Bad wafer recovery method and device and related equipment

By testing and determining the recovery type of IO modules in the core particles, efficient recycling and packaging of the core particles is achieved, solving the problem that the core particles cannot work normally during the chip manufacturing process and reducing the cost of the chip.

CN120198109APending Publication Date: 2025-06-24HAIGUANG INFORMATION TECH (SUZHOU) CO LTD
View PDF 0 Cites 0 Cited by

Patent Information

Application Number
CN202411319656.X
Authority / Receiving Office
CN · China
Patent Type
Applications(China)
Current Assignee / Owner
Filing Date
2024-09-20
Publication Date
2025-06-24

AI Technical Summary

Technical Problem

As the chip manufacturing process enters the nanoprocess, the yield of chip mass production decreases, resulting in inevitable process deviations, dust loss and other factors in the production and manufacturing process, resulting in the core particles not working normally and excessive waste of chips.

Method used

By obtaining the preset chip, test the failure status of multiple groups of IO modules in the core particles, determine the recovery type of the core particles, and record the IO number of the failed IO module, determine the packaging strategy of the core particles corresponding to the recovery type based on the recorded IO number, and then package the core particles corresponding to the recovery type to obtain the recovery chip.

Benefits of technology

The recycling rate of core particles is improved, the cost of chip is reduced, and the recovery chip is achieved without affecting its function while recycling defective core particles.

✦ Generated by Eureka AI based on patent content.

Smart Images

  • Figure CN120198109A_ABST
    Figure CN120198109A_ABST
Patent Text Reader

Abstract

The embodiment of the invention provides a bad chip recovery method and device and related equipment, and the method comprises the steps: obtaining a preset chip, the preset chip comprises a plurality of core particles, the core particles comprise a plurality of groups of IO modules which are symmetrically distributed, and the IO modules are correspondingly provided with IO numbers for indicating the placement positions of the IO modules in the core particles; wherein different chip particles are provided with different packaging positions in the preset chip; testing the plurality of groups of IO modules, and determining failure conditions of the plurality of groups of IO modules; according to the failure condition, the recovery type of the core particles where the multiple groups of IO modules are located is determined, and the IO numbers of the failure IO modules are recorded; according to the recorded IO number, determining a packaging strategy of the core particles of the corresponding recovery type; and according to the packaging position and the packaging strategy, packaging the core particles of the corresponding recovery type to obtain a recovered chip, thereby improving the recovery utilization rate of the core particles and reducing the cost of the chip.
Need to check novelty before this filing date? Find Prior Art

Description

Technical Field

[0001] The embodiments of the present application relate to the field of chip technology, and particularly to a defective die recycling method, device and related equipment thereof. Background Art

[0002] A die is a unit chip that can implement certain functions and contains an interconnection interface. Multiple dies can be packaged through 2D, 2.5D, 3D and other packaging technologies to obtain a chiplet, and the dies can be interconnected based on the interconnection interface, thereby forming a chip with expected functions, such as a SOC (System On Chip) chip.

[0003] However, as the chip manufacturing process enters the nanometer process, the yield of chip mass production has decreased significantly. This is mainly because during the production and manufacturing process of the die, inevitable process deviations, dust falling and other factors lead to abnormal operation. To avoid excessive waste of chips, defective dies can be recycled. Defective dies are screened out through testing means and can be reused after corresponding processing.

[0004] Under this background, how to provide a defective die recycling solution to improve the recycling utilization rate of dies and reduce the cost of chips has become a technical problem that needs to be solved urgently by those skilled in the art. Summary of the Invention

[0005] In view of this, the embodiments of the present application provide a defective die recycling method, device and related equipment thereof to achieve improving the recycling utilization rate of dies and reducing the cost of chips.

[0006] In a first aspect, the embodiments of the present application provide a defective die recycling method, including:

[0007] Obtain a preset chip, where the preset chip includes multiple dies, the dies include multiple groups of IO modules symmetrically distributed, and the IO modules are correspondingly provided with IO numbers for indicating the placement positions of the IO modules in the die; wherein, different dies are set at different packaging positions in the preset chip;

[0008] Test the multiple groups of IO modules to determine the failure conditions of the multiple groups of IO modules;

[0009] According to the failure conditions, determine the recycling types of the dies where the multiple groups of IO modules are located, and record the IO numbers of the failed IO modules;

[0010] According to the recorded IO numbers, determine the packaging strategies of the dies corresponding to the recycling types;

[0011] According to the packaging positions and the packaging strategies, package the dies corresponding to the recycling types to obtain recycled chips.

[0012] Optionally, the IO module includes an IO controller and an IO physical layer;

[0013] Testing the multiple groups of IO modules to determine the failure conditions of the multiple groups of IO modules specifically includes: testing the IO controllers and IO physical layers of the multiple groups of IO modules respectively to determine the defect information of the IO controllers and IO physical layers in the multiple groups of IO modules.

[0014] Optionally, based on the failure conditions of the multiple groups of IO modules, determining the recycling types of the dies where the multiple groups of IO modules are located, and recording the IO numbers of the failed IO modules and the die numbers of the dies corresponding to the recycling types, specifically including: based on the defect information of the IO controllers and IO physical layers in the multiple groups of IO modules, determining the recycling types of the dies where the multiple groups of IO modules are located, and recording the IO numbers of the IO modules where the failed IO controllers and IO physical layers are located in the defect information.

[0015] Optionally, the IO controller includes a first management and configuration unit and a data processing unit, and a first scan chain is provided corresponding to the first management and configuration unit and the data processing unit. The first scan chain is used to perform a scan test on the execution logics of the first management and configuration unit and the data processing unit; the first management and configuration unit includes a first configuration domain and a first core domain. The first core domain is used to configure the core functions of the IO controller, and a second scan chain is provided corresponding to the first core domain. The second scan chain is used to perform a scan test on the execution logic of the first core domain; the first configuration domain is used to configure the communication parameters of the IO controller;

[0016] Testing the IO controllers of the multiple groups of IO modules includes:

[0017] For the IO controllers of the multiple groups of IO modules, respectively based on the first scan chain, scanning and testing the execution logics of the first management and configuration unit and the data processing unit; and based on the second scan chain, scanning and testing the execution logic of the first core domain.

[0018] Optionally, the IO physical layer includes a second management and configuration unit and a physical layer unit, and a third scan chain is provided corresponding to the second management and configuration unit and the physical layer unit. The third scan chain is used to perform a scan test on the execution logics of the second management and configuration unit and the physical layer unit; the second management and configuration unit includes a second configuration domain and a second core domain. The second core domain is used to configure the core functions of the IO physical layer, and a fourth scan chain is provided corresponding to the second core domain. The fourth scan chain is used to perform a scan test on the execution logic of the second core domain; the second configuration domain is used to configure the component information of the IO physical layer;

[0019] Test the IO physical layer of the multiple groups of IO modules, including:

[0020] For the IO physical layer of the multiple groups of IO modules, respectively based on the third scan chain, scan and test the execution logics of the second management configuration unit and the physical layer unit; and, based on the fourth scan chain, scan and test the execution logic of the second core domain.

[0021] Optionally, the defect information includes:

[0022] The IO controller of one or more groups of IO modules in the multiple groups of IO modules fails and / or the IO physical layer fails, and the first core domains of the first management configuration units of the IO controllers of the multiple groups of IO modules are all valid, and the second core domains of the second management configuration units of the IO physical layers are all valid;

[0023] Or,

[0024] The first core domain of the first management configuration unit of the IO controller of at least one group of IO modules in the multiple groups of IO modules fails, and / or the second core domain of the second management configuration unit of the IO physical layer fails.

[0025] Optionally, the recycling type includes recyclable and non-recyclable;

[0026] Determine the recycling type of the chiplet where the multiple groups of IO modules are located according to the defect information of the IO controller and the IO physical layer in the multiple groups of IO modules, including:

[0027] If the defect information is that the IO controller of one or more groups of IO modules in the multiple groups of IO modules fails and / or the IO physical layer fails, and the first core domains of the first management configuration units of the IO controllers of the multiple groups of IO modules are all valid, and the second core domains of the second management configuration units of the IO physical layers are all valid, determine that the recycling type of the chiplet where the multiple groups of IO modules are located is recyclable;

[0028] If the defect information is that the first core domain of the first management configuration unit of the IO controller of at least one group of IO modules in the multiple groups of IO modules fails, and / or the second core domain of the second management configuration unit of the IO physical layer fails, determine that the recycling type of the chiplet where the multiple groups of IO modules are located is non-recyclable.

