Test circuits and semiconductor devices configured to perform tests
By designing test circuits to optimize the testing process of stacked storage systems, the problem of excessively long testing times in existing technologies has been solved, enabling efficient fault identification and data transmission.
Patent Information
- Application Number
- CN202511312002.9
- Authority / Receiving Office
- CN · China
- Patent Type
- Applications(China)
- Current Assignee / Owner
- Priority Date
- 2025-05-07
- Filing Date
- 2025-09-15
- Publication Date
- 2026-11-10
AI Technical Summary
Existing stacked storage systems struggle to efficiently identify and reduce faults during testing, leading to excessively long testing times and impacting data transmission efficiency.
A test circuit was designed, including a data transmission generation circuit, a selection circuit, an error flag generation circuit, an error flag counter, and a fault signal generation circuit. By generating error flags and counting signals, the transmission of data is driven, reducing the need to determine faults for each read operation and optimizing the test process.
By reducing the need to determine faults for each read operation, testing time is shortened, data transfer efficiency is improved, and it is suitable for high-bandwidth storage systems.
Smart Images

Figure CN123072476A_ABST
Abstract
Description
[0001] Cross-reference to related applications
[0002] This application claims priority to Korean Application No. 10-2025-0059295, filed on May 7, 2025, with the Korean Intellectual Property Office, the entire contents of which are incorporated herein by reference. Technical Field
[0003] This disclosure generally relates to test circuits and semiconductor devices, and more specifically, to test circuits and semiconductor devices configured to perform tests. Background Technology
[0004] Stacked memory systems, such as high-bandwidth memory (HBM) devices, are used in a wide range of applications due to their high bandwidth and energy efficiency. Unlike traditional memory systems that use parallel data buses, stacked memory systems consist of stacked memory devices comprising base dies and core dies interconnected via through-silicon vias (TSVs). These stacked memory devices include physical interfaces, such as a physical layer for communicating with the processor. This physical layer is designed for high-speed data transfer and efficient communication. Summary of the Invention
[0005] This disclosure provides a test circuit that may include: a transmission data generation circuit configured to generate an error flag based on selected compressed data generated from a channel compression test or a pseudo-channel compression test, compare an error count signal generated by counting the error flags with a target code to generate a fault signal, and drive transmission data based on the fault signal; and a register transmission level circuit configured to generate test output data from the transmission data and transmit the test output data to a direct access area.
[0006] This disclosure provides a test circuit that may include: a selection circuit configured to select and output one of compressed data and pseudo-compressed data as selected compressed data based on a compression test selection signal; an error flag generation circuit configured to generate an error flag based on the selected compressed data; an error flag counter configured to generate an error count signal based on the error flag and a read strobe signal; a fault signal generation circuit configured to compare the error count signal with a target code to generate a fault signal; a first drive circuit configured to generate first transmission data from the selected compressed data and the fault signal based on a first test mode signal; and a second drive circuit configured to generate second transmission data from the selected compressed data and the fault signal based on a second test mode signal.
[0007] This disclosure provides a semiconductor device that may include a plurality of core dies stacked on a base die. In this disclosure, each of the plurality of core dies may include a plurality of core regions assigned to a plurality of channels, and each of the plurality of channels may include a plurality of pseudo-channels. In this disclosure, the base die may include test circuitry configured to generate error flags based on selected compressed data generated in a channel compression test or a pseudo-channel compression test, compare an error count signal generated by counting the error flags with a target code to generate a fault signal, and drive data transmission based on the fault signal. Attached Figure Description
[0008] Figure 1 A test circuit according to an embodiment of the present disclosure is shown.
[0009] Figure 2 The core die is shown according to an embodiment of this disclosure.
[0010] Figure 3 It is a table during the operation of selecting compressed data in the selection circuit according to an embodiment of the present disclosure.
[0011] Figure 4 This is a table showing the operation period during which a fault signal is generated in the fault signal generation circuit according to an embodiment of the present disclosure by comparing an error count signal with a target code.
[0012] Figure 5 This is a table during the operation of driving the first transmission data in the first driving circuit according to an embodiment of the present disclosure.
[0013] Figure 6 This is a table during the operation of driving the second transmission data in the second driving circuit according to an embodiment of the present disclosure.
[0014] Figure 7 The basic die is shown according to an embodiment of this disclosure.
[0015] Figure 8 A storage system according to an embodiment of the present disclosure is shown. Detailed Implementation
[0016] In the following description of the embodiments, when a parameter is referred to as “preset” or “predetermined,” it may be intended to indicate that the value of the parameter is predetermined when the parameter is used in a process or algorithm. The value of the parameter may be set at the start of the process or algorithm, or it may be set during the execution of the process or algorithm. For example, a preset number of times means that the value of the parameter is determined before the parameter is used in the process or algorithm. In some embodiments, the value of the parameter is determined before the process or algorithm begins. In other embodiments, the value of the parameter is determined during the process or algorithm but before the parameter is used in the process or algorithm.
