match between the SelCh 0 _ 1 signal and its respective CID ⁇ 2 > value (the SelCh 0 _ 1 signal is 1 and the CID ⁇ 2 > value for core die 503 a is 0) and the comparison circuit 852 e identifies a match between SelCh 0 _ 2 and its respective CID ⁇ 2 > value (both are 1).
- the comparison circuit 852 a provides an inactive SIDCh 0 En_ 1 signal so that the command circuit 856 a remains inactive and the core die 503 a is not accessed.
- the transmitter circuit 854 a is active because of the inactive SIDCh 0 En_ 1 signal, and provides to the core die 503 e the CTL signals and the clock signal ck_t_ 0 as the ck_t_ 1 signal.
- the comparison circuit 852 e provides an active SIDCh 0 En_ 2 signal to activate the command circuit 856 e to respond to the memory command included in the CTL signals and to the clock signal ck_t_ 1 provided by the core die 503 a.
- the active SIDCh 0 En_ 2 signal additionally disables the transmitter control circuit 854 e to prevent the ck_t_ 1 signal and the CTL signals from being provided to any other core dies.
- the activated command circuit 856 e provides signals to the circuits of the core die 503 e to perform operations for the memory command, for example, the control circuit 858 e may be activated by the command circuit 856 e to control execution of the operation.
- FIG. 9 is a schematic diagram of a repair circuit 900 according to an embodiment of the disclosure.
- the repair circuit may be included in the repair circuit 832 of FIG. 8 in some embodiments of the disclosure.
- the repair circuit 900 includes a channel selection control circuit 910 and a channel selection circuit 920 .
- the channel selection control circuit 910 includes logic circuits 912 ( 0 )- 912 ( 7 ). Each logic circuit 912 ( 0 )- 912 ( 7 ) corresponds to a channel of a multi-channel memory device, and receives respective enable signals MasterChnEn and SlaveChnEn.
- the MasterChnEn and SlaveChnEn signals are provided from a storage circuit as the ChEn signals, for example, as shown in FIG. 8 .
- the MasterChnEn signal has a logic level that indicates whether the corresponding channel in the master stack group is operational
- the SlaveChnEn signal has a logic level that indicates whether the corresponding channel in the master stack group is operational.
- the MasterCh 0 En signal has a high logic level (e.g., logic level “1”) when Channel 0 of the stack group 504 a is operable and has a low logic level (e.g., logic level “0”) when Channel 0 of the stack group 504 a is inoperable.
- the SlaveCh 0 En signal has a high logic level (e.g., logic level “1”) when Channel 0 of the stack group 504 b is operable and has a low logic level (e.g., logic level “0”) when Channel 0 of the stack group 504 b is inoperable.
- Operability of Channels 1 - 7 for the stack groups 504 a and 504 b are also indicated in the same manner by the respective MasterChnEn and SlaveChnEn signals.
- the outputs of the logic circuits 912 ( 0 )- 912 ( 7 ) are provided to logic circuit 914 .
- the logic circuit 914 performs a logic operation on the outputs of the logic gates 912 ( 0 )- 912 ( 7 ) and provide a control signal SEL to the channel selection circuit 920 .
- the logic circuits 912 ( 0 )- 912 ( 7 ) and 914 of the channel selection control circuit 910 are shown as AND logic circuits in the embodiment of FIG. 9 . However, different logic circuits and/or a different number of logic circuits may be used in other embodiments without departing from the scope of the disclosure.
- the channel selection circuit 920 includes multiplexers 924 ( 0 )- 924 ( 7 ) (which may also be referred to as selector circuits) that receive at a first input respective stack identification (SID) information SIDChn and receive at a second input from a respective inverter 922 ( 0 )- 922 ( 7 ) the complement of respective MasterChnEn signal.
- SIDChn may be included in the SID information provided by the CTL signals.
- the multiplexers 924 ( 0 )- 924 ( 7 ) provide respective channel select signals SelectCh 0 -SelectCh 7 (collectively referred to as SelectChn).
- the multiplexers 924 ( 0 )- 924 ( 7 ) provide the SID information SIDChn or the complement of the MasterChnEn signal as the respective SelectChn signal based on the SEL signal from the channel selection control circuit 910 .
- a high logic level SEL signal controls the multiplexers 924 ( 0 )- 924 ( 7 ) to provide the SID information as the respective SelectChn signals and a low logic level SEL signal controls the multiplexers 924 ( 0 )- 924 ( 7 ) to provide the MasterChnEn signal as the respective SelectChn signals.
- the channel selection control circuit 910 provides a high logic level SEL signal when all of the channels of all of the stack groups are operable (i.e., all of the MasterChnEn and SlaveChnEn signals have a high logic level), but provides a low logic level SEL signal when any of the channels of either stack group are inoperable any of the MasterChnEn and SlaveChnEn signals have a low logic level).
- the multiplexers 924 ( 0 )- 924 ( 7 ) of the channel selection circuit 920 provide SIDChn as respective SelectChn signals.
- the SelectChn signals are provided to the core dies 503 a - 503 h where the value of the SelectChn signal is compared with respective CID value to determine whether to activate the corresponding channel for the stack group 504 a or 504 b.
- the SelectCh 0 signal when SIDCh 0 is a low logic level, and SIDCh 0 -SIDCh 7 are provided by the multiplexers 924 ( 0 )- 924 ( 7 ) as the SelectChn signals, the SelectCh 0 signal will be a low logic level to activate Channel 0 of the stack group 504 a. Conversely, when SIDCh 0 is a high logic level, and SIDCh 0 -SIDCh 7 are provided by the multiplexers 924 ( 0 )- 924 ( 7 ) as the SelectChn signals, the SelectCh 0 signal will be a high logic level to activate Channel 0 of the stack group 504 b.
- the multiplexers 924 ( 0 )- 924 ( 7 ) of the channel selection circuit 920 provide the complement of the MasterChnEn signals as respective SelectChn signals.
- the SelectChn signals for the channels of the memory device that are operable for the stack group 504 a will have a low logic level.
- the SelectChn signals for the channels of the memory device that are inoperable for the stack group 504 a will have a high logic level.
- the low logic level SelectChn signals for the operable channels of stack group 504 a cause the activation of the corresponding core dies 503 a - 503 h of the stack group 504 a when the SID information provided to the memory device indicates selection of the stack group 504 a.
- the high logic level SelectChn signals for the inoperable channels of stack group 504 a cause the activation of the corresponding core dies 503 a - 503 h of the stack group 504 b instead when the SID information provided to the memory device indicates selection of the stack group 504 a.
- an inoperable channel of stack group 504 a may be repaired by replacing it with an operable channel of stack group 504 b, if available.
- such a memory device has inoperable channels and would not be operable as an 8 channel, two stack group memory device, the memory device may nonetheless be operable as an 8 channel, one stack group memory device.
- FIGS. 7, 8, and 9 An example of repairing an inoperable channel in stack group 504 a will now be described with reference to FIGS. 7, 8, and 9 .
- the example is not intended to limit the scope of the disclosure to the particular details of the example, and is provided to facilitate understanding of the operation according to some embodiments of the disclosure.
- Channel 0 of stack group 504 a and Channel 3 of stack group 504 b of the memory device 500 are inoperable, and the remaining channels of stack group 504 a and 504 b are operable.