[0029] Optionally, if the recycling type of the chiplet where the multiple groups of IO modules are located is recyclable, then execute to determine the packaging strategy of the chiplet of the corresponding recycling type according to the recorded IO number, and, according to the packaging position and the packaging strategy, package the chiplet of the corresponding recycling type to obtain a recycled chip;

[0030] If the recycling type of the die where the multiple groups of IO modules are located is non-recyclable, discard the die.

[0031] Optionally, in the preset chip, for the multiple groups of IO modules corresponding to the die, one or more defective IO numbers are preset; among them, the multiple IO modules corresponding to the multiple defective IO numbers are symmetrically distributed in the die;

[0032] The encapsulation strategy includes translational encapsulation and rotational encapsulation;

[0033] The determining the encapsulation strategy of the die corresponding to the recycling type according to the recorded IO number includes:

[0034] If the recorded IO number is the same as the defective IO number, determine that the encapsulation strategy of the die corresponding to the recycling type is translational encapsulation;

[0035] If the recorded IO number corresponds to the defective IO number, determine that the encapsulation strategy of the die corresponding to the recycling type is rotational encapsulation.

[0036] Optionally, after encapsulating the die corresponding to the recycling type according to the encapsulation position and the encapsulation strategy to obtain a recycled chip, it further includes:

[0037] Initialize the recycled chip and check the position identifier of the die, where the position identifier is used to mark the usage status of the multiple IO modules of the die.

[0038] In a second aspect, an embodiment of the present application provides a defective chip recycling device, including:

[0039] A chip acquisition unit, configured to acquire a preset chip, where the preset chip includes multiple dice, the die includes multiple groups of symmetrically distributed IO modules, and the IO modules are correspondingly provided with IO numbers for indicating the placement positions of the IO modules in the die; among them, different dice are provided with different encapsulation positions in the preset chip;

[0040] A failure test unit, configured to test the multiple groups of IO modules to determine the failure conditions of the multiple groups of IO modules;

[0041] A recycling determination unit, configured to determine the recycling type of the die where the multiple groups of IO modules are located according to the failure conditions, and record the IO numbers of the failed IO modules;

[0042] A strategy determination unit, configured to determine the encapsulation strategy of the die corresponding to the recycling type according to the recorded IO number;

[0043] A defective chip encapsulation unit, configured to encapsulate the die corresponding to the recycling type according to the encapsulation position and the encapsulation strategy to obtain a recycled chip.

[0044] Optionally, it further includes:

[0045] An initialization unit, configured to initialize the recycling chip and check the position identifier of the die, where the position identifier is used to mark the usage status of multiple IO modules of the die.

[0046] In a third aspect, an embodiment of the present application provides a chip, which is obtained by recycling based on the defective die recycling method described in the first aspect above.

[0047] Optionally, the chip includes multiple dies, each die includes multiple groups of symmetrically distributed IO modules, and the IO module includes an IO controller and an IO physical layer.

[0048] Optionally, the IO controller includes a first management and configuration unit and a data processing unit, and a first scan chain is provided corresponding to the first management and configuration unit and the data processing unit, and the first scan chain is used to perform a scan test on the execution logic of the first management and configuration unit and the data processing unit; the first management and configuration unit includes a first configuration domain and a first core domain, the first core domain is used to configure the core functions of the IO controller, and a second scan chain is provided corresponding to the first core domain, and the second scan chain is used to perform a scan test on the execution logic of the first core domain; the first configuration domain is used to configure the communication parameters of the IO controller.

[0049] Optionally, the IO physical layer includes a second management and configuration unit and a physical layer unit, and a third scan chain is provided corresponding to the second management and configuration unit and the physical layer unit, and the third scan chain is used to perform a scan test on the execution logic of the second management and configuration unit and the physical layer unit; the second management and configuration unit includes a second configuration domain and a second core domain, the second core domain is used to configure the core functions of the IO physical layer, and a fourth scan chain is provided corresponding to the second core domain, and the fourth scan chain is used to perform a scan test on the execution logic of the second core domain; the second configuration domain is used to configure the component information of the IO physical layer.

[0050] In a fourth aspect, an embodiment of the present application provides a computing device, including the defective die recycling device described in the second aspect above.

[0051] In a fifth aspect, an embodiment of the present application provides a storage medium, which stores one or more computer-executable instructions, and when the one or more computer-executable instructions are executed, the defective die recycling method described in the first aspect above is implemented.

[0052] In a sixth aspect, an embodiment of the present application provides a computer program product, which is characterized by including one or more computer-executable instructions that, when executed, implement the defective die recovery method described in the first aspect above.

[0053] The defective die recovery method provided by the embodiment of the present application obtains a preset chip including a plurality of dies. When the die includes multiple groups of IO modules symmetrically distributed, the multiple groups of IO modules can be tested to determine the failure conditions of the multiple groups of IO modules. Then, according to the failure conditions, the recovery type (i.e., the recovery strategy) of the die is determined, improving the recovery utilization rate of the die. Moreover, when the IO module is provided with an IO number for indicating the placement position of the IO module in the die, and different dies are provided with different packaging positions in the preset chip, it is possible to record the IO numbers of the failed IO modules according to the failure conditions of the multiple groups of IO modules. Based on the recorded IO numbers, the packaging strategy of the die corresponding to the recovery type is determined. Thus, according to the packaging position and the packaging strategy, the die corresponding to the recovery type can be packaged to obtain a recovered chip, enabling the obtained recovered chip to have its function unaffected while recovering defective dies, reducing the cost of the chip. BRIEF DESCRIPTION OF THE DRAWINGS

[0054] To more clearly illustrate the technical solutions in the embodiments of the present application or the prior art, the following will briefly introduce the drawings required for use in the description of the embodiments or the prior art. Obviously, the drawings in the following description are only the embodiments of the present application. For those of ordinary skill in the art, without creative efforts, other drawings can also be obtained based on the provided drawings.

[0055] Figure 1 is an optional structural schematic diagram of a chip.

[0056] Figure 2 and Figure 3 are optional architecture schematic diagrams of different CPU chips composed of multiple dies.

[0057] Figure 4 is an optional flowchart of the defective die recovery method provided by the embodiment of the present application.

[0058] Figure 5 is an internal structural schematic diagram of CPU Chip 1 provided by the embodiment of the present application.

[0059] Figure 6 is an optional internal structural schematic diagram of the IO controller provided by the embodiment of the present application.

[0060] Figure 7 is an optional internal structural schematic diagram of the IO physical layer provided by the embodiment of the present application.

[0061] Figure 8 is another optional flowchart diagram of the defective chip recycling method provided by the embodiments of the present application.

[0062] Figure 9 and Figure 10 is a schematic diagram of the package of the CPU chip 1 provided by the embodiments of the present application.

[0063] Figure 11 is an optional block diagram of the defective chip recycling device provided by the embodiments of the present application. Detailed implementation manners

[0064] Next, the technical solutions in the embodiments of the present application will be clearly and completely described in conjunction with the accompanying drawings in the embodiments of the present application. Obviously, the described embodiments are only a part of the embodiments of the present application, rather than all the embodiments. All other embodiments obtained by those of ordinary skill in the art based on the embodiments of the present application without creative efforts shall fall within the protection scope of the present application.

[0065] For ease of understanding the composition of the chip, Figure 1 an optional structural schematic diagram of a chip is exemplarily shown. As Figure 1 shown, the chip may include multiple die (die 1, die 2, die 3, and die 4 are shown as examples in the figure). Among them, die 1, die 2, die 3, and die 4 may be the same type of die with the same function, or different types of die with different functions. The die 1, die 2, die 3, and die 4 are packaged and integrated through a substrate to obtain Figure 1 the shown chip.