[0017] Terms such as “first” and “second” are used to distinguish multiple components without implying the size, order, priority, number, or importance of the components. For example, in one example, the first component might be named the second component, while in another example, the second component might be named the first component.
[0018] When one component is determined to be "connected" to another component, these components can be connected directly or through an intermediary component between them. When two components are determined to be "directly connected," one component connects directly to the other without any intermediary component between them.
[0019] Logic "high" and logic "low" levels can be used to describe the logic levels of electrical signals. Signals with a logic high level are distinct from signals with a logic low level. For example, when a signal at a first voltage level corresponds to a logic high level signal, a signal at a second voltage level corresponds to a logic low level signal. In one embodiment, a logic high level can be a voltage level higher than a logic low level. According to embodiments, the logic levels of signals can be different or opposite. For example, a signal at a logic high level in one embodiment can be at a logic low level in another embodiment, and vice versa.
[0020] Embodiments of this disclosure are described in detail with reference to the accompanying drawings. Specific structural or functional descriptions of the embodiments are provided as examples to illustrate the concepts disclosed in this application. Examples or embodiments based on these concepts may be implemented in various forms, and the scope of this disclosure is not limited to the examples or embodiments described in this specification.
[0021] Figure 1 A test circuit 1 according to an embodiment of the present disclosure is shown.
[0022] like Figure 1 As shown, the test circuit 1 includes a data transmission generation circuit 11 and a register transfer level (RTL) circuit 13. The data transmission generation circuit 11 includes a selection circuit 111, a fault determination circuit 112, a first drive circuit (DRV1) 119-1, and a second drive circuit (DRV2) 119-2.
[0023] Selection circuit 111 selects and outputs either compressed data IOCOMP or pseudo-compressed data PCCOMP as the selected compressed data COMP-SEL based on the compression test selection signal TMC-SEL. When the compression test selection signal TMC-SEL is at a first logic level, selection circuit 111 selects and outputs compressed data IOCOMP as the selected compressed data COMP-SEL, and when the compression test selection signal TMC-SEL is at a second logic level, selection circuit 111 selects and outputs pseudo-compressed data PCCOMP as the selected compressed data COMP-SEL. When performing a channel compression test, the compression test selection signal TMC-SEL is set to the first logic level, and when performing a pseudo-channel compression test, the compression test selection signal TMC-SEL is set to the second logic level. The first logic level can be set to a logic high level, and the second logic level can be set to a logic low level, but this is only an example, and this disclosure is not limited thereto.
[0024] Channel compression testing is performed as follows: On a storage device (e.g., ...), Figure 8 After the storage cell array (not shown) included in the storage device 37) performs a write operation storing data of the same logic level through multiple channels, it generates compressed data IOCOMP by comparing the data stored in the storage cell array with the data output through multiple channels for each pseudo-channel and burst length according to a read operation. For example, the first bit of the compressed data IOCOMP <1> It can be generated by comparing the data with the first burst length output through the first to fourth channels that are output through the first pseudo-channel. When all the data with the first burst length output through the first to fourth channels that are output through the first pseudo-channel are the same, the first bit of the compressed data is IOCOMP. <1> It can be set to a first logic level, and when at least one of the data output through the first to fourth channels with a first burst length, or the data output through the first pseudo-channel, is different from the other data, the first bit of the compressed data, IOCOMP, is... <1> It can be set to the second logic level. As another example, the second bit of the compressed data, IOCOMP. <2> It can be generated by comparing the data output through the first to fourth channels that has a second burst length and is output through the first pseudo-channel. Each channel may include multiple pseudo-channels, and when the burst length is set to four (4), data with the first to fourth burst lengths can be sequentially output to each data pad, but this is only an example and the disclosure is not limited thereto.
[0025] A pseudo-channel compression test can be performed as follows: After a write operation is performed on a memory cell array included in the storage device, storing data of the same logic level through multiple channels, a pseudo-compressed data PCCOMP is generated by comparing the data stored in the memory cell array with data of multiple burst lengths output through a specific pseudo-channel based on a read operation. For example, the first bit of the pseudo-compressed data PCCOMP... <1> This can be generated by comparing data with first to fourth burst lengths output through the first pseudo-channel. In this case, when all data with first to fourth burst lengths output through the first pseudo-channel are identical, the first bit of the pseudo-compressed data is PCCOMP. <1> It can be set to the first logic level, and when at least one of the data with the first to fourth burst lengths output through the first pseudo-channel is different from the other data, the first bit of the pseudo-compressed data, PCCOMP, is set to the first logic level. <1> It can be set to the second logic level. As another example, data with first to fourth burst lengths output through the second pseudo-channel can be compared, and the second bit of the pseudo-compressed data, PCCOMP, can be output. <2> .