- the operability of the channels of the memory device 500 may be determined during a testing phase, which may occur during manufacturing, and in some embodiments of the disclosure, also after manufacturing of the memory device.
- the identifications of the inoperable and operable channels are stored in the memory device, for example, with reference to FIG. 8 , as information stored in the storage circuit 830 of the interface die 502 .
- the memory device 500 is inoperable as an 8 channel, two stack group memory device.
- the memory device 500 may be repaired to be operable as an 8 channel, one stack group memory device by replacing inoperable Channel 0 of stack group 504 a with operable Channel 0 of stack group 504 b.
- the inoperable Channel 0 of stack group 504 a and inoperable Channel 3 of stack group 504 b results in, with reference to FIG. 9 , a MasterCh 0 En signal having a low logic level, a SlaveCh 3 En signal having a low logic level, and MasterChnEn and SlaveChnEn signals for the other channels of stack groups 504 a and 504 b have a high logic level.
- the MasterChnEn signals is a low logic level and the MasterCh 1 En-MasterCh 7 signals are a high logic level.
- the channel selection control circuit 910 provides a low logic level SEL signal, which controls the multiplexers 924 ( 0 )- 924 ( 7 ) of the channel selection circuit 920 to provide the complement of the MasterChnEn signals as respective SelectChn signals.
- the SelectCh 0 signal is a high logic level and the SelectCh 1 -SelectCh 7 signals are a low logic level when provided to the core dies of the memory device.
- Channels 1 - 7 of the stack group 504 a may be activated when the corresponding channel is selected by SID information received by the interface die 502 , whereas when SID information received by the interface die 502 selects Channel 0 , Channel 0 of the stack group 504 b is activated instead of Channel 0 of the stack group 504 a.
- replacing an inoperable channel of stack group 504 a with a corresponding operable channel of stack group 504 b may allow the memory device to be operable as an 8 channel, one stack group memory device.
Landscapes
- Engineering & Computer Science (AREA)
- Theoretical Computer Science (AREA)
- Quality & Reliability (AREA)
- Physics & Mathematics (AREA)
- General Engineering & Computer Science (AREA)
- General Physics & Mathematics (AREA)
- Microelectronics & Electronic Packaging (AREA)
- For Increasing The Reliability Of Semiconductor Memories (AREA)
- Dram (AREA)
Abstract
Description
- High data reliability, high speed of memory access, lower power consumption and reduced chip size are features that are demanded from semiconductor memory. In recent years, three-dimensional (3D) memory devices have been introduced. Some 3D memory devices are formed by stacking memory dies (or “dice”) vertically and interconnecting the memory dies using through-silicon (or through-substrate) vias (TSVs). Benefits of the 3D memory devices include shorter interconnects which reduce signal delays and power consumption, a large number of vertical vias between layers which allow wide bandwidth buses between functional blocks in different layers, and a considerably smaller footprint. Thus, the 3D memory devices contribute to higher memory access speed, lower power consumption and chip size reduction. Example 3D memory devices include Hybrid Memory Cube (HMC) and High Bandwidth Memory (HBM). For example, High Bandwidth Memory (HBM) is a type of memory including a high-performance dynamic random access memory (DRAM) interface die and vertically stacked DRAM dies.
- After assembly of the stacked memory dies and interface die, the memory device cannot be easily disassembled in the event one or more of the stacked memory dies become inoperable. A memory die may become inoperable for a variety of reasons, for example, one or more of the circuits of the memory die do not function properly, memory cells of the memory die are defective and cannot correctly and/or adequately store data, etc. Although each die typically includes a limited amount of redundant memory circuits for repairing defective memory cells, the redundant memory circuits may not be capable of repairing all defective memory cells of the memory die, and additionally, inoperable circuits of a memory die may not be repairable. As a result of one or more inoperable memory dies assembled in the stack, if an inoperable die cannot be repaired in place, the entire memory device is scrapped, even if several of or all other memory dies in the stack and the interface are operable.
- It may be desirable to provide repairability to memory devices having a plurality of memory dies and an interface die in order to recover an operable memory device although one or more of the memory dies is inoperable.
- BRIEF DESCRIPTION OF THE DRAWINGS
-
FIG. 1 is a wiring diagram of a memory device and a processor according to an embodiment of the disclosure. -
FIG. 2A is a schematic diagram of a memory device stack including an interface (I/F) die and a plurality of core dies. -
FIG. 2B is a schematic diagram of a portion of the memory device stack ofFIG. 2A . -
FIG. 2C is a schematic diagram of the memory device stack ofFIG. 2A including the I/F die and the plurality of core dies. -
FIG. 3A is a schematic diagram of a memory device stack including an interface (I/F) die and a plurality of core dies according to an embodiment of the disclosure. -
FIG. 3B is a schematic diagram of the memory device stack ofFIG. 3A including the I/F die and the plurality of core dies according to an embodiment of the disclosure. -
FIG. 4A is a wiring diagram of the memory device stack ofFIG. 2A including an I/F die and a plurality of core dies. -
FIG. 4B is a wiring diagram of the memory device stack ofFIG. 3A including an I/F die and a plurality of core dies according to an embodiment of the disclosure. -
FIG. 5 is a block diagram of a memory device according to an embodiment of the disclosure. -
FIG. 6 is a block diagram of the memory device ofFIG. 5 having two inoperable channels. -
FIG. 7 is a block diagram of the memory device ofFIG. 5 having the two inoperable channels ofFIG. 6 and showing an operable channel of one stack group replacing an inoperable channel of another stack group according to an embodiment of the disclosure. -
FIG. 8 is a schematic diagram of circuits included in the memory device ofFIG. 5 according to an embodiment of the disclosure. -
FIG. 9 is a schematic diagram of a repair circuit according to an embodiment of the disclosure. - Certain details are set forth below to provide a sufficient understanding of examples of the disclosure. However, it will be clear to one having skill in the art that examples of the disclosure may be practiced without these particular details. Moreover, the particular examples of the present disclosure described herein should not be construed to limit the scope of the disclosure to these particular examples. In other instances, well-known circuits, control signals, timing protocols, and software operations have not been shown in detail in order to avoid unnecessarily obscuring the disclosure. Additionally, terms such as “couples” and “coupled” mean that two components may be directly or indirectly electrically coupled. Indirectly coupled may imply that two components are coupled through one or more intermediate components.