[0066] It should be noted that in the chip design stage, the overall architecture of the chip can be decomposed according to different computing units or functional units required for the chip to achieve the expected function, and one or more die corresponding to different computing units or functional units can be designed according to the architecture decomposition scheme of the chip, and the die are interconnected and packaged together to achieve the expected function of the chip as a whole. Among them, since a die is a unit chip that can achieve a certain function and contains an interconnection interface, multiple die that constitute the same computing unit or functional unit can be called a chiplet, and the multiple die in the chiplet can be the same or different. And, based on the interconnection and communication between different modules, the die in the same chiplet can also be used to form different chiplets. Moreover, with the continuous development of semiconductor technology, a single die can contain various functional blocks, so as to achieve effects such as computing, etc. Therefore, in certain specific scenarios, the single die can be regarded as a special chiplet.

[0067] In an optional example, Figure 2 and Figure 3Shows a schematic diagram of an alternative architecture of different CPU chips composed of multiple die, Figure 2 is CPU chip 1 with 32 cores, Figure 3 is CPU chip 2 with 64 cores.

[0068] As Figure 2 shown, this 32-core CPU chip 1 may include 2 IO die (IO die m1 and IO die m2 respectively) and 2 CPU die (CPU die n1 and CPU die n2 respectively). Among them, the 2 IO die and the 2 CPU die can be regarded as a chipset, such as chipset A. The 2 IO die are responsible for input / output (I / O) tasks; the 2 CPU die are used to implement the 32-core channel of the CPU.

[0069] As Figure 3 shown, this 64-core CPU chip 2 may include 2 IO die (IO die M1 and IO die M2 respectively) and 4 CPU die (CPU die N1, CPU die N2, CPU die N3 and CPU die N4 respectively). Among them, the 2 IO die and the 4 CPU die can be regarded as a chipset, such as chipset B. The 2 IO die are responsible for input / output (I / O) tasks; the 4 CPU die are used to implement the 64-core channel of the CPU.

[0070] It should be further noted that the production and manufacturing of die refer to etching a pre-determined pattern on a wafer through a lithography machine to obtain multiple identical unit chips. However, due to inevitable process deviations, dust falling, and other factors during the production and manufacturing of die, there are defective die that cannot work properly among the obtained die. Therefore, in order to avoid the chips obtained by packaging defective die from being discarded because they cannot achieve the corresponding functions, resulting in excessive waste of chips, it is necessary to recycle defective die. Among them, the recycling of defective die refers to screening out the defective die during the production and manufacturing process through testing means, and continuing to use the defective chips after corresponding processing.

[0071] As an alternative implementation, defective die recycling can be performed on the computing cores that make up the chip. For example, multiple CPU computing cores (i.e., CPU die) are integrated inside the CPU chip. When some computing cores of the CPU chip are defective, the defective computing cores can be turned off to achieve the recycling of defective die. Moreover, the CPU chip with the defective computing cores turned off can also be recycled and reused as a chip product with reduced performance.

[0072] However, with the improvement of chip manufacturing technology and the increase in the input / output (I / O) bandwidth requirements for data execution, the IO performance of chips is also constantly improving, the IO design is becoming more and more complex, and the probability of defects in IO and IO control logic is getting higher and higher. Therefore, when the IO of the core particle and the IO-related communication controller [such as the DDR (Double Data Rate) physical layer and DDR controller] logic are defective, the defective core particle cannot be recycled, which reduces the recycling rate of the core particle. Furthermore, due to the reduced recycling rate of the core particle, the cost of the chip is also correspondingly increased, such as the chip R&D cost, raw material cost, manufacturing cost, etc.

[0073] In view of this, the embodiment of the present application proposes an improved bad chip recycling scheme, which determines the recycling strategy based on the defect conditions of multiple IO modules of the core particle, so as to improve the recycling rate of the core particle and reduce the cost of the chip. Among them, the core particle is responsible for performing specific I / O operations and is a hardware component at the physical level; the IO module is a communication module, which can be an IO interface module, which is responsible for and manages various hardware interface standards (such as PCI Express, USB, SPI, etc.). Among them, some IO modules can be used in conjunction with system programs.

[0074] Based on the above ideas, as an optional implementation, the embodiment of the present application can obtain a preset chip including multiple core particles. When the core particle contains multiple groups of symmetrically distributed IO modules, the multiple groups of IO modules can be tested to determine the failure conditions of the multiple groups of IO modules, and then determine the recovery type (i.e., recovery strategy) of the core particles according to the failure conditions, thereby improving the recycling rate of the core particles; and, when the IO module is provided with an IO number for indicating the placement position of the IO module in the core particle, and different core particles are provided with different packaging positions in the preset chip, the IO number of the failed IO module can be recorded according to the failure conditions of the multiple groups of IO modules, so that the packaging strategy of the core particles of the corresponding recycling type can be determined based on the recorded IO number, so that the core particles of the corresponding recycling type can be packaged according to the packaging position and the packaging strategy to obtain recycled chips, so that while recycling defective core particles, recycled chips whose functions are not affected can be obtained, thereby reducing the cost of chips.

[0075] in, Figure 4 The optional flow chart of the bad chip recycling method of the embodiment of the present application is exemplarily shown. Figure 4 As shown, the following steps may be included:

[0076] Step S100, obtaining a preset chip.

[0077] The preset chip refers to a pre-designed chip for implementing a chip with expected functions. The preset chip may include multiple dielets, where the functions implemented by the multiple dielets may be the same or different. For example: IO dielets, CPU dielets. When the implemented functions are the same, the multiple dielets may be of the same type; when the implemented functions are different, the multiple dielets may be of different types. Different dielets are set at different packaging positions in the preset chip.

[0078] The dielet may include multiple groups of IO modules symmetrically distributed. Among them, the multiple groups of IO modules may include functional IO modules and redundant IO modules. The functional IO modules can be understood as the IO modules required to implement the product functions; the redundant IO modules can be IO modules with the same function implementation as the functional IO modules, so as to ensure the symmetrical distribution of multiple groups of IO modules in the dielet. And, the IO modules are set with IO numbers, which are used to indicate the placement positions of the IO modules in the dielet.

[0079] It should be noted that the function of the dielet can be realized based on the execution of the IO module. The IO module can be understood as a communication module. Therefore, in order to improve the flexibility and efficiency of the system, one or more groups of IO modules can be integrated in the dielet to provide multiple standard communication interfaces, so that the same dielet can interact with different external devices. Among them, in the IO dielet, the IO module as the communication module can be specifically an IO controller; in the CPU dielet, the IO module as the communication module can be specifically DDR. Taking the CPU dielet in CPU chip 1 as an example, Figure 5 Exemplarily shows the internal structure schematic diagram of CPU chip 1. As Figure 5 shown, CPU chip 1 contains two CPU dielets, and each CPU dielet may contain 8 groups of IO modules symmetrically distributed (the 8 groups of IO modules are communication modules. In the example taking the CPU dielet as an example, Figure 8 the DDRs shown as communication modules are DDR00 to DDR07, and DDR00' to DDR07' respectively). Among them, in the 8 groups of IO modules, there can be 6 groups of functional IO modules and 2 groups of redundant IO modules.

[0080] Step S110, test multiple groups of IO modules to determine the failure conditions of the multiple groups of IO modules.

[0081] Since defects can be screened out by testing means, and the dielet has multiple groups of IO modules, therefore, the failure conditions of the multiple groups of IO modules can be determined by testing the multiple groups of IO modules.

[0082] Step S120, according to the failure conditions of the multiple groups of IO modules, determine the recycling type of the dielets where the multiple groups of IO modules are located, and record the IO numbers of the failed IO modules.

[0083] Among them, the recycling type of the die in the embodiments of the present application can be understood as the recycling strategy corresponding to the defective die. For example, defective die A can be recycled; or, defective die A cannot be recycled. And based on the failure conditions obtained by testing multiple groups of IO modules, the recycling type of the die can correspond to the failure conditions of multiple groups of IO modules, and is used to indicate the recycling status of the die. The failure conditions of multiple groups of IO modules refer to the situation of defective IO modules existing in multiple groups of IO modules. By determining the failure conditions of multiple groups of IO modules, the IO numbers of the failed IO modules can be recorded.