[0026] The fault determination circuit 112 generates an activated fault signal F-EN based on the selected compressed data COMP-SEL when the error flag E-FLAG is activated a preset number of times. The fault determination circuit 112 includes an error flag generation circuit (E-FLAG GEN) 113, an error flag counter (E-FLAG CNT) 115, and a fault signal generation circuit (F-EN GEN) 117.
[0027] Error flag generation circuit 113 is electrically connected to selection circuit 111 and receives selected compressed data COMP-SEL from selection circuit 111. Error flag generation circuit 113 generates error flag E-FLAG based on the selected compressed data COMP-SEL. When a channel compression test is performed with output compressed data IOCOMP as the selected compressed data COMP-SEL, and the logic level of at least one bit in the selected compressed data COMP-SEL is different from the logic level of the other bits, error flag generation circuit 113 generates error flag E-FLAG activated at a first logic level. When a channel compression test is performed with output compressed data IOCOMP as the selected compressed data COMP-SEL, and the logic levels of all bits in the selected compressed data COMP-SEL are the same, error flag generation circuit 113 generates error flag E-FLAG deactivated at a second logic level. When a pseudo-channel compression test is performed with the output pseudo-compressed data PCCOMP as the selected compressed data COMP-SEL, and the logic level of at least one bit in the selected compressed data COMP-SEL is different from the logic level of the other bits, the error flag generation circuit 113 generates an error flag E-FLAG that is activated at the first logic level. When a pseudo-channel compression test is performed with the output pseudo-compressed data PCCOMP as the selected compressed data COMP-SEL, and all bits in the selected compressed data COMP-SEL are the same, the error flag generation circuit 113 generates an error flag E-FLAG that is deactivated at the second logic level.
[0028] Error flag counter 115 is electrically connected to error flag generation circuit 113 and receives error flag E-FLAG from error flag generation circuit 113. Error flag counter 115 generates error count signal E-CNT based on error flag E-FLAG and read strobe signal RDQS. Error flag counter 115 counts the activated error flag E-FLAG synchronously with the read strobe signal RDQS, which is activated during read operations for channel compression test or pseudo channel compression test, to generate error count signal E-CNT.
[0029] Error flag counter 115 generates an error count signal E-CNT consisting of a binary bit group. This error count signal E-CNT is counted and incremented by 1 bit each time an error flag E-FLAG is generated. The error flag E-FLAG is activated according to a channel compression test when at least one of the data stored in the memory cell array allocated to all channels differs from the other data. For example, when the error flag E-FLAG is activated for the first time synchronously with the read strobe signal RDQS of the channel compression test, error flag counter 115 generates an error count signal E-CNT<2:0> consisting of a binary bit group with "001", and when the error flag E-FLAG is activated for the second time synchronously with the read strobe signal RDQS of the channel compression test, error flag counter 115 generates an error count signal E-CNT<2:0> consisting of a binary bit group with "010".
[0030] Error flag counter 115 generates an error count signal E-CNT with a binary bit group. This error count signal E-CNT is counted and incremented by 1 bit each time an error flag E-FLAG is generated. The error flag E-FLAG is activated according to a pseudo-channel compression test when one of the data stored in the memory cell array allocated to all pseudo-channels differs from the other data. For example, when the error flag E-FLAG is activated for the first time synchronously with the read strobe signal RDQS of the pseudo-channel compression test, error flag counter 115 generates an error count signal E-CNT<2:0> with a binary bit group of "001", and when the error flag E-FLAG is activated for the second time synchronously with the read strobe signal RDQS of the pseudo-channel compression test, error flag counter 115 generates an error count signal E-CNT<2:0> with a binary bit group of "010".
[0031] A fault signal generation circuit 117 is electrically connected to an error flag counter 115 and receives an error count signal E-CNT from the error flag counter 115. The fault signal generation circuit 117 compares the error count signal E-CNT with the target code TCD to generate a fault signal F-EN. When the binary bit set of the error count signal E-CNT is the same as the binary bit set of the target code TCD, the fault signal generation circuit 117 generates a fault signal F-EN that is activated at a first logic level. When the binary bit set of the error count signal E-CNT is different from the binary bit set of the target code TCD, the fault signal generation circuit 117 generates a fault signal F-EN that is deactivated at a second logic level. The target code TCD can be set to have a binary bit set corresponding to a preset number of times the error flag E-FLAG is activated, which is used to activate the fault signal F-EN during channel compression testing.