-
FIG. 1 is a wiring diagram of amemory device 1 and aprocessor 2 according to an embodiment of the disclosure. In some embodiments of the disclosure, thememory device 1 is a High Bandwidth Memory (HBM). In some embodiments of the disclosure, theprocessor 2 may be a graphical processor unit or a memory controller. Thememory device 1 may include terminals coupled by balls 3 (e.g., microbumps) to aninterposer 5. Theprocessor 2 may include terminals coupled by balls 4 (e.g., microbumps) to theinterposer 5 and further to the corresponding terminals of thememory device 1 through theinterposer 5. The interposer may be stacked on a packaging substrate (not shown) byballs 6. For example, theinterposer 5 may be made of silicon. -
FIG. 2A is a schematic diagram of a memory device stack including an interface (I/F) die 22 and a plurality of DRAM core dies 23. For example, the number of the plurality ofcore dies 23 in thememory device stack 21 may be four. Each of thecore dies 23 may be a memory die including a memory array for storing data and further including circuits for performing memory operations.FIG. 2B is a schematic diagram of a portion of thememory device stack 21. The I/F die 22 and the plurality ofcore dies 23 may be coupled by a plurality of conductive vias 27 (e.g., through silicon (substrate) via (TSV)). The I/F die 22 may be on theballs 3. For example, a combination of theconductive vias 27 and theballs 3 may function as interconnects.FIG. 2C is a schematic diagram of thememory device stack 21 including the I/F die 22 and the plurality ofcore dies 23. Thememory device stack 21 may have two 128-bit channels per core die for a total of eight input/output channels and a width of 1024 bits in total. For example, each core die of the plurality of the core dies 23 may include two channels. A channel of a core die represents a separable addressable memory space that may be accessed to read data from the channel and to write data to the channel. In this example, the core dies 23 a, 23 b, 23 c and 23 d include Channels A and C, Channels B and D, Channels E and G, and Channels F and H, respectively. For example, a clock frequency, a command sequence, and data can be independently provided for each channel. -
FIG. 4A is a wiring diagram of thememory device stack 21 including the I/F die 22 and the plurality of core dies 23. The I/F die 22 of thememory device 21 providesinterfaces native IOs 27 a to 27 f may be implemented as conductive vias. For example, the conductive vias may have a spiral structure. Each core die 23 may include a command circuit for each channel. For example, the core dies 23 a to 23 d may includecommand circuits 26 a to 26 d for Channel A, Channel B, Channel E and Channel F, respectively. Thus, clock signals, command signals and data signals for each channel may be transmitted independently and a plurality of data buses and their respective channels can operate individually. -
FIG. 3A is a schematic diagram of amemory device stack 31 that may also be thememory device 1 ofFIG. 1 , including an interface (I/F) die 32 and a plurality of memory core dies 33 such as DRAM dies according to an embodiment of the disclosure. The core dies 33 may be stacked with one another. For example, plurality of core dies 33 in thememory device stack 31 are stacked and may include eight core dies.FIG. 3B is a schematic diagram of thememory device stack 31 including the I/F die 32 and the plurality of core dies 33 according to an embodiment of the disclosure. Thememory device stack 31 may have two 128-bit channels per core die for a total of eight input/output channels and a width of 1024 bits in total. For example, each core die of the plurality of the core dies 33 may include two channels. Thememory device stack 31 may be logically divided into multiple stack groups. In this example, astack group 34 a having a stack identifier (SID) “0” includes the core dies 33 a, 33 b, 33 c and 33 d including Channels A and C, Channels B and D, Channels E and G, and Channels F and H, respectively. Astack group 34 b having a stack ID (SID) “1” includes the core dies 33 e, 33 f, 33 g and 33 h including Channels A and C, Channels B and D, Channels E and G, and Channels F and respectively. Thus, a destination dies among a plurality of core dies in each channel (e.g., core dies 33 a and 33 e of Channel A) addressed in a command may be identified by the SID. The core dies 33 a, 33 b, 33 c, and 33 d may represent a “stack” and the core dies 33 e, 33 f, 33 g, and 33 h may represent another stack that is stacked on the stack of core dies 33 a, 33 b, 33 c, and 33 d. -
FIG. 4B is a wiring diagram of thememory device stack 31 including the I/F die 32 and the plurality of core dies 33 according to an embodiment of the disclosure. The I/F die 32 of thememory device 31 providesinterfaces stack group 34 a and memory arrays of Channels A, B, E and F of thestack group 34 b may be coupled to the same native input/output lines (IOs) 37 a, 37 b, 37 e and 37 f, respectively. For example, memory arrays of Channel A of the core die 33 a in thestack group 34 a and memory arrays of Channel A of the core die 33 e in thestack group 34 b may be coupled to thenative IO 37 a. Each core die 33 may include a command circuit for each channel. For example, the core dies 33 a to 33 d in thestack group 34 a may includecommand circuits 36 a to 36 d for Channel A, Channel B, Channel E and Channel F, respectively. The core dies 33 e to 33 h in thestack group 34 b may includecommand circuits 36 e to 36 h for Channel A, Channel B, Channel E and Channel F, respectively. Each command circuit 36 may detect the SID in a command, check whether the SID in the command matches with an SID of the stack group of the core dies 33 including the command circuit 36, and decode the command if the SID matches and memory access actions responsive to the command may be performed. For example, when theinterface 38 a transmits a command on the input/output line 37 a, thecommand circuit 36 a receives the command and check whether the SID in the command is “0”. Thecommand circuit 36 a processes the command if the SID is “0” and ignores the command if the SID is “1”. Thecommand circuit 36 e also receives the command and check whether the SID in the command is “1”. Thecommand circuit 36 e processes the command if the SID is “1” and ignores the command if the SID is “0”. Thus, clock signals, command signals and data signals for each channel on each die may be transmitted independently. -
FIG. 5 is a block diagram of amemory device 500 according to an embodiment of the disclosure. Thememory device 500 may be a memory device stack in some embodiments of the disclosure. Thememory device 500 includes an interface (I/F) die 502 and a plurality of core dies 503. Each of the core dies 503 may be a memory die including a memory array for storing data and further including circuits for performing memory operations. Thememory device 500 is shown inFIG. 5 as including eight core dies 503 a-503 h. However, greater or fewer core dies may be included in other embodiments of the disclosure. - The
memory device 500 may have multiple channels. For example, each core die 503 of thememory device 500 may include two channels. In the embodiment ofFIG. 5 , the core die 503 a includesChannels Channels Channels Channels Channels Channels Channels Channels - The Channels 0-7 are coupled to the interface die 502 through
signal lines 527. In the embodiment ofFIG. 5 , four sets of signal lines are provided forChannels Channels Channel 0 of core dies 503 a and 503 e, another set of signal lines may provide signals between the interface die 502 andChannel 1 of core dies 503 b and 503 f, and so on, for the other channels. The signal lines 527 may include conductive vias, for example, through-silicon (or through-substrate) vias (TSVs). - The core dies 503 are logically arranged into two stack groups. For example, a
stack group 504 a having a stack identifier (SID) “0” includes the core dies 503 a, 503 b, 503 c, and 503 d, and includesChannels stack group 504 a may be referred to as a “master” stack group. Astack group 504 b having a stack ID (SID) “1” includes the core dies core dies 503 e, 503 f, 503 g, and 503 h, and includesChannels stack group 504 b may be referred to as a “slave” stack group. Eachstack group stack groups - In operation, memory commands and addresses are provided to the interface die 502 from a (graphic) processor or a memory controller (see
FIG. 1 ). The commands and addresses are then provided to the core dies 503 over the signal lines. The core dies 503 provide data over thesignal lines 527 to the interface die 502 for read commands and the core dies 503 are provided with data over thesignal lines 527 by the interface die 503 for write commands. - One or more channels of the
memory device 500 may be or become inoperable (e.g., defective). Theentire memory device 500 may be inoperable because of the one or more inoperable channels. -
FIG. 6 is a block diagram of thememory device 500 having two inoperable channels. A first inoperable channel isChannel 0 of core die 503 a, which is included instack group 504 a and a second inoperable isChannel 3 of core die 503 f, which is included instack group 504 b. Both inoperable channels are identified inFIG. 6 by the “X” next to the respective channel. All the remaining channels are operable. Rather than theentire memory device 500 being inoperable for having an inoperable channel in bothstack groups memory device 500 includes circuits for replacing an inoperable channel of one stack group with an operable channel of another stack group. The inoperable channel of one stack group may be replaced by accessing the operable channel of the other stack group instead of accessing the inoperable channel. The resultingmemory device 500 may be used as a single stack group memory device having eight operable channels. While the memory capacity of the resultingmemory device 500 is less than if all the channels for bothstack groups -