[0084] It can be understood that in the embodiments of the present application, by testing multiple groups of IO modules, the failure conditions of multiple groups of IO modules are determined, and then the recycling type of the die is determined, so that based on the failure conditions of multiple groups of IO modules, corresponding recycling strategies are obtained, such as the die can be recycled or cannot be recycled, which can effectively improve the recycling utilization rate of the die.

[0085] Step S130: Determine the packaging strategy of the die corresponding to the recycling type according to the recorded IO numbers.

[0086] Since the IO module is correspondingly provided with an IO number indicating its placement position in the die, when multiple IO modules are symmetrically distributed in the die, it is possible that multiple IO modules all have defects in the die. Therefore, in order to maximize the recycling of defective chips, isolate the potential failure influence range, and increase its ability to resist random defects, the packaging strategy of the die corresponding to the recycling type can be determined according to the IO numbers of the failed IO modules in the recorded multiple groups of IO modules, so as to optimize the chip structure and enhance the reliability and stability of the defective die. The packaging strategy can be the packaging method of the die where the IO module is located determined based on the recorded IO numbers, such as translational packaging, rotational packaging, etc.

[0087] Step S140: Package the die corresponding to the recycling type according to the packaging position and the packaging strategy to obtain a recycled chip.

[0088] After determining the recycling type of the die and the packaging strategy of the die in the embodiments of the present application, the die corresponding to the recycling type can be packaged according to the preset packaging position of the die and the packaging strategy to obtain a recycled chip, thereby realizing the recycling of defective dies. At the same time, based on the reasonably planned packaging position of the die, when recycling defective dies to obtain a chip with expected functions, the cost of the chip is also reduced without affecting the expected functions of the chip.

[0089] It can be seen that in the embodiments of the present application, by obtaining a preset chip including multiple die, when the die includes multiple groups of IO modules symmetrically distributed, the multiple groups of IO modules can be tested to determine the failure conditions of the multiple groups of IO modules, and then according to the failure conditions, the recovery type (i.e., the recovery strategy) of the die can be determined, improving the recovery utilization rate of the die; and, when the IO module is provided with an IO number for indicating its placement position in the die, and different die are provided with different packaging positions in the preset chip, it is possible to record the IO numbers of the failed IO modules according to the failure conditions of the multiple groups of IO modules, so that based on the recorded IO numbers, the packaging strategy of the die corresponding to the recovery type can be determined, and thus the die corresponding to the recovery type can be packaged according to the packaging position and the packaging strategy to obtain a recovered chip, so that while recovering the defective die, a recovered chip with unaffected function is obtained, reducing the cost of the chip.

[0090] In some embodiments, to implement the functions of the IO module, the IO module may include an IO controller and an IO physical layer. The IO controller is responsible for managing the communication between the hardware device and the CPU and performing control operations for data transmission, such as sending commands, receiving data, processing interrupts, etc. The IO physical layer is responsible for data transmission and data conversion for the CPU to process. Among them, in the example where the IO module as a communication module is specifically a DDR module, the DDR module may include a DDR controller and a DDR physical layer.

[0091] Based on the design of the IO controller and the IO physical layer, when testing multiple groups of IO modules to determine the failure conditions of the multiple groups of IO modules, it may specifically be: testing the IO controllers and IO physical layers of the multiple groups of IO modules separately to determine the defect information of the IO controllers and IO physical layers in the multiple groups of IO modules, and the defect information is the information of the failed IO controllers and IO physical layers in the multiple groups of IO modules.

[0092] Further, in some embodiments, based on the determination of the defect information of the IO controllers and IO physical layers in the multiple groups of IO modules, when determining the recovery type of the die where the multiple groups of IO modules are located according to the failure conditions of the multiple groups of IO modules and recording the IO numbers of the failed IO modules, it may specifically be to determine the recovery type of the die where the multiple groups of IO modules are located according to the defect information of the IO controllers and IO physical layers in the multiple groups of IO modules, and record the IO numbers of the IO modules where the failed IO controllers and IO physical layers are located in the defect information.

[0093] In some embodiments, to implement the test of the IO controller Figure 6 Exemplarily shows the internal optional structural schematic diagram of the IO controller in the embodiments of the present application.

[0094] As shown Figure 6 in the figure, the IO controller may include a first management configuration unit and a data processing unit, and a first scan chain is provided corresponding to the first management configuration unit and the data processing unit.

[0095] The first management configuration unit is responsible for the initialization settings, configuration management, and connection status of the IO controller. The first management configuration unit obtains the configuration information of the IO controller, such as the baud rate, protocol type, etc., through a bus or other dedicated interfaces, and performs configuration management on the IO controller to ensure the correct interaction between the IO controller and other devices.

[0096] The data processing unit is used to receive and send data. When data needs to be transmitted, the data processing unit will encapsulate, de-encapsulate, construct, and split data packets according to a preset protocol stack. For an IO controller with input and output functions, when performing an input operation, the data processing unit can receive data from an external device; when performing an output operation, the data processing unit can package the data and send it to a specified target.

[0097] The first scan chain is used to perform a scan test on the execution logic of the first management configuration unit and the data processing unit, so as to verify the basic functions of the first management configuration unit and the data processing unit, and ensure that the first management configuration module and the data processing unit can work properly according to the expected design. For example: the normal interaction between the first management configuration unit and the data processing unit, and the correct response to input instructions and processing of data.

[0098] Moreover, to facilitate accurately determining defect information in the IO controller, the IO controller in the embodiment of the present application may have a more refined structure. Referring to Figure 6 the figure, the first management configuration unit of the IO controller in the embodiment of the present application may include a first core domain and a first configuration domain, and a second scan chain is provided corresponding to the first core domain.

[0099] The first core domain focuses on the core functions and basic hardware control of the IO controller, and is used to configure the core functions of the IO controller. The core functions may include, for example, at least one of a clock signal, a reset logic, and a low-power state of the physical layer.

[0100] The first configuration domain focuses on the IO configuration requirements and is used to configure the communication parameters of the IO controller. The communication parameters are parameters for controlling the communication process between the IO controller and other devices, such as the occupied transmission channels, interrupt lines, etc.

[0101] The second scan chain is used to perform a scan test on the execution logic of the first core domain to implement a test on the manufacturing correctness of the execution logic of the first core domain.

[0102] Based on the first scan chain and the second scan chain set by the IO controller, when testing the IO controllers of multiple groups of IO modules, it can be: for the IO controllers of the multiple groups of IO modules, respectively based on the first scan chain, scanning and testing the execution logics of the first management configuration unit and the data processing unit; and, based on the second scan chain, scanning and testing the execution logic of the first core domain.

[0103] In some embodiments, to implement the test of the IO physical layer, Figure 7 An exemplary internal optional structure diagram of the IO physical layer in the embodiments of the present application is shown.

[0104] As Figure 7 shown, the IO physical layer may include a second management configuration unit and a physical layer unit, and a third scan chain is provided corresponding to the second management configuration unit and the physical layer unit.

[0105] The second management configuration unit is responsible for the initialization setting, configuration management, and connection state of the physical layer. The second management configuration unit obtains the configuration information of the physical layer, such as the baud rate, protocol type, etc., through a bus or other dedicated interfaces, and performs configuration management on the physical layer to ensure the correct information interaction between the physical layer and other devices.

[0106] The physical layer unit is used to complete the electrical transmission and signal processing of data. For example: since different interface standards may use different voltage levels, the physical layer unit will perform necessary voltage conversion to ensure the stable transmission of data.

[0107] The third scan chain is used to scan and test the execution logics of the second management configuration unit and the physical layer unit, so as to realize the basic function verification of the second management configuration unit and the physical layer unit, and ensure that the second management configuration module and the physical layer unit can work properly according to the expected design. For example: the normal interaction between the second management configuration unit and the physical layer unit, and the correct response to input instructions and processing of data.

[0108] Moreover, to facilitate the accurate determination of defect information in the IO physical layer, the IO physical layer in the embodiments of the present application may have a more refined structure. Referring to Figure 7 shown, the second management configuration unit of the IO physical layer in the embodiments of the present application may include a second core domain and a second configuration domain, and a fourth scan chain is provided corresponding to the second core domain.