[0032] The data transmission generation circuit 11 can drive the transmission data (e.g., first transmission data TD1 and second transmission data TD2) based on the fault signal F-EN. For example, the first drive circuit 119-1 is electrically connected to the selection circuit 111 and the fault signal generation circuit 117, receiving selected compressed data COMP-SEL from the selection circuit 111 and the fault signal generation circuit 117. The first drive circuit 119-1 generates the first transmission data TD1 from the selected compressed data COMP-SEL and the fault signal F-EN based on the first test mode signal TM-CP. When a channel compression test or a pseudo-channel compression test is performed, the first test mode signal TM-CP can be set to a first logic level, and when neither a channel compression test nor a pseudo-channel compression test is performed, the first test mode signal TM-CP can be set to a second logic level. When a channel compression test or a pseudo-channel compression test is performed and the first test mode signal TM-CP set to the first logic level is received, the first drive circuit 119-1 generates the fault signal F-EN as the first transmission data TD1. When neither channel compression test nor pseudo channel compression test is performed and the first test mode signal TM-CP set to the second logic level is received, the first driver circuit 119-1 generates selected compressed data COMP-SEL as the first transmission data TD1.
[0033] The data transmission generation circuit 11 can drive the transmission data (e.g., first transmission data TD1 and second transmission data TD2) based on the fault signal F-EN. For example, the second drive circuit 119-2 is electrically connected to the selection circuit 111 and the fault signal generation circuit 117, receiving selected compressed data COMP-SEL from the selection circuit 111 and the fault signal generation circuit 117. The second drive circuit 119-2 generates the second transmission data TD2 based on the selected compressed data COMP-SEL and the fault signal F-EN based on the second test mode signal TM-S. When a fixed test is performed, the second test mode signal TM-S can be set to a first logic level, and when the fixed test is not performed, the second test mode signal TM-S can be set to a second logic level. When a fixed test is performed, the second test mode signal TM-S set to the first logic level is received, and the fault signal F-EN is activated, the second drive circuit 119-2 generates the second transmission data TD2 fixed at the first logic level. When the fixed test is not performed and a second test mode signal TM-S, set to the second logic level, is received, the second driver circuit 119-2 generates selected compressed data COMP-SEL as the second transmission data TD2. The fixed test can be performed to set the second transmission data TD2, fixed to the first logic level, to be output from the second driver circuit 119-2 when the fault signal F-EN is activated.
[0034] Register transfer level circuit 13 is electrically connected to transfer data generation circuit 11 and receives first transfer data TD1 and second transfer data TD2 from transfer data generation circuit 11. Register transfer level circuit 13 outputs first transfer data TD1 and second transfer data TD2 as test output data T-DA. Test output data T-DA is output through pads in the direct access area, for example, Figure 7 Direct access area (DA).
[0035] The test circuit 1 configured as described above does not determine whether a fault has occurred for each read operation during channel compression testing or pseudo-channel compression testing. Instead, it outputs an activated fault signal F-EN when a preset number of errors occur in multiple read operations, thereby reducing test time in one embodiment. Furthermore, when an error occurs for the first time in multiple read operations during a fixed test, the test circuit 1 outputs a fault signal F-EN that is fixed in an active state, thereby reducing test time in one embodiment.
[0036] Figure 2 The core die 15 is shown according to an embodiment of this disclosure.
[0037] like Figure 2As shown, core die 15 includes multiple core regions allocated to channels CH0 through CH31. Each core region includes a memory cell array allocated to each of channels CH0 through CH31 and receives and stores data during channel compression testing or pseudo-channel compression testing. Each of channels CH0 through CH31 includes a first pseudo-channel PC0 and a second pseudo-channel PC1. During channel compression testing, each core region compares data output through a channel selected from channels CH0 through CH31 for each pseudo-channel and burst length to generate compressed data IOCOMP. During pseudo-channel compression testing, each core region compares data of multiple burst lengths output through either the first pseudo-channel PC0 or the second pseudo-channel PC1 to generate pseudo-compressed data PCCOMP. Although the allocation of core regions included in core die 15 to channels CH0 through CH31, including the first pseudo-channel PC0 and the second pseudo-channel PC1, has been illustrated by way of example, this is merely an example and the disclosure is not limited thereto.
[0038] Figure 3 In accordance with embodiments of this disclosure, such as Figure 1 The selection circuit 111 shown selects the table during the operation of the selected compressed data COMP-SEL.
[0039] Reference Figure 3 and Figure 1 When the compression test selection signal TMC-SEL is at a logic low level "L", the selection circuit 111 selects and outputs the compressed data IOCOMP as the selected compressed data COMP-SEL. When the compression test selection signal TMC-SEL is at a logic high level "H", the selection circuit 111 selects and outputs the pseudo compressed data PCCOMP as the selected compressed data COMP-SEL.
[0040] Figure 4 In accordance with embodiments of this disclosure, such as Figure 1 The table shows the operation period during which the fault signal F-EN is generated by comparing the error count signal E-CNT with the target code TCD in the fault signal generation circuit 117 shown.
[0041] Reference Figure 4 The first line and Figure 1When the error flag E-FLAG is activated for the first time due to a channel compression test or a pseudo channel compression test, and the error count signal E-CNT<2:0> with a binary bit group of "001" is generated, since the binary bit group of the target code TCD<2:0> with a binary bit group of "111" is different from the binary bit group of the error count signal E-CNT<2:0>, the fault signal generation circuit 117 generates a fault signal F-EN that is deactivated at the logic low level "L".