FIG. 7 is a block diagram of thememory device 500 having the two inoperable channels ofFIG. 6 and showing an operable channel of one stack group replacing an inoperable channel of another stack group according to an embodiment of the disclosure. As shown inFIG. 7 . theinoperable Channel 0 ofstack group 504 a is replaced by theoperable Channel 0 ofstack group 504 b. Instead of accessing theinoperable Channel 0 ofstack group 504 a, theoperable Channel 0 ofstack group 504 b is accessed. All the other (i.e., operable) channels ofstack group 504 a are accessed normally. As a result, thememory device 500 may be used as an eight channel memory device with a single stack group. -
FIG. 8 is a schematic diagram of circuits included in the interface die 502 and core dies 503 a-503 h of thememory device 500 according to an embodiment of the disclosure. Each of the core dies 503 a-503 h may be a memory die including memory for storing data and further including circuits for performing memory operations. The core dies 503 a-503 h are arranged in twostack groups Stack group 504 a includes core dies 503 a-503 d andstack group 504 b includes core dies 503 e-503 h.FIG. 8 shows details of a channel for core die 503 a ofstack group 504 a and for core die 503 e ofstack group 504 b. Some or all of the circuits shown and described below with respect to the core dies may be replicated and/or combined for additional channels included in a core die. Moreover, each of the other core dies 503 b-503 d and 503 f-503 h of thememory device 500 may include the same or similar circuits as shown inFIG. 8 for core dies 503 a and 503 e. - The interface die 502 includes
receiver circuits receiver circuit 810 receives a clock signal ck_t and provides the ck_t signal to atransmitter circuit 820. The ck_t signal may be used for timing various memory operations, for example memory access operations. Thereceiver circuit 812 receives control signals CTL, The CTL signals may include various commands for performing various memory operations. For example, the CTL signals may include memory access commands such as read commands, write commands, etc. to be performed by the core dies 503 a-503 h. The CTL signals may also include commands, as represented by AFCTL, for programming storage circuit elements included in storage circuit 830 (which may also be referred to as a defective information store circuit). - The storage circuit elements of the
storage circuit 830 may be programmed in response to the AFCTL command to store, for example, information related to the operability of each channel of the core dies 503 a-503 h (e.g., defective information for the channel of the core dies 503 a-503 h). The storage circuit elements of thestorage circuit 830 may include non-volatile memory that will continue to store the information even when power is not provided. The storage circuit elements of thestorage circuit 830 may be antifuse circuits in sonic embodiments of the disclosure. The information related to the operability of each channel of the core dies 503 a-503 h may be determined by product testing during the manufacture of thememory device 500, and the information programmed in response to the AFCTL command, which may be provided by a memory tester. - The
storage circuit 830 provides signals ChEn (which may also be referred to as defective signals) to arepair circuit 832. The ChEn signals may be based on the information stored in thestorage circuit 830. For example, the ChEn signals provided by thestorage circuit 830 may indicate which channels of the core dies 503 a-503 h are operable (or inoperable). For example, a first logic level of the ChEn signal may indicate that the respective channel is operable and a second logic level of the ChEn signal may indicate that the respective channel is inoperable. The ChEn signals include a respective signal for each channel of thememory device 500. For example, in the embodiment ofFIG. 8 , where thememory device 500 includes 8 channels and two stack groups, each of the 8 channels for both of the two stack groups has a respective ChEn signal. Each of the ChEn signals may indicate whether the respective channel is operable. - The
receiver circuit 814 receives data DQ and provides the data DQ to an input-output (IO)driver circuit 824. The data DQ received by thereceiver circuit 814 may be write data to be stored in memory of the core dies 503 a-503 h associated with a write operation. Thetransmitter circuit 816 provides data DQ from the 1( )driver circuit 824. The data provided by thetransmitter circuit 816 may be read data from the memory of the core dies 503 a-503 h associated with a read operation. - The CTL signals may also include, for example, stack identification (SID) information related to a memory access operation. The SID information may indicate the stack group and channel associated with the memory access operation. The SID information may be provided to the
repair circuit 832. As will be described in more detail below, the SID information may be compared by therepair circuit 832 to the information stored in thestorage circuit 830 in order to determine whether to access a channel of the stack group corresponding to the SID information, or to access a corresponding channel of an alternative stack group, such as when the channel of the stack group corresponding to the SID information is inoperable. - The
repair circuit 832 provides channel select signals SelectChn based on the comparison of the SID information and the information stored in thestorage circuit 830. The channel select signals SelectChn include a respective channel select signal for each of the channels. An active channel select signal selects the corresponding channel for a memory operation. For example, where each of the stack groups include eight channels, therepair circuit 832 provides eight select signals to the core dies 503 a-503 h. A logic level of the respective select signal will select the corresponding channel of the core dies 503 a-503 h for a memory operation. For example, a first logic level of the respective channel select signal may select a first stack group and a second logic level of the respective select signal may select a second stack group. The channel select signals SelectChn are provided by therepair circuit 832 throughtransmitter circuit 818 to the core dies 503 a-503 h. - As previously described, the core dies 503 a-503 h are coupled to the interface die 502 through signal lines. The signal lines may include conductive vias (or TSVs). In some embodiments of the disclosure, the conductive vias may have a spiral structure. Each of the core dies 503 a-503 h includes a receiver circuit 840 that receives a clock signal ck_t and further includes a receiver circuit 842 that receives the control signals CTL. The receiver circuits 840 and 842 provide the ck_t signal and the CTL signals to a transmitter control circuit 854. The transmitter control circuit 854 provides the ck_t and CTL signals to other core dies of the core dies 503 a-503 h when activated and does not provide the ck_t and CTL signals to other core dies of the core dies 503 a-503 h when not activated. The receiver circuit 840 also provides the ck_t signal and the CTL signals to a command circuit 856.
- A control circuit 858 controls various memory circuits to perform operations for a memory command when activated by the command circuit 856. The control circuit 858 may also receive and provide data DQ, for example, receiving write data to be written to memory array 860, or providing read data from the memory array 860. The data DQ may be received from the
IO driver 824 of the interface die 502, or from the control circuit 858 of other core dies of the core dies 503 a-503 h. The control circuit 858 may be controlled by the command circuit 856. The command circuit 856 may control the control circuit 858 based on the ck_t signal from the receiver circuit 840 and on the CTL signals from thereceiver circuit 842 a. - The core dies 503 a-503 h further include a receiver circuit 844 that receives a respective channel select signal SelectChn from the interface die 502 and provides the channel select signal to a transmitter circuit 846 and a comparison circuit 852. The transmitter circuit 846 provides the channel select signal to other core dies of the core dies 503 a-503 h that include a corresponding channel. The comparison circuit 852 compares the channel select signal with a respective chip identification (CID) count, or portion of a multibit CID count. In some embodiments of the disclosure, for example, as shown in
FIG. 8 , the most significant bit of the CID count (e.g., CID<2> of a three bit CID count) may be compared by the comparison circuit 852 with the channel select signal. - The comparison circuit 852 of a core chip compares the respective SelectChn signal with CID<2> to detect selection of that core chip for a memory access operation. The comparison circuit 852 provides an active SIDChnEn signal (e.g., active high logic level) when the core chip is selected and provides an inactive SIDChnEn signal (e.g., inactive low logic level) when the core chip is not selected. The SIDChnEn signal is provided to the transmitter control circuit 854 and the command circuit 856. An active SIDChnEn signal activates the command circuit 856 to perform memory operations and does not activate the transmitter control circuit 854, whereas an inactive SIDChnEn signal does not activate the command circuit 856 but activates the transmitter control circuit 854. As a result, memory operations are not performed by the core die, although the clock signal ck_t and CTL signals are provided to the other core die.