[0109] The second core domain focuses on the core functions and basic signal control of the IO physical layer, and is used to configure the core functions of the IO physical layer. The core functions may include, for example, at least one of a clock signal, a reset logic, and a low-power state of the physical layer.

[0110] The second configuration domain focuses on the configuration requirements of the IO physical layer and is used to configure the component information of the IO physical layer. The component information is the configuration information for the IO physical layer to manage and control input / output devices, such as the setting information of hardware components like storage and management device controllers, channels, and ports.

[0111] The fourth scan chain is used to perform a scan test on the execution logic of the second core domain to achieve the test of the manufacturing correctness of the execution logic of the second core domain.

[0112] Based on the third scan chain and the fourth scan chain set for the IO physical layer, when testing the IO physical layers of multiple groups of IO modules, it can be: for the IO physical layers of the multiple groups of IO modules, respectively based on the third scan chain, scan and test the execution logic of the second management configuration unit and the physical layer unit; and, based on the fourth scan chain, scan and test the execution logic of the second core domain.

[0113] It should be noted that the scan chains in the embodiments of the present application (such as the first scan chain, the second scan chain, the third scan chain, and the fourth scan chain) can be inserted after the functional design of the IO controller or the IO physical layer is completed. Thus, ordinary registers in the IO controller or the IO physical layer can be replaced with scan flip-flops, and corresponding tests are performed based on the scan chains to effectively screen out defective chips and improve product quality. Moreover, since the first core domain of the first management configuration unit and the second core domain of the second management configuration unit focus on the management and configuration of the core functions of the device, it is necessary to ensure the normal operation of the core domain. If there are defects in the first core domain of the first management configuration unit of an IO controller and / or there are defects in the second core domain of the second management configuration unit of an IO physical layer, then the die where the IO module to which the IO controller and / or the IO physical layer belongs will not be recyclable, that is: the first core domain of the first management configuration unit of the IO controllers of multiple IO modules of the recyclable die and the second core domain of the second management configuration unit of the IO physical layer must both be valid.

[0114] In some embodiments, based on the die including multiple groups of IO modules, each IO module including an IO controller and an IO physical layer, and the internal refinement structures of the IO controller and the IO physical layer, when testing the IO controllers and IO physical layers of multiple groups of IO modules respectively, the defect information determined for the IO controllers and IO physical layers in the multiple groups of IO modules can include various situations, as follows.

[0115] Case 1: The IO controller and / or the IO physical layer of one or more groups of IO modules in multiple groups of IO modules fails, and the first core domains of the first management configuration units of the IO controllers of the multiple groups of IO modules are all valid, and the second core domains of the second management configuration units of the IO physical layers are all valid;

[0116] Case 2: The first core domain of the first management configuration unit of the IO controller of at least one group of IO modules in multiple groups of IO modules fails, and / or, the second core domain of the second management configuration unit of the IO physical layer fails.

[0117] It can be understood that since the first core domain of the first management configuration unit of the IO controller and the second core domain of the second management configuration unit of the IO physical layer focus on the management and configuration of the core functions of the device, therefore, when the IO controller and / or the IO physical layer of one or more groups of IO modules in multiple groups of IO modules fails, and the first core domains of the first management configuration units of the IO controllers of the multiple groups of IO modules are all valid, and the second core domains of the second management configuration units of the IO physical layers are all valid, it can be determined that although there are defects in the die where the multiple groups of IO modules are located, it can still be used continuously. When the first core domain of the first management configuration unit of the IO controller of at least one group of IO modules in multiple groups of IO modules fails, and / or, the second core domain of the second management configuration unit of the IO physical layer fails, it can be determined that there are irreversible defects in the die where the multiple groups of IO modules are located, and it is an unreliable die.

[0118] In addition, when the IO controller and the IO physical layer of one or more groups of IO modules in multiple groups of IO modules do not fail, and the first core domains of the first management configuration units of the IO controllers of the multiple groups of IO modules are all valid, and the second core domains of the second management configuration units of the IO physical layers are all valid, it can be determined that there are no defects in the die where the multiple groups of IO modules are located, and it is a reliable die.

[0119] In some embodiments, corresponding to the defect information content of the IO controllers and IO physical layers of multiple groups of IO modules, the recycling types of the die can include recyclable and non-recyclable.

[0120] Thus, when determining the recycling type of the die where the multiple groups of IO modules are located according to the defect information of the IO controllers and IO physical layers in the multiple groups of IO modules, if the defect information is Case 1, that is, the IO controller and / or the IO physical layer of one or more groups of IO modules in multiple groups of IO modules fails, and the first core domains of the first management configuration units of the IO controllers of the multiple groups of IO modules are all valid, and the second core domains of the second management configuration units of the IO physical layers are all valid, then it can be determined that the recycling type of the die where the multiple groups of IO modules are located is recyclable.

[0121] If the information on the failed area is Case 2, that is, the first core domain of the first management and configuration unit of the IO controller in at least one group of IO modules fails, and / or the second core domain of the second management and configuration unit of the IO physical layer fails, it can be determined that the recycling type of the die where the multiple groups of IO modules are located is non-recyclable.

[0122] In an optional example, if the recycling type of the die where the multiple groups of IO modules are located is recyclable, indicating that the die can be recycled and reused without affecting the chip function, steps S130 and S140 can be executed.

[0123] In an optional example, if the recycling type of the die where the multiple groups of IO modules are located is non-recyclable, the die can be discarded.

[0124] It should be noted that to manufacture a chip with the expected function, after determining that a die is non-recyclable and discarding it, a new die of the same type can be replaced, and steps S120, S130, and S140 are executed in sequence until a chip with the expected function is obtained.

[0125] In some embodiments, to improve the reliability of the chip, in the preset chip, for the multiple groups of IO modules corresponding to the die, one or more defective IO numbers are preset. Among them, if there are multiple preset defective IO numbers, the multiple IO modules corresponding to the multiple defective IO numbers are symmetrically distributed in the die.

[0126] Moreover, since when multiple IO modules are symmetrically distributed in the die, the probability of defects occurring in the multiple IO modules in the die is the same or similar. To average the defect risk borne by the die as a whole, isolate the potential failure impact range, and increase the die's ability to resist random defects, the die packaging strategy can include translational packaging and rotational packaging.

[0127] Thus, when determining the packaging strategy of the die corresponding to the recycling type according to the recorded IO number, as an optional implementation, if the recorded IO number is the same as the defective IO number, it can be determined that the packaging strategy of the die corresponding to the recycling type is translational packaging, that is, the die is translated to a predetermined position for packaging. As another optional implementation, if the recorded IO number corresponds to the defective IO number, for example: the recorded IO number and the defective IO number are arranged symmetrically about the axis of symmetry of the die, it can be determined that the packaging strategy of the die corresponding to the recycling type is rotational packaging, that is, the die is rotated to a predetermined position for packaging, such as rotating 180°.

[0128] It should be noted that for the same type of core particles, the multiple IO modules contained therein and the corresponding IO number sequence are substantially the same. If the recorded IO number is the same as the defective IO number, but the recorded IO number is located in a different core particle of the same type, the core particle where the recorded IO number is located can be translated and packaged to the packaging position corresponding to the core particle where the defective IO number is located, so as to meet the requirement that the defective IO number indicates that the IO module corresponding to the position is a defective IO module.

[0129] In some embodiments, to avoid normal logic execution in the chiplet or increase the power consumption of the chiplet, Figure 8 Another optional flow chart of the bad chip recycling method in the embodiment of the present application is shown as an example. Figure 8 As shown, after step S140, step S150 may be further included to initialize the chip recovery and check the position mark of the chip.

[0130] The core position identifier may be used to mark the usage status of multiple IO modules of the core.

[0131] By initializing the recycling chip and checking the position identification of the core particle, the use status of each IO module in the core particle can be confirmed to ensure the correct hardware configuration and resource allocation optimization of the core particle. When an unused idle IO module is found, the IO module can be placed in a low power state to reduce the power consumption of the core particle, and the clock signal of the IO module can be turned off, and the reset signal can be placed in a valid state to perform normal startup and configuration operations.