[0042] Reference Figure 4 The second line and Figure 1 When the error flag E-FLAG is activated for the second time due to a channel compression test or a pseudo channel compression test, and the error count signal E-CNT<2:0> with a binary bit group of "010" is generated, since the binary bit group of the target code TCD<2:0> with a binary bit group of "111" is different from the binary bit group of the error count signal E-CNT<2:0>, the fault signal generation circuit 117 generates a fault signal F-EN that is deactivated at the logic low level "L".
[0043] Reference Figure 4 The third line and Figure 1 When the error flag E-FLAG is activated for the third time due to a channel compression test or a pseudo channel compression test, and the error count signal E-CNT<2:0> with a binary bit group of "011" is generated, since the binary bit group of the target code TCD<2:0> with a binary bit group of "111" is different from the binary bit group of the error count signal E-CNT<2:0>, the fault signal generation circuit 117 generates a fault signal F-EN that is deactivated at the logic low level "L".
[0044] Reference Figure 4 The fourth line and Figure 1 When the error flag E-FLAG is activated for the fourth time due to a channel compression test or a pseudo channel compression test, and the error count signal E-CNT<2:0> with a binary bit group of "100" is generated, since the binary bit group of the target code TCD<2:0> with a binary bit group of "111" is different from the binary bit group of the error count signal E-CNT<2:0>, the fault signal generation circuit 117 generates a fault signal F-EN that is deactivated at the logic low level "L".
[0045] Reference Figure 4 The fifth line and Figure 1When the error flag E-FLAG is activated for the fifth time due to a channel compression test or a pseudo channel compression test, and the error count signal E-CNT<2:0> with a binary bit group of "101" is generated, since the binary bit group of the target code TCD<2:0> with a binary bit group of "111" is different from the binary bit group of the error count signal E-CNT<2:0>, the fault signal generation circuit 117 generates a fault signal F-EN that is deactivated at the logic low level "L".
[0046] Reference Figure 4 The sixth line and Figure 1 When the error flag E-FLAG is activated for the sixth time due to a channel compression test or a pseudo channel compression test, and the error count signal E-CNT<2:0> with a binary bit group of "110" is generated, since the binary bit group of the target code TCD<2:0> with a binary bit group of "111" is different from the binary bit group of the error count signal E-CNT<2:0>, the fault signal generation circuit 117 generates a fault signal F-EN that is deactivated at the logic low level "L".
[0047] Reference Figure 4 The seventh line and Figure 1 When the error flag E-FLAG is activated for the seventh time due to a channel compression test or a pseudo channel compression test, and the error count signal E-CNT<2:0> with a binary bit group of "111" is generated, since the binary bit group of the target code TCD<2:0> with a binary bit group of "111" is the same as the binary bit group of the error count signal E-CNT<2:0>, the fault signal generation circuit 117 generates the fault signal F-EN activated at the logic high level "H".
[0048] Figure 5 In accordance with embodiments of this disclosure, such as Figure 1 The table shows the operation of driving the first transmission data TD1 in the first driving circuit 119-1.
[0049] like Figure 5 As shown, when the channel compression test and pseudo-channel compression test are not executed and the first test mode signal TM-CP, which is set to logic low level "L", is received, the first driver circuit 119-1 generates selected compressed data COMP-SEL as the first transmission data TD1. When the channel compression test or pseudo-channel compression test is executed and the first test mode signal TM-CP, which is set to logic high level "H", is received, the first driver circuit 119-1 generates a fault signal F-EN as the first transmission data TD1.
[0050] Figure 6In accordance with embodiments of this disclosure, such as Figure 1 The table shows the operation of driving the second transmission data TD2 in the second driving circuit 119-2.
[0051] like Figure 6 As shown, when the fixed test is not performed and the second test mode signal TM-S, which is set to logic low level "L", is received, the second drive circuit 119-2 generates selected compressed data COMP-SEL as the second transmission data TD2. When the fixed test is performed, the second test mode signal TM-S, which is set to logic high level "H", is received and the fault signal F-EN is activated, the second drive circuit 119-2 generates the second transmission data TD2, which is fixed at the first logic level.
[0052] Figure 7 The base die 2 is shown according to an embodiment of the present disclosure.
[0053] like Figure 7 As shown, the basic die 2 includes a physical layer (PHY), a serial-to-parallel conversion circuit (PHYSER), alignment circuits 21-1, 21-2, 21-3, 21-4, 21-5 and 21-6, a through silicon via region (TSV), a direct access region (DA) and a logic circuit region (DARTL).