- The CID count (or portion of the CID count) is provided by a CID counter circuit 850. The CID counter circuit 850 of a core die provides a CID count that is associated with that core die. For example, the CID counter circuit of all the core dies 503 a-503 h may be reset to an initial value when reset (e.g., power up, reset operation, etc.) and decremented (or incremented in some embodiments) for each of the core dies 503 a-503 h in the
memory device 500. For example, the CID counter circuits 850 of the core dies 503 a-503 h may be reset to the highest count (or the lowest count). The highest count may be 111 for a three-bit count CID<2:0> (the lowest count may be 000 for a three-bit count CID<2:0>). The CID count is decremented (or incremented) by one for each succeeding lower core die in thememory device 500. As a result, the uppermost core die 503 h of thememory device 500 may have CID<2:0>=111 (or 000), the next lower core die 503 g of thememory device 500 may have CID<2:0>=110 (or 001), the next lower core die 503 f of thememory device 500 may have a CID<2:0>=101 (or 010), and so on until the lowermost core die 503 a of thememory device 500 has a CID<2:0>=000 (or 111). The most significant bit of the CID<2:0> count for a core die, that is, CID<2>, has a value that represents the SID of the stack group in which that core die is included. For example, with reference to thememory device 500, the four uppermost core die 503 e-503 h all have a CID<2> value=1 (or 0), representing an SID=1 (or 0) for thestack group 504 b, and the four lowermost core die 503 a-503 d all have a CID<2> value=0 (or 1), representing an SID=0 (or 1) for thestack group 504 a. - In operation, a memory command is included in the CTL signal provided to the interface die 502. Associated with the memory command is SID information identifying the stack group and channel to which the memory command is directed. For example, the SID associated with a read command identifies the stack group from which data is read, and the SID associated with a write command identifies the stack group to which data is written. The
repair circuit 832 compares the SID information with the ChEn signals from thestorage circuit 830. As previously described, the ChEn signals may indicate which channels of the core dies 503 a-503 h are operable. Therepair circuit 832 provides respective channel select signals SelectChn to the core dies 503 a-503 h to select the channel of the core die identified by the SID information. - Based on the ChEn signals, the
repair circuit 832 provides the SelectChn signals to select the channel of the core die of the stack group actually corresponding to the SID information, or to select the same channel of a core die of another stack group, such as when the channel of the core die of the stack group that actually corresponds to the SID information is inoperable. - For the purpose of providing a non-limiting example, it is assumed that
Channel 0 of core die 503 a is operable, as indicated by information stored in thestorage circuit 830 forChannel 0 of core die 503 a, and that a memory command is directed to memory ofChannel 0 ofstack group 504 a (i.e., core die 503 a includingChannel 0 will be accessed), as indicated by the SID information. - The SID information is provided to the
repair circuit 832 and compared to the information from thestorage circuit 830 as represented by the ChEn signal. With the information from thestorage circuit 830 indicating thatChannel 0 ofstack group 504 a is operable, therepair circuit 832 provides a channel select signal SelectCh0 to selectChannel 0 ofstack group 504 a for access. In the present example, SelectCh0=0 for this condition. - The SelectCh0 signal is provided by the
transmitter circuit 818 to core die 503 a ofstack group 504 a and to core die 503 e ofstack group 504 b. As previously described, in some embodiments of the disclosure, signals may be provided to corresponding core dies of different stack groups through conductive vias having a spiral structure, for example, as previously described with reference toFIG. 4B . Returning toFIG. 8 , thereceiver circuit 844 a of the core die 503 a provides the SelCh0_0 signal as a SelCh0_1 signal to thecomparison circuit 852 a and to thetransmitter circuit 846 a. Thetransmitter circuit 846 a provides the SelCh0_1 signal to the core die 503 e. The core die 503 e includesChannel 0 forstack group 504 b. Areceiver circuit 844 e provides the SelCh0_2 to thecomparison circuit 852 e and to thetransmitter circuit 846 e. The SelCh0_1 and SelCh0_2 signals have the same logic level as the SelCh0_0 signal provided by thetransmitter circuit 818 of the interface die 502. - The
comparison circuit 852 a (of core die 503 a) compares the SelCh0_1 signal to the CID<2> value from theCID counter circuit 850 a and thecomparison circuit 852 e (of core die 503 e) compares the SelCh0_2 signal to the CID<2> value from theCID counter circuit 850 e. As previously described, the CID<2> value from theCID counter circuit 850 a is 0 representing an SID=0 for thestack group 504 a and the CID<2> value from theCID counter circuit 850 e is 1 representing an SID=1 for thestack group 504 b. Thecomparison circuit 852 a identifies a match between SelCh0_1 signal and its respective CID<2> value (both are 0) and thecomparison circuit 852 e does not identify a match between SelCh0_2 and its respective CID<2> (the SelCh0_2 signal is 0 and the CID<2> value for core die 503 e is 1). - The
comparison circuit 852 a provides an active SIDCh0En_1 signal to activate thecommand circuit 856 a to respond to the memory command included in the CTL signals and to the clock signal ck_t. The active SIDCh0En_1 signal does not activate thetransmitter control circuit 854 a to prevent the ck_t signal and the CTL signals from being provided to the core die 503 e. The activatedcommand circuit 856 a provides signals to the circuits of the core die 503 a to perform operations for the memory command, for example, thecontrol circuit 858 a may be activated by thecommand circuit 856 a to control execution of the operation. In contrast, thecomparison circuit 852 e provides an inactive SIDCh0En_2 signal so that thecommand circuit 856 e remains inactive and the core die 503 e is not accessed. - For the purpose of providing another non-limiting example, it is assumed that
Channel 0 of core die 503 a is inoperable butChannel 0 of core die 503 e is operable, as indicated by the information stored in thestorage circuit 830 forChannel 0 of core dies 503 a and 503 e, and it is further assumed that a memory command is directed to memory ofChannel 0 of stack group 504, as indicated by the SID information. - The SID information is provided to the
repair circuit 832 and compared to the information from thestorage circuit 830 as represented by the ChEn signal. With the information from thestorage circuit 830 indicating thatChannel 0 ofstack group 504 a (core die 503 a) is inoperable butChannel 0 ofstack group 504 b (core die 503 e) is operable, therepair circuit 832 provides a channel select signal SelectCh0 to selectChannel 0 ofstack group 504 b for access. In the present example, SelectCh0=1 for this condition. Thus, by repairing theinoperable Channel 0 ofstack group 504 a by replacing it with theoperable Channel 0 ofstack group 504 b, thememory device 500 may nonetheless be operable as a memory device with lesser memory capacity. - The SelectCh0 signal is provided by the