[0132] In order to facilitate understanding of the bad chip recycling method of the embodiment of the present application, Figure 5 The CPU core in the CPU chip 1 shown is used as an example for explanation. CPU core n1 and CPU core n2 are CPU cores with the same functions. Both CPU core n1 and CPU core n2 contain 8 groups of DDR modules, and each DDR module contains a DDR controller and a DDR physical layer. Each DDR controller is provided with a first management configuration unit and a data processing unit, and a first scan chain is provided corresponding to the first management configuration unit and the data processing unit; the first management configuration unit is provided with a first configuration domain and a first core domain, and a second scan chain is provided corresponding to the first core domain. Each DDR physical layer is provided with a second management configuration unit and a physical layer unit, and a third scan chain is provided corresponding to the second management configuration unit and the physical layer unit; the second management configuration unit is provided with a second configuration domain and a second core domain, and a fourth scan chain is provided corresponding to the second core domain. Corresponding Figure 5 The CPU chip 1 shown, Figure 9 and Figure 10The schematic diagram of the packaging of the CPU chip 1 is exemplarily shown. In the CPU chip 1, the packaging position of the CPU chip n1 is L1, and the packaging position of the CPU chip n2 is L2.

[0133] First, the eight groups of DDR modules (i.e., DDR00 to DDR07) of the CPU chip n1 are tested. Specifically, the first management configuration unit and the execution logic of the data processing unit of the DDR controller can be scanned and tested based on the first scan chain, and the execution logic of the first core domain can be scanned and tested based on the second scan chain. In addition, the second management configuration unit and the execution logic of the physical layer unit of the DDR physical layer can be scanned and tested based on the third scan chain, and the execution logic of the second core domain can be scanned and tested based on the fourth scan chain.

[0134] If it is tested that the DDR controller and DDR physical layer in DDR00 to DDR07 are valid, and the first core domain managed by the first management configuration of the DDR controller is valid, and the second core domain managed by the second management configuration of the DDR physical layer is valid, then it can be determined that the CPU chip n1 has no defects and can operate normally and effectively, and the CPU chip n1 can be packaged in the L1 position according to the preset position.

[0135] If it is found through testing that DDR00 or DDR05 in DDR00 to DDR07 fails, and the first core domain of the first management configuration management of the DDR controller in DDR00 to DDR07 is valid, and the second core domain of the second management configuration management of the DDR physical layer in DDR00 to DDR07 is valid, then it can be determined that DDR00 or DDR05 of CPU chip n1 fails, and thus it is determined that CPU chip n1 has a defect. However, since the first core domain of the first management configuration management of each DDR controller is valid, and the second core domain of the second management configuration management of each DDR physical layer is valid, the defect of DDR00 or DDR05 does not affect the normal operation of CPU chip n1. Therefore, the recycling status of CPU chip n1 can be marked as recyclable, and the number of the failed DDR module (i.e., 00 or 05) can be recorded. Among them, if the defective IO numbers in the CPU chip are pre-set to 0 and 5, then according to the recorded 00 or 05, it can be determined that the packaging strategy of CPU chip n1 where DDR00 to DDR07 is located is translation packaging. Thus, CPU chip n1 can be translationally packaged at the L1 position, such as Figure 9 shown.

[0136] Similarly, if it is tested that DDR02’ or DDR07’ of CPU die n2 fails, and it is preset that the defective IO numbers in the CPU die are 2 and 7, then according to the recorded 02’ or 07’, the packaging strategy of the CPU die n2 where DDR02’ or DDR07’ is located can be determined as translational packaging. Thus, the CPU die n2 can be translationally packaged at the L2 position.

[0137] If it is tested that DDR03 or DDR06 in DDR00 to DDR07 fails, and the first core domains managed by the first management configuration of the DDR controllers in DDR00 to DDR07 are all valid, and the second core domains managed by the second management configuration of the DDR physical layers in DDR00 to DDR07 are all valid, then it can be determined that DDR03 or DDR06 of CPU die n1 fails, thereby determining that there is a defect in CPU die n1. However, since the first core domains managed by the first management configuration of each DDR controller are all valid, and the second core domains managed by the second management configuration of each DDR physical layer are all valid, the defect of DDR03 or DDR06 does not affect the normal operation of CPU die n1. Therefore, the recycling status of CPU die n1 can be marked as recyclable, and the numbers of the failed DDR modules (i.e., 03 or 06) are recorded. Among them, if it is preset that the defective IO numbers in the CPU die are 3 and 6, then according to the recorded 03 or 06, the packaging strategy of the CPU die n1 where DDR03 to DDR06 are located can be determined as rotational packaging. Thus, the CPU die n1 can be rotated 180° and packaged at the L1 position, as Figure 10 shown.

[0138] Similarly, if it is tested that DDR01’ or DDR04’ of CPU die n2 fails, and it is preset that the defective IO numbers in the CPU die are 1 and 4, then according to the recorded 01’ or 04’, the packaging strategy of the CPU die n2 where DDR01’ or DDR04’ is located can be determined as rotational packaging. Thus, the CPU die n2 can be rotated 180° and packaged at the L2 position.

[0139] If it is tested that the first core domain of the first configuration management unit of the DDR controller in DDR00 of CPU die n1 fails, and / or the second core domain of the second configuration management unit of the DDR physical layer fails, then the recycling type of CPU die n1 can be marked as non-recyclable. For the non-recyclable CPU die n1, it can be discarded and replaced with a new CPU die of the same type, such as CPU die n13, and steps such as testing are performed on CPU die n3.

[0140] It should be noted that Figure 9 or Figure 10Taking the CPU die as an example, the defective die recovery method is described below. For the defective die recovery method implemented based on the I / O die, the above content can be referred to by analogy, and the embodiments of this application will not elaborate on this too much.

[0141] By packaging the recoverable CPU die n1 at the L1 position, the recovered CPU chip 1 with its function unaffected can be obtained. Then, initialize this CPU chip 1, check the position identifier of the CPU die n1. If DDR00 in the CPU die n1 is idle, put DDR00 in the CPU die n1 into the low-power state, turn off its clock signal, and put the reset signal in the effective state to avoid normal logic execution in DDR00 or increase the power consumption of the CPU die n1.

[0142] It can be seen that in the embodiments of this application, by obtaining a preset chip including multiple dies, when the die includes multiple groups of I / O modules symmetrically distributed, the multiple groups of I / O modules can be tested to determine the failure conditions of the multiple groups of I / O modules. Then, according to the failure conditions, the recovery type (i.e., the recovery strategy) of the die can be determined, improving the recovery utilization rate of the die. And, when the I / O module is set with an I / O number and different dies are set with different packaging positions in the preset chip, it is possible to record the I / O numbers of the failed I / O modules according to the failure conditions of the multiple groups of I / O modules, so that based on the recorded I / O numbers, the packaging strategy of the die corresponding to the recovery type can be determined. Thus, according to the packaging position and the packaging strategy, the die corresponding to the recovery type can be packaged to obtain a recovered chip, enabling the obtained recovered chip with its function unaffected while recovering the defective die, reducing the cost of the chip.

[0143] Next, the device for defective die recovery provided by the embodiments of this application is introduced. The device content described below can be considered as a chip recovery device or a computing device, which is a functional module required to implement the defective die recovery method provided by the embodiments of this application. The device content described below can be correspondingly referred to the method content described above.

[0144] Figure 11 The optional block diagram of the defective die recovery device provided by the embodiments of this application is shown. As Figure 11 shown, the device may include: a chip acquisition unit 111, a failure test unit 112, a recovery determination unit 113, a strategy determination unit 114, and a defective die packaging unit 115.

[0145] The chip acquisition unit 111 is configured to acquire a preset chip, where the preset chip includes a plurality of dielets, each dielet contains multiple groups of IO modules symmetrically distributed, and the IO modules are respectively provided with IO numbers for indicating the placement positions of the IO modules in the dielet; wherein, different dielets are arranged at different packaging positions in the preset chip;

[0146] The failure test unit 112 is configured to test the multiple groups of IO modules to determine the failure conditions of the multiple groups of IO modules;

[0147] The recycling determination unit 113 is configured to determine the recycling type of the dielet where the multiple groups of IO modules are located according to the failure conditions, and record the IO numbers of the failed IO modules;

[0148] The strategy determination unit 114 is configured to determine the packaging strategy of the dielet corresponding to the recycling type according to the recorded IO numbers;

[0149] The defective chip packaging unit 115 is configured to package the dielet corresponding to the recycling type according to the packaging position and the packaging strategy to obtain a recycled chip.