[0054] Physical layer (PHY) control communicates with the processor (e.g., via multiple channels) through the core die. Figure 8 The physical layer (PHY) facilitates the transfer of data, commands, and addresses between processors 39. It includes data control circuitry and command / address control circuitry. The data control circuitry controls data transmission through multiple channels, while the command / address control circuitry controls the transmission of commands and addresses through multiple channels. The physical layer (PHY) is used for high-speed data transmission, maintaining signal integrity, and low-power design in high-bandwidth memory (HBM) systems.
[0055] A serial-to-parallel conversion circuit (PHYSER) converts serial signals transmitted serially through the physical layer (PHY) into parallel signals. Alignment circuits 21-1, 21-2, 21-3, 21-4, 21-5, and 21-6 align the converted parallel signals before transmitting them to a through-silicon via (TSV) region. Alignment circuits 21-1, 21-2, 21-3, and 21-4 can be implemented as data alignment circuits, while alignment circuits 21-5 and 21-6 can be implemented as command, address, and data alignment circuits; however, these are merely examples, and this disclosure is not limited thereto. The position and number of alignment circuits 21-1, 21-2, 21-3, 21-4, 21-5, and 21-6 can be modified in various ways depending on the embodiment. The serial-to-parallel conversion circuit (PHYSER) includes... Figure 1The transmission data generation circuit 11 shown outputs an activated fault signal F-EN when a preset number of errors occur during multiple read operations, or outputs a fault signal F-EN that is fixed in an activated state when an error occurs for the first time during multiple read operations in a fixed test, without having to determine whether a fault has occurred for each read operation when a channel compression test or pseudo-channel compression test is performed.
[0056] The through-silicon via (TSV) region includes multiple TSVs that connect the base die 2 and multiple core dies, and transmit parallel signals from the serial-to-parallel converter (PHYSER) to the core dies, or output parallel signals from the core dies to the serial-to-parallel converter (PHYSER).
[0057] Direct access areas (DAs) include those used for accessing data without passing through a processor (e.g., ...). Figure 8 In the case of the memory controller, etc., in the processor 39), the pads of the core die are directly accessed. Signals for performing various tests, such as tests based on the IEEE 1500 standard, can be transmitted through the direct access area (DA). In one embodiment, the direct access area (DA) can provide low latency and high bandwidth, and can reduce power consumption.
[0058] The logic circuit region (DA RTL) can be implemented as register-transfer-level (RTL) circuitry and can control or test the core die stacked on top of the base die 2 (e.g., Figure 8 The operation of core die 321-1 to core die 321-L. The logic circuit region (DA RTL) includes... Figure 1 The register transfer level circuit 13 shown in the figure outputs from... Figure 1 The first transmission data TD1 and the second transmission data TD2 received by the transmission data generation circuit 11 shown are used as test output data T-DA.
[0059] Figure 8 A storage system 3 according to an embodiment of the present disclosure is shown.
[0060] like Figure 8 As shown, the storage system 3 includes a printed circuit board (PCB) 31, a substrate 33, an interposer 35, a storage device 37, and a processor 39.
[0061] Printed circuit board 31 connects various electronic components to form an electronic circuit. The electronic circuit includes a storage system 3. A copper (Cu) layer, a solder mask, silkscreen printing, etc., are formed on the printed circuit board 31. Circuit paths for transmitting or carrying signals or electricity are formed in the copper (Cu) layer. In one embodiment, the solder mask prevents or mitigates damage to the circuit and protects specific areas of the soldered components. Silkscreen printing uses characters or symbols printed on the surface of the printed circuit board 31 to indicate the location or information of the electronic components.
[0062] A substrate 33 is disposed on a printed circuit board 31 (with bump pads therein, such as bump pad 311) and mechanically supports the interposer 35, storage device 37, and processor 39. The substrate 33 serves as the physical basis of the printed circuit board 31 and is an insulator. The substrate 33 may include materials such as flame retardant 4 (FR4) (an insulator made of glass fiber and epoxy resin), ceramics (which can withstand high temperatures, have suitable thermal conductivity, and are used in high-frequency circuits), and polyimide (which is used as a basic material for flexible PCBs due to its flexibility).
[0063] Intermediate layer 35 is disposed on substrate 33 (with bump pads therein) and includes wiring connecting electronic components (e.g., storage device 37 and processor 39) that have mismatched size specifications or pin arrangements or different spacing. Intermediate layer 35 converts signals from different interfaces, such as Double Data Rate (DDR), High Bandwidth Memory (HBM), and Peripheral Component Interconnect Fast (PCIe).