transmitter circuit 818 to core die 503 a ofstack group 504 a and to core die 503 e ofstack group 504 b. Thereceiver circuit 844 a of the core die 503 a provides the SelCh0_0 signal as a SelCh0_1 signal to thecomparison circuit 852 a and to thetransmitter circuit 846 a. Thetransmitter circuit 846 a provides the SelCh0_1 signal to the core die 503 e. The core die 503 e includesChannel 0 forstack group 504 b. Areceiver circuit 844 e provides the SelCh0_2 to thecomparison circuit 852 e and to thetransmitter circuit 846 e. As previously described, the SelCh0_1 and SelCh0_2 signals have the same logic level as the SelCh0_0 signal provided by thetransmitter circuit 818 of the interface die 502. - The
comparison circuit 852 a (of core die 503 a) compares the SelCh0_1 signal to the CID<2> value from theCID counter circuit 850 a and thecomparison circuit 852 e (of core die 503 e) compares the SelCh0_2 signal to the CID<2> value from theCID counter circuit 850 e. As previously described, the CID<2> value from theCID counter circuit 850 a is 0 representing an STD=0 for thestack group 504 a and the CID<2> value from theCID counter circuit 850 e is 1 representing an SID=1 for thestack group 504 b. - The
comparison circuit 852 a does not identify a match between the SelCh0_1 signal and its respective CID<2> value (the SelCh0_1 signal is 1 and the CID<2> value for core die 503 a is 0) and thecomparison circuit 852 e identifies a match between SelCh0_2 and its respective CID<2> value (both are 1). - The
comparison circuit 852 a provides an inactive SIDCh0En_1 signal so that thecommand circuit 856 a remains inactive and the core die 503 a is not accessed. Thetransmitter circuit 854 a is active because of the inactive SIDCh0En_1 signal, and provides to the core die 503 e the CTL signals and the clock signal ck_t_0 as the ck_t_1 signal. In contrast, thecomparison circuit 852 e provides an active SIDCh0En_2 signal to activate thecommand circuit 856 e to respond to the memory command included in the CTL signals and to the clock signal ck_t_1 provided by the core die 503 a. The active SIDCh0En_2 signal additionally disables thetransmitter control circuit 854 e to prevent the ck_t_1 signal and the CTL signals from being provided to any other core dies. The activatedcommand circuit 856 e provides signals to the circuits of the core die 503 e to perform operations for the memory command, for example, thecontrol circuit 858 e may be activated by thecommand circuit 856 e to control execution of the operation. -
FIG. 9 is a schematic diagram of arepair circuit 900 according to an embodiment of the disclosure. The repair circuit may be included in therepair circuit 832 ofFIG. 8 in some embodiments of the disclosure. - The
repair circuit 900 includes a channelselection control circuit 910 and achannel selection circuit 920. The channelselection control circuit 910 includes logic circuits 912(0)-912(7). Each logic circuit 912(0)-912(7) corresponds to a channel of a multi-channel memory device, and receives respective enable signals MasterChnEn and SlaveChnEn. In some embodiments, the MasterChnEn and SlaveChnEn signals are provided from a storage circuit as the ChEn signals, for example, as shown inFIG. 8 . - The MasterChnEn signal has a logic level that indicates whether the corresponding channel in the master stack group is operational, and the SlaveChnEn signal has a logic level that indicates whether the corresponding channel in the master stack group is operational. For example, with reference to
memory device 500, the MasterCh0En signal has a high logic level (e.g., logic level “1”) whenChannel 0 of thestack group 504 a is operable and has a low logic level (e.g., logic level “0”) whenChannel 0 of thestack group 504 a is inoperable. Similarly, the SlaveCh0En signal has a high logic level (e.g., logic level “1”) whenChannel 0 of thestack group 504 b is operable and has a low logic level (e.g., logic level “0”) whenChannel 0 of thestack group 504 b is inoperable. Operability of Channels 1-7 for thestack groups logic circuit 914. Thelogic circuit 914 performs a logic operation on the outputs of the logic gates 912(0)-912(7) and provide a control signal SEL to thechannel selection circuit 920. The logic circuits 912(0)-912(7) and 914 of the channelselection control circuit 910 are shown as AND logic circuits in the embodiment ofFIG. 9 . However, different logic circuits and/or a different number of logic circuits may be used in other embodiments without departing from the scope of the disclosure. - The
channel selection circuit 920 includes multiplexers 924(0)-924(7) (which may also be referred to as selector circuits) that receive at a first input respective stack identification (SID) information SIDChn and receive at a second input from a respective inverter 922(0)-922(7) the complement of respective MasterChnEn signal. With reference toFIG. 8 , SIDChn may be included in the SID information provided by the CTL signals. The multiplexers 924(0)-924(7) provide respective channel select signals SelectCh0-SelectCh7 (collectively referred to as SelectChn). The multiplexers 924(0)-924(7) provide the SID information SIDChn or the complement of the MasterChnEn signal as the respective SelectChn signal based on the SEL signal from the channelselection control circuit 910. In particular, a high logic level SEL signal controls the multiplexers 924(0)-924(7) to provide the SID information as the respective SelectChn signals and a low logic level SEL signal controls the multiplexers 924(0)-924(7) to provide the MasterChnEn signal as the respective SelectChn signals. - In operation, the channel
selection control circuit 910 provides a high logic level SEL signal when all of the channels of all of the stack groups are operable (i.e., all of the MasterChnEn and SlaveChnEn signals have a high logic level), but provides a low logic level SEL signal when any of the channels of either stack group are inoperable any of the MasterChnEn and SlaveChnEn signals have a low logic level). - When all of the channels of a memory device are operable (i.e., all the MasterChnEn and SlaveChnEn signals have a high logic level), the multiplexers 924(0)-924(7) of the
channel selection circuit 920 provide SIDChn as respective SelectChn signals. As previously described with reference toFIG. 8 , the SelectChn signals are provided to the core dies 503 a-503 h where the value of the SelectChn signal is compared with respective CID value to determine whether to activate the corresponding channel for thestack group Channel 0 of thestack group 504 a. Conversely, when SIDCh0 is a high logic level, and SIDCh0-SIDCh7 are provided by the multiplexers 924(0)-924(7) as the SelectChn signals, the SelectCh0 signal will be a high logic level to activateChannel 0 of thestack group 504 b. - When one or more of the channels of the memory device is inoperable (i.e., at least one of the MasterChnEn and SlaveChnEn signals have a low logic level), the multiplexers 924(0)-924(7) of the