[0150] Optionally, the IO module includes an IO controller and an IO physical layer;

[0151] The failure test unit 112 tests the multiple groups of IO modules to determine the failure conditions of the multiple groups of IO modules, specifically: testing the IO controllers and IO physical layers of the multiple groups of IO modules respectively to determine the defect information of the IO controllers and IO physical layers in the multiple groups of IO modules.

[0152] Optionally, the recycling determination unit 113 determines the recycling type of the dielet where the multiple groups of IO modules are located according to the failure conditions of the multiple groups of IO modules, and records the IO numbers of the failed IO modules, specifically: determining the recycling type of the dielet where the multiple groups of IO modules are located according to the defect information of the IO controllers and IO physical layers in the multiple groups of IO modules, and recording the IO numbers of the IO modules where the failed IO controllers and IO physical layers are located in the defect information.

[0153] Optionally,

[0154] The IO controller includes a first management and configuration unit and a data processing unit, and a first scan chain is provided corresponding to the first management and configuration unit and the data processing unit. The first scan chain is used to perform a scan test on the execution logics of the first management and configuration unit and the data processing unit. The first management and configuration unit includes a first configuration domain and a first core domain. The first core domain is used to configure the core functions of the IO controller, and a second scan chain is provided corresponding to the first core domain. The second scan chain is used to perform a scan test on the execution logic of the first core domain. The first configuration domain is used to configure the communication parameters of the IO controller;

[0155] The failure test unit 112 tests the IO controllers of the multiple groups of IO modules, including:

[0156] For the IO controllers of the multiple groups of IO modules, respectively based on the first scan chain, scan and test the execution logics of the first management and configuration unit and the data processing unit; and, based on the second scan chain, scan and test the execution logic of the first core domain.

[0157] Optionally,

[0158] The IO physical layer includes a second management and configuration unit and a physical layer unit, and a third scan chain is provided corresponding to the second management and configuration unit and the physical layer unit. The third scan chain is used to perform a scan test on the execution logics of the second management and configuration unit and the physical layer unit. The second management and configuration unit includes a second configuration domain and a second core domain. The second core domain is used to configure the core functions of the IO physical layer, and a fourth scan chain is provided corresponding to the second core domain. The fourth scan chain is used to perform a scan test on the execution logic of the second core domain. The second configuration domain is used to configure the component information of the IO physical layer;

[0159] The failure test unit 112 tests the IO physical layers of the multiple groups of IO modules, including:

[0160] For the IO physical layers of the multiple groups of IO modules, respectively based on the third scan chain, scan and test the execution logics of the second management and configuration unit and the physical layer unit; and, based on the fourth scan chain, scan and test the execution logic of the second core domain.

[0161] Optionally, the defect information includes:

[0162] The IO controllers of one or more groups of IO modules among the multiple groups of IO modules fail and / or the IO physical layers fail, and the first core domains of the first management and configuration units of the IO controllers of the multiple groups of IO modules are all valid, and the second core domains of the second management and configuration units of the IO physical layers are all valid;

[0163] Or,

[0164] The first core domain of the first management and configuration unit of the IO controller of at least one group of IO modules among the multiple groups of IO modules fails, and / or, the second core domain of the second management and configuration unit of the IO physical layer fails.

[0165] Optionally, the recycling type includes recyclable and non-recyclable;

[0166] The recycling determination unit 113 determines the recycling type of the die where the multiple groups of IO modules are located according to the defect information of the IO controllers and the IO physical layers in the multiple groups of IO modules, including:

[0167] If the defect information is that one or more groups of IO modules among the multiple groups of IO modules have failed IO controllers and / or failed IO physical layers, and the first core domains of the first management and configuration units of the IO controllers of the multiple groups of IO modules are all valid, and the second core domains of the second management and configuration units of the IO physical layers are all valid, determine that the recycling type of the die where the multiple groups of IO modules are located is recyclable;

[0168] If the defect information is that the first core domain of the first management and configuration unit of the IO controller of at least one group of IO modules among the multiple groups of IO modules fails, and / or, the second core domain of the second management and configuration unit of the IO physical layer fails, determine that the recycling type of the die where the multiple groups of IO modules are located is non-recyclable.

[0169] Optionally,

[0170] If the recycling type of the die where the multiple groups of IO modules are located is recyclable, then the policy determination unit 114 executes to determine the packaging policy of the die of the corresponding recycling type according to the recorded IO numbers, and, the defective die packaging unit 115 executes to package the die of the corresponding recycling type according to the packaging position and the packaging policy to obtain a recycled chip;

[0171] If the recycling type of the die where the multiple groups of IO modules are located is non-recyclable, then the recycling determination unit 113 discards the die.

[0172] Optionally, in the preset chip, one or more defective IO numbers are preset for the multiple groups of IO modules corresponding to the die; among them, the multiple IO modules corresponding to the multiple defective IO numbers are symmetrically distributed in the die;

[0173] The packaging policy includes translational packaging and rotational packaging;

[0174] The defective die packaging unit 115 determines the packaging policy of the die of the corresponding recycling type according to the recorded IO numbers, including:

[0175] If the recorded IO number is the same as the defective IO number, determine that the packaging strategy for the die of the corresponding recycling type is translational packaging;

[0176] If the recorded IO number corresponds to the defective IO number, determine that the packaging strategy for the die of the corresponding recycling type is rotational packaging.

[0177] Optionally, the defective die recycling device further includes:

[0178] An initialization unit 116, configured to initialize the recycled chip and check the position identifier of the die, where the position identifier is used to mark the usage status of multiple IO modules of the die.

[0179] It can be seen that the defective die recycling device according to the embodiment of the present application can determine the recycling strategy (i.e., the recycling type) of the die based on the defective condition of the die, improving the recycling utilization rate of the die. Moreover, while recycling the defective die, a recycled chip with unaffected functions is obtained, reducing the cost of the chip.

[0180] The embodiment of the present application further provides a chip, which is obtained by recycling based on the defective die recycling method provided by the embodiment of the present application.

[0181] Optionally, the chip includes multiple dice, the die includes multiple groups of symmetrically distributed IO modules, and the IO module includes an IO controller and an IO physical layer.

[0182] Optionally, the IO controller includes a first management and configuration unit and a data processing unit, and a first scan chain is provided corresponding to the first management and configuration unit and the data processing unit, and the first scan chain is used to perform a scan test on the execution logic of the first management and configuration unit and the data processing unit; the first management and configuration unit includes a first configuration domain and a first core domain, the first core domain is used to configure the core function of the IO controller, and a second scan chain is provided corresponding to the first core domain, and the second scan chain is used to perform a scan test on the execution logic of the first core domain; the first configuration domain is used to configure the communication parameters of the IO controller.

[0183] Optionally, the IO physical layer includes a second management and configuration unit and a physical layer unit, and a third scan chain is provided corresponding to the second management and configuration unit and the physical layer unit. The third scan chain is used to perform a scan test on the execution logic of the second management and configuration unit and the physical layer unit. The second management and configuration unit includes a second configuration domain and a second core domain. The second core domain is used to configure the core functions of the IO physical layer, and a fourth scan chain is provided corresponding to the second core domain. The fourth scan chain is used to perform a scan test on the execution logic of the second core domain. The second configuration domain is used to configure the component information of the IO physical layer.

[0184] The embodiments of the present application further provide a computing device, such as a server device or a terminal device. The computing device may include a defective chip recovery device as provided in the embodiments of the present application.

[0185] The embodiments of the present application further provide a storage medium. The storage medium stores one or more computer-executable instructions. When the one or more computer-executable instructions are executed, the defective chip recovery method provided in the embodiments of the present application is implemented.

[0186] The embodiments of the present application further provide a computer program product. The computer program product includes one or more computer-executable instructions. When the one or more computer-executable instructions are executed, the defective chip recovery method provided in the embodiments of the present application is implemented.