[0064] Storage device 37 is disposed on interposer layer 35, with pads (e.g., microbump pads 313) therebetween. Storage device 37 stores data received from processor 39 or outputs stored data to processor 39 under the control of processor 39. Storage device 37 includes base die 320 and multiple core dies 321-1 to 321-L, where L is an integer greater than 1. Core dies 321-1 to 321-L are stacked on base die 320, with microbump pads therebetween. Base die 320 and core dies 321-1 to 321-L are vertically connected to each other using through-silicon vias (TSVs) and microbump pads. Base die 320 controls efficient data transfer between processor 39 and core dies 321-1 to 321-L. Base die 320 receives input / output power supply voltage (drain voltage for I / O, also known as output stage drain power supply voltage) VDDQ as an operating voltage used during the operation of internal circuitry included in base die 320. The base die 320 receives the input / output power supply voltage VDDQ from the printed circuit board 31 via the substrate 33 and the interposer 35. The input / output power supply voltage VDDQ is the voltage supplied to the buffer transmitting data and is distinct from or different from the power supply voltage VDD. Core dies 321-1 to 321-L use the peripheral voltage VPERI as their operating voltage during the operation of the internal circuitry included in core dies 321-1 to 321-L. Core dies 321-1 to 321-L generate the peripheral voltage VPERI from the input / output power supply voltage VDDQ received via the base die 320. Core dies 321-1 to 321-L generate the peripheral voltage VPERI at a voltage level lower than the input / output power supply voltage VDDQ and use the peripheral voltage VPERI as their operating voltage. Each of the core dies 321-1 to 321-L includes multiple channel regions, for example, eight independently operating channel regions or forty-six channel regions. Each of the multiple channel regions is assigned an independently operating channel for receiving or transmitting data. The number L of core dies 321-1 to 321-L can be four, eight, thirty-two, forty-six, etc. For example, when each of core dies 321-1 to 321-12 has eight channels, core dies 321-1 to 321-4, core dies 321-5 to 321-8, and core dies 321-9 to 321-12 each include a thirty-two channel region, and transmit and receive data with processor 39 in units of rank (memory column) including thirty-two channels.
[0065] The base die 320 may include Figure 1The test circuit 1 shown in the diagram. The test circuit 1 included in the base die 320 does not determine whether a fault has occurred for each read operation during channel compression testing or pseudo-channel compression testing. Instead, the test circuit 1 outputs an activated fault signal F-EN when a preset number of errors occur during multiple read operations, or outputs a fault signal F-EN that is fixed in an activated state when an error occurs for the first time during multiple read operations in a fixed test, thereby reducing test time in one embodiment.
[0066] These concepts have been disclosed in conjunction with various examples and embodiments. Those skilled in the art will understand that various modifications, additions, and substitutions can be made without departing from the scope and technical concepts of this disclosure. The embodiments disclosed in this specification should be considered from an illustrative rather than restrictive perspective. The scope of this disclosure is not limited to the foregoing description. All variations within the meaning and equivalents of the claims are included within its scope.
Claims
1. A test circuit, comprising: The data generation circuit for transmission has the following characteristics: An error flag is generated based on selected compressed data, which is generated from a channel compression test or a pseudo-channel compression test. An error count signal is generated by comparing an error count signal with a target code to produce a fault signal, the error count signal being generated by counting the error flags; and Drive data transmission based on the fault signal; and A register-transfer-level circuit generates test output data from the transmitted data and transmits the test output data to the direct access area.
2. The test circuit according to claim 1, wherein, The transmission data generation circuit includes a selection circuit that selects and outputs one of compressed data and pseudo-compressed data as the selected compressed data based on a compression test selection signal.
3. The test circuit according to claim 2, wherein, The selection circuit receives the compressed data, which is generated during the channel compression test by outputting data through multiple channels from the data stored in the storage cell array for each pseudo-channel and burst length comparison.
4. The test circuit according to claim 2, wherein, The selection circuit receives the pseudo-compressed data, which is generated during the pseudo-channel compression test by comparing multiple burst lengths of data output through a specific pseudo-channel from the data stored in the storage cell array.
5. The test circuit according to claim 2, wherein, The selection circuit: Receive the compression test selection signal at the first logic level during the channel compression test to select the compressed data as the selected compressed data; as well as The compression test selection signal, which is at the second logic level during the pseudo-channel compression test, is received to select the pseudo-compressed data as the selected compressed data.
6. The test circuit according to claim 1, wherein, The data transmission generation circuit includes a fault determination circuit, which generates an activated fault signal based on the selected compressed data when the error flag is activated a preset number of times.
7. The test circuit according to claim 6, wherein, The fault determination circuit includes an error flag generation circuit that generates the error flag based on the selected compressed data.
8. The test circuit according to claim 7, wherein, The error flag generation circuit: When the logic level of at least one bit among all the bits included in the selected compressed data is different from that of the other bits, the error flag activated at the first logic level is generated. as well as When all bits included in the selected compressed data have the same logic level, the error flag is generated and deactivated at the second logic level.
9. The test circuit according to claim 6, wherein, The fault determination circuit includes an error flag counter that generates the error count signal based on the error flag and a read strobe signal.
10. The test circuit according to claim 9, wherein, The error flag counter counts the activated error flags synchronously with the read strobe signal activated during the read operation for the channel compression test or the pseudo-channel compression test to generate the error count signal.