channel selection circuit 920 provide the complement of the MasterChnEn signals as respective SelectChn signals. As a result, the SelectChn signals for the channels of the memory device that are operable for thestack group 504 a will have a low logic level. The SelectChn signals for the channels of the memory device that are inoperable for thestack group 504 a will have a high logic level. The low logic level SelectChn signals for the operable channels ofstack group 504 a cause the activation of the corresponding core dies 503 a-503 h of thestack group 504 a when the SID information provided to the memory device indicates selection of thestack group 504 a. However, the high logic level SelectChn signals for the inoperable channels ofstack group 504 a cause the activation of the corresponding core dies 503 a-503 h of thestack group 504 b instead when the SID information provided to the memory device indicates selection of thestack group 504 a. In this manner, an inoperable channel ofstack group 504 a may be repaired by replacing it with an operable channel ofstack group 504 b, if available. Although such a memory device has inoperable channels and would not be operable as an 8 channel, two stack group memory device, the memory device may nonetheless be operable as an 8 channel, one stack group memory device. - An example of repairing an inoperable channel in
stack group 504 a will now be described with reference toFIGS. 7, 8, and 9 . The example is not intended to limit the scope of the disclosure to the particular details of the example, and is provided to facilitate understanding of the operation according to some embodiments of the disclosure. - Referring to
FIG. 7 ,Channel 0 ofstack group 504 a andChannel 3 ofstack group 504 b of thememory device 500 are inoperable, and the remaining channels ofstack group memory device 500 may be determined during a testing phase, which may occur during manufacturing, and in some embodiments of the disclosure, also after manufacturing of the memory device. The identifications of the inoperable and operable channels are stored in the memory device, for example, with reference toFIG. 8 , as information stored in thestorage circuit 830 of the interface die 502. - As a result of
inoperable Channel 0 of thestack group 504 a andinoperable Channel 3 ofstack group 504 b, thememory device 500 is inoperable as an 8 channel, two stack group memory device. However, thememory device 500 may be repaired to be operable as an 8 channel, one stack group memory device by replacinginoperable Channel 0 ofstack group 504 a withoperable Channel 0 ofstack group 504 b. - The
inoperable Channel 0 ofstack group 504 a andinoperable Channel 3 ofstack group 504 b results in, with reference toFIG. 9 , a MasterCh0En signal having a low logic level, a SlaveCh3En signal having a low logic level, and MasterChnEn and SlaveChnEn signals for the other channels ofstack groups selection control circuit 910 provides a low logic level SEL signal, which controls the multiplexers 924(0)-924(7) of thechannel selection circuit 920 to provide the complement of the MasterChnEn signals as respective SelectChn signals. As a result, the SelectCh0 signal is a high logic level and the SelectCh1-SelectCh7 signals are a low logic level when provided to the core dies of the memory device. - By providing the complement of the MasterChnEn signals as respective SelectChn signals, Channels 1-7 of the
stack group 504 a may be activated when the corresponding channel is selected by SID information received by the interface die 502, whereas when SID information received by the interface die 502 selectsChannel 0,Channel 0 of thestack group 504 b is activated instead ofChannel 0 of thestack group 504 a. As previously described, replacing an inoperable channel ofstack group 504 a with a corresponding operable channel ofstack group 504 b may allow the memory device to be operable as an 8 channel, one stack group memory device. - From the foregoing it will be appreciated that, although specific embodiments of the disclosure have been described herein for purposes of illustration, various modifications may be made without deviating from the spirit and scope of the disclosure. Accordingly, the scope disclosure should not be limited any of the specific embodiments described herein.
Claims (12)
Priority Applications (2)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
US15/808,771 US10282264B1 (en) | 2017-11-09 | 2017-11-09 | Apparatus and methods for repairing memory devices including a plurality of memory die and an interface |
US16/352,743 US11113162B2 (en) | 2017-11-09 | 2019-03-13 | Apparatuses and methods for repairing memory devices including a plurality of memory die and an interface |
Applications Claiming Priority (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
US15/808,771 US10282264B1 (en) | 2017-11-09 | 2017-11-09 | Apparatus and methods for repairing memory devices including a plurality of memory die and an interface |
Related Child Applications (1)
Application Number | Title | Priority Date | Filing Date |
---|---|---|---|
US16/352,743 Continuation US11113162B2 (en) | 2017-11-09 | 2019-03-13 | Apparatuses and methods for repairing memory devices including a plurality of memory die and an interface |
Publications (2)
Publication Number | Publication Date |
---|---|
US10282264B1 US10282264B1 (en) | 2019-05-07 |
US20190138412A1 true US20190138412A1 (en) | 2019-05-09 |
Family
ID=66327311
Family Applications (2)
Application Number | Title | Priority Date | Filing Date |
---|---|---|---|
US15/808,771 Active US10282264B1 (en) | 2017-11-09 | 2017-11-09 | Apparatus and methods for repairing memory devices including a plurality of memory die and an interface |
US16/352,743 Active US11113162B2 (en) | 2017-11-09 | 2019-03-13 | Apparatuses and methods for repairing memory devices including a plurality of memory die and an interface |
Family Applications After (1)
Application Number | Title | Priority Date | Filing Date |
---|---|---|---|
US16/352,743 Active US11113162B2 (en) | 2017-11-09 | 2019-03-13 | Apparatuses and methods for repairing memory devices including a plurality of memory die and an interface |
Country Status (1)
Country | Link |
---|---|
US (2) | US10282264B1 (en) |
Cited By (7)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
US11158552B2 (en) | 2018-12-26 | 2021-10-26 | AP Memory Technology Corp. | Semiconductor device and method to manufacture the same |
US11315916B2 (en) * | 2018-12-26 | 2022-04-26 | AP Memory Technology Corp. | Method of assembling microelectronic package and method of operating the same |
US11328997B2 (en) * | 2018-04-05 | 2022-05-10 | Micron Technology, Inc. | Through-core via |
US11380614B2 (en) | 2018-12-26 | 2022-07-05 | AP Memory Technology Corp. | Circuit assembly |
US11417628B2 (en) | 2018-12-26 | 2022-08-16 | Ap Memory Technology Corporation | Method for manufacturing semiconductor structure |
US11672111B2 (en) | 2018-12-26 | 2023-06-06 | Ap Memory Technology Corporation | Semiconductor structure and method for manufacturing a plurality thereof |
US11776584B2 (en) * | 2018-07-31 | 2023-10-03 | Micron Technology, Inc. | Bank and channel structure of stacked semiconductor device |
Families Citing this family (2)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN112655088A (en) * | 2018-09-12 | 2021-04-13 | 华为技术有限公司 | IC die-to-IC die interconnect using error correction code and data path interleaving |
US20210311638A1 (en) * | 2020-04-07 | 2021-10-07 | Micron Technology, Inc. | Apparatuses and methods for die replacement in stacked memory |
Family Cites Families (14)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
US7203779B2 (en) | 2002-01-30 | 2007-04-10 | Stmicroelectronics, Inc. | Fast turn-off slow turn-on arbitrator for reducing tri-state driver power dissipation on a shared bus |
US8041881B2 (en) | 2006-07-31 | 2011-10-18 | Google Inc. | Memory device with emulated characteristics |
US20080028137A1 (en) | 2006-07-31 | 2008-01-31 | Schakel Keith R | Method and Apparatus For Refresh Management of Memory Modules |
JP2011081884A (en) | 2009-10-09 | 2011-04-21 | Elpida Memory Inc | Semiconductor memory device and information processing system equipped with the same |
US9432298B1 (en) * | 2011-12-09 | 2016-08-30 | P4tents1, LLC | System, method, and computer program product for improving memory systems |
JP5932236B2 (en) | 2011-04-13 | 2016-06-08 | ピーエスフォー ルクスコ エスエイアールエルPS4 Luxco S.a.r.l. | Semiconductor device and system |
US20130159587A1 (en) * | 2011-12-15 | 2013-06-20 | Aaron Nygren | Interconnect Redundancy for Multi-Interconnect Device |
JP2013131533A (en) | 2011-12-20 | 2013-07-04 | Elpida Memory Inc | Semiconductor device |
US8869007B2 (en) | 2012-06-14 | 2014-10-21 | International Business Machines Corporation | Three dimensional (3D) memory device sparing |
US9202551B2 (en) | 2012-06-28 | 2015-12-01 | Intel Corporation | Flexible command addressing for memory |
KR102216559B1 (en) | 2013-12-09 | 2021-02-17 | 삼성전자주식회사 | Semiconductor memory device for use in multi chip package |
KR102339780B1 (en) * | 2015-10-29 | 2021-12-15 | 삼성전자주식회사 | Semiconductor device having chip ID generartation circuit |
US10079049B2 (en) | 2016-06-08 | 2018-09-18 | Micron Technology, Inc. | Stack access control for memory device |
KR102650497B1 (en) * | 2017-02-28 | 2024-03-25 | 에스케이하이닉스 주식회사 | Stacked semiconductor device |
-
2017
- 2017-11-09 US US15/808,771 patent/US10282264B1/en active Active
-
2019
- 2019-03-13 US US16/352,743 patent/US11113162B2/en active Active
Cited By (12)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
US11328997B2 (en) * | 2018-04-05 | 2022-05-10 | Micron Technology, Inc. | Through-core via |
US20220262733A1 (en) * | 2018-04-05 | 2022-08-18 | Micron Technology, Inc. | Through-Core Via |
US12040279B2 (en) * | 2018-04-05 | 2024-07-16 | Micron Technology, Inc. | Through-core via |
US11776584B2 (en) * | 2018-07-31 | 2023-10-03 | Micron Technology, Inc. | Bank and channel structure of stacked semiconductor device |
US11158552B2 (en) | 2018-12-26 | 2021-10-26 | AP Memory Technology Corp. | Semiconductor device and method to manufacture the same |
US11315916B2 (en) * | 2018-12-26 | 2022-04-26 | AP Memory Technology Corp. | Method of assembling microelectronic package and method of operating the same |
US11380614B2 (en) | 2018-12-26 | 2022-07-05 | AP Memory Technology Corp. | Circuit assembly |
US11417628B2 (en) | 2018-12-26 | 2022-08-16 | Ap Memory Technology Corporation | Method for manufacturing semiconductor structure |
US11652011B2 (en) | 2018-12-26 | 2023-05-16 | AP Memory Technology Corp. | Method to manufacture semiconductor device |
US11672111B2 (en) | 2018-12-26 | 2023-06-06 | Ap Memory Technology Corporation | Semiconductor structure and method for manufacturing a plurality thereof |
US11887974B2 (en) | 2018-12-26 | 2024-01-30 | AP Memory Technology Corp. | Method of operating microelectronic package |
US12074103B2 (en) | 2018-12-26 | 2024-08-27 | AP Memory Technology Corp. | Circuit assembly |
Also Published As
Publication number | Publication date |
---|---|
US11113162B2 (en) | 2021-09-07 |
US20190213094A1 (en) | 2019-07-11 |
US10282264B1 (en) | 2019-05-07 |
Similar Documents
Publication | Publication Date | Title |
---|---|---|
US11113162B2 (en) | Apparatuses and methods for repairing memory devices including a plurality of memory die and an interface | |
US10347355B2 (en) | Device and method for repairing memory cell and memory system including the device | |
CN106548807B (en) | Repair circuit, semiconductor device using the same, and semiconductor system | |
KR102399014B1 (en) | Shared error detection and correction memory | |
US9401227B2 (en) | Post package repair device | |
US20150286529A1 (en) | Memory device having controller with local memory | |
US8675431B2 (en) | Semiconductor memory device and defective cell relieving method | |
US10032525B2 (en) | Fuse circuit, repair control circuit, and semiconductor apparatus including the same | |
US8817547B2 (en) | Apparatuses and methods for unit identification in a master/slave memory stack | |
CN115428086A (en) | Apparatus and method for die replacement in stacked memory | |
US10489312B2 (en) | Stack access control for memory device | |
CN212675921U (en) | Memory device and memory system | |
EP3933843B1 (en) | Memory testing method and related device | |
US10032523B2 (en) | Memory device including extra capacity and stacked memory device including the same | |
CN114187954A (en) | Memory device, testing method and using method thereof, and memory system |
Legal Events
Date | Code | Title | Description |
---|---|---|---|
AS | Assignment |
Owner name: MICRON TECHNOLOGY, INC., IDAHO Free format text: ASSIGNMENT OF ASSIGNORS INTEREST;ASSIGNOR:OGASAWARA, MASASHI;REEL/FRAME:044086/0658 Effective date: 20171108 |
|
FEPP | Fee payment procedure |
Free format text: ENTITY STATUS SET TO UNDISCOUNTED (ORIGINAL EVENT CODE: BIG.); ENTITY STATUS OF PATENT OWNER: LARGE ENTITY |
|
AS | Assignment |
Owner name: MORGAN STANLEY SENIOR FUNDING, INC., AS COLLATERAL Free format text: SUPPLEMENT NO. 7 TO PATENT SECURITY AGREEMENT;ASSIGNOR:MICRON TECHNOLOGY, INC.;REEL/FRAME:045267/0833 Effective date: 20180123 Owner name: MORGAN STANLEY SENIOR FUNDING, INC., AS COLLATERAL AGENT, MARYLAND Free format text: SUPPLEMENT NO. 7 TO PATENT SECURITY AGREEMENT;ASSIGNOR:MICRON TECHNOLOGY, INC.;REEL/FRAME:045267/0833 Effective date: 20180123 |
|
AS | Assignment |
Owner name: JPMORGAN CHASE BANK, N.A., AS COLLATERAL AGENT, IL Free format text: SECURITY INTEREST;ASSIGNORS:MICRON TECHNOLOGY, INC.;MICRON SEMICONDUCTOR PRODUCTS, INC.;REEL/FRAME:047540/0001 Effective date: 20180703 Owner name: JPMORGAN CHASE BANK, N.A., AS COLLATERAL AGENT, ILLINOIS Free format text: SECURITY INTEREST;ASSIGNORS:MICRON TECHNOLOGY, INC.;MICRON SEMICONDUCTOR PRODUCTS, INC.;REEL/FRAME:047540/0001 Effective date: 20180703 |
|
STCF | Information on status: patent grant |
Free format text: PATENTED CASE |
|
AS | Assignment |
Owner name: MICRON TECHNOLOGY, INC., CALIFORNIA Free format text: RELEASE BY SECURED PARTY;ASSIGNOR:MORGAN STANLEY SENIOR FUNDING, INC., AS COLLATERAL AGENT;REEL/FRAME:050716/0678 Effective date: 20190731 |
|
AS | Assignment |
Owner name: MICRON SEMICONDUCTOR PRODUCTS, INC., IDAHO Free format text: RELEASE BY SECURED PARTY;ASSIGNOR:JPMORGAN CHASE BANK, N.A., AS COLLATERAL AGENT;REEL/FRAME:051028/0001 Effective date: 20190731 Owner name: MICRON TECHNOLOGY, INC., IDAHO Free format text: RELEASE BY SECURED PARTY;ASSIGNOR:JPMORGAN CHASE BANK, N.A., AS COLLATERAL AGENT;REEL/FRAME:051028/0001 Effective date: 20190731 |
|
CC | Certificate of correction | ||
MAFP | Maintenance fee payment |
Free format text: PAYMENT OF MAINTENANCE FEE, 4TH YEAR, LARGE ENTITY (ORIGINAL EVENT CODE: M1551); ENTITY STATUS OF PATENT OWNER: LARGE ENTITY Year of fee payment: 4 |