11. The test circuit according to claim 6, wherein, The fault determination circuit includes a fault signal generation circuit, which compares the error count signal with the target code to generate the fault signal.
12. The test circuit according to claim 11, wherein, The fault signal generation circuit: When the binary bit group of the error count signal is the same as the binary bit group of the target code, the fault signal activated at the first logic level is generated. as well as When the binary bit group of the error count signal is different from the binary bit group of the target code, the fault signal is generated and deactivated at the second logic level.
13. The test circuit according to claim 1, in, The transmitted data includes first transmitted data and second transmitted data, and The transmission data generation circuit includes a first driving circuit, which generates the first transmission data from the selected compressed data and the fault signal based on a first test mode signal.
14. The test circuit according to claim 13, wherein, First driving circuit: When the channel compression test or the pseudo channel compression test is executed, the first test mode signal set to the first logic level is received to drive the first transmission data based on the fault signal; as well as When the channel compression test and the pseudo-channel compression test are not performed, the first test mode signal set to the second logic level is received to drive the first transmission data based on the selected compressed data.
15. The test circuit according to claim 1, in, The transmitted data includes first transmitted data and second transmitted data, and The transmission data generation circuit includes a second driving circuit, which generates the second transmission data from the selected compressed data and the fault signal based on a second test mode signal.
16. The test circuit according to claim 15, wherein, Second driving circuit: During the fixed test execution, the second transmission data at a fixed logic level is generated based on the fault signal; and When the fixed test is not performed, the second transmission data is generated based on the selected compressed data.
17. A test circuit, comprising: The selection circuit selects and outputs one of the compressed data and pseudo-compressed data as the selected compressed data based on the compression test selection signal; An error flag generation circuit that generates an error flag based on the selected compressed data; An error flag counter that generates an error count signal based on the error flag and the read strobe signal; A fault signal generation circuit compares the error count signal with the target code to generate a fault signal; A first driving circuit generates first transmission data from the selected compressed data and the fault signal based on a first test mode signal; as well as The second driving circuit generates second transmission data from the selected compressed data and the fault signal based on the second test mode signal.
18. The test circuit according to claim 17, wherein, The selection circuit receives the compressed data, which is generated during the channel compression test by outputting data through multiple channels from the data stored in the storage cell array for each pseudo-channel and burst length comparison.
19. The test circuit according to claim 17, wherein, The selection circuit receives the pseudo-compressed data, which is generated during the pseudo-channel compression test by comparing multiple burst lengths of data output through a specific pseudo-channel from the data stored in the storage cell array.
20. The test circuit according to claim 17, wherein, The selection circuit: Receive the compression test selection signal at the first logic level during the channel compression test to select the compressed data as the selected compressed data; as well as The compression test selection signal, which is at the second logic level during the pseudo-channel compression test, is received to select the pseudo-compressed data as the selected compressed data.
21. The test circuit according to claim 17, wherein, The error flag generation circuit: When the logic level of at least one bit among all the bits included in the selected compressed data is different from that of the other bits, the error flag activated at the first logic level is generated. as well as When all bits included in the selected compressed data have the same logic level, the error flag is generated and deactivated at the second logic level.
22. The test circuit according to claim 17, wherein, The error flag counter counts the activated error flags synchronously with the read strobe signal activated during the read operation for the channel compression test or the pseudo-channel compression test to generate the error count signal.
23. The test circuit according to claim 17, wherein, The fault signal generation circuit: When the binary bit group of the error count signal is the same as the binary bit group of the target code, the fault signal activated at the first logic level is generated. as well as When the binary bit group of the error count signal is different from the binary bit group of the target code, the fault signal is generated and deactivated at the second logic level.
24. The test circuit according to claim 17, wherein, First driving circuit: When the channel compression test and the pseudo channel compression test are executed, the first test mode signal set to the first logic level is received to drive the first transmission data based on the fault signal; as well as When the channel compression test and the pseudo-channel compression test are not performed, the first test mode signal set to the second logic level is received to drive the first transmission data based on the selected compressed data.
25. The test circuit according to claim 17, wherein, Second driving circuit: During the fixed test execution, the second transmission data at a fixed logic level is generated based on the fault signal; and When the fixed test is not performed, the second transmission data is generated based on the selected compressed data.
26. A semiconductor device comprising a plurality of core dies stacked on a base die, in, Each of the plurality of core dies includes a plurality of core regions allocated to a plurality of channels, and each of the plurality of channels includes a plurality of pseudo channels, and The basic die includes a test circuit, wherein the test circuit is: An error flag is generated based on selected compressed data, which is generated during a channel compression test or a pseudo-channel compression test. An error count signal is generated by comparing an error count signal with a target code to produce a fault signal, the error count signal being generated by counting the error flags; and Data transmission is driven based on the fault signal.
Citation Information
Patent Citations
An iris device, a camera device and an optical apparatus
KR1020250059295A