[0187] The above describes multiple embodiment solutions provided in the embodiments of the present application. The optional methods described in each embodiment solution can be combined and cross-referenced with each other without conflict, thereby extending multiple possible embodiment solutions. All of these can be considered as the embodiment solutions disclosed and made public in the embodiments of the present application.

[0188] Although the embodiments of the present application are disclosed as above, the present application is not limited thereto. Any person skilled in the art can make various changes and modifications without departing from the spirit and scope of the present application. Therefore, the protection scope of the present application should be subject to the scope defined by the claims.

Claims

1. A bad chip recycling method, characterized in that: include: Obtain a preset chip, wherein the preset chip includes a plurality of core particles, wherein the core particles include a plurality of symmetrically distributed groups of IO modules, and the IO modules are correspondingly provided with IO numbers for indicating the placement positions of the IO modules in the core particles; wherein different core particles are provided with different packaging positions in the preset chip; Testing the multiple groups of IO modules to determine failure conditions of the multiple groups of IO modules; According to the failure situation, determining the recycling type of the core particles where the multiple groups of IO modules are located, and recording the IO numbers of the failed IO modules; Determine a packaging strategy for a chip of a corresponding recycling type according to the recorded IO number; According to the packaging position and the packaging strategy, the core particles of the corresponding recycling type are packaged to obtain recycled chips.

2. The method according to claim 1, characterized in that The IO module includes an IO controller and an IO physical layer; The testing of the multiple groups of IO modules to determine the failure conditions of the multiple groups of IO modules is specifically: testing the IO controllers and IO physical layers of the multiple groups of IO modules respectively to determine the defect information of the IO controllers and IO physical layers in the multiple groups of IO modules.

3. The method according to claim 2, characterized in that The method of determining the recovery type of the core particles where the multiple groups of IO modules are located according to the failure conditions of the multiple groups of IO modules, and recording the IO numbers of the failed IO modules, is specifically: determining the recovery type of the core particles where the multiple groups of IO modules are located according to the defect information of the IO controllers and IO physical layers in the multiple groups of IO modules, and recording the IO numbers of the IO modules where the failed IO controllers and IO physical layers in the defect information are located.

4. The method according to claim 3, characterized in that The IO controller includes a first management configuration unit and a data processing unit, and a first scan chain is provided corresponding to the first management configuration unit and the data processing unit, and the first scan chain is used to scan and test the execution logic of the first management configuration unit and the data processing unit; the first management configuration unit includes a first configuration domain and a first core domain, the first core domain is used to configure the core function of the IO controller, and a second scan chain is provided corresponding to the first core domain, and the second scan chain is used to scan and test the execution logic of the first core domain; The first configuration domain is used to configure communication parameters of the IO controller; Testing the IO controllers of the multiple groups of IO modules includes: For the IO controllers of the multiple groups of IO modules, respectively, based on the first scan chain, scan and test the execution logic of the first management configuration unit and the data processing unit; And, scanning and testing the execution logic of the first core domain based on the second scan chain.

5. The method according to claim 3, characterized in that: The IO physical layer includes a second management configuration unit and a physical layer unit, and a third scan chain is provided corresponding to the second management configuration unit and the physical layer unit, and the third scan chain is used to scan and test the execution logic of the second management configuration unit and the physical layer unit; the second management configuration unit includes a second configuration domain and a second core domain, the second core domain is used to configure the core function of the IO physical layer, and a fourth scan chain is provided corresponding to the second core domain, and the fourth scan chain is used to scan and test the execution logic of the second core domain; The second configuration field is used to configure component information of the IO physical layer; Testing the IO physical layers of the multiple groups of IO modules includes: For the IO physical layers of the multiple groups of IO modules, respectively, based on the third scan chain, scan and test the execution logic of the second management configuration unit and the physical layer unit; And, scanning and testing the execution logic of the second core domain based on the fourth scan chain.

6. The method according to any one of claims 4 or 5, characterized in that: The defect information includes: The IO controllers of one or more groups of IO modules in the multiple groups of IO modules fail and / or the IO physical layers fail, and the first core domains of the first management configuration units of the IO controllers of the multiple groups of IO modules are all valid, and the second core domains of the second management configuration units of the IO physical layers are all valid; or, The first core domain of the first management configuration unit of the IO controller of at least one group of IO modules among the multiple groups of IO modules fails, and / or the second core domain of the second management configuration unit of the IO physical layer fails.

7. The method according to claim 6, characterized in that The recycling types include recyclable and non-recyclable; The step of determining the recycling type of the core particles where the multiple groups of IO modules are located according to the defect information of the IO controllers and the IO physical layers in the multiple groups of IO modules includes: If the defect information indicates that the IO controller of one or more groups of IO modules in the multiple groups of IO modules fails and / or the IO physical layer fails, and the first core domains of the first management configuration units of the IO controllers of the multiple groups of IO modules are all valid, and the second core domains of the second management configuration units of the IO physical layer are all valid, it is determined that the recycling type of the core particles where the multiple groups of IO modules are located is recyclable; If the defect information is that the first core domain of the first management configuration unit of the IO controller of at least one group of IO modules among the multiple groups of IO modules has failed, and / or the second core domain of the second management configuration unit of the IO physical layer has failed, it is determined that the recycling type of the core particles where the multiple groups of IO modules are located is non-recyclable.

8. The method according to claim 7, characterized in that If the recycling type of the core particles where the multiple groups of IO modules are located is recyclable, then determining the packaging strategy of the core particles of the corresponding recycling type according to the recorded IO numbers, and packaging the core particles of the corresponding recycling type according to the packaging positions and the packaging strategy to obtain recycled chips; If the recycling type of the core particles where the multiple groups of IO modules are located is non-recyclable, the core particles are discarded.

9. The method according to claim 1, characterized in that: In the preset chip, multiple groups of IO modules corresponding to the core particle are pre-set with one or more defective IO numbers; wherein the multiple IO modules corresponding to the multiple defective IO numbers are symmetrically distributed in the core particle; The packaging strategy includes translation packaging and rotation packaging; The step of determining a packaging strategy for a chip of a corresponding recycling type according to the recorded IO number includes: If the recorded IO number is the same as the defective IO number, determining that the packaging strategy of the chip of the corresponding recycling type is translation packaging; If the recorded IO number corresponds to the defective IO number, it is determined that the packaging strategy of the chip of the corresponding recycling type is rotation packaging.

10. The method according to claim 1, characterized in that After packaging the corresponding recycling type of core particles according to the packaging position and the packaging strategy to obtain the recycled chips, the method further includes: Initialize the recycling chip and check the position mark of the chip, where the position mark is used to mark the use status of multiple IO modules of the chip.

11. A bad film recovery device, characterized in that: include: A chip acquisition unit, used for acquiring a preset chip, wherein the preset chip includes a plurality of core particles, wherein the core particles include a plurality of symmetrically distributed groups of IO modules, wherein the IO modules are correspondingly provided with IO numbers for indicating the placement positions of the IO modules in the core particles; wherein different core particles are provided with different packaging positions in the preset chip; A failure testing unit, used to test the multiple groups of IO modules to determine the failure conditions of the multiple groups of IO modules; A recycling determination unit, used to determine the recycling type of the core particles where the multiple groups of IO modules are located according to the failure situation, and record the IO number of the failed IO module; A strategy determination unit, used to determine a packaging strategy for a chip of a corresponding recycling type according to the recorded IO number; The bad chip packaging unit is used to package the core particles of the corresponding recycling type according to the packaging position and the packaging strategy to obtain recycled chips.

12. The bad film recovery device according to claim 11, characterized in that: Also includes: The initialization unit is used to initialize the recycling chip and check the position mark of the chip, where the position mark is used to mark the use status of multiple IO modules of the chip.

13. A chip, characterized in that: The chip is recovered based on the bad chip recovery method according to any one of claims 1 to 10.

14. The chip according to claim 13, characterized in that: The chip includes a plurality of core particles, each of which includes a plurality of symmetrically distributed IO modules, and each of the IO modules includes an IO controller and an IO physical layer.

15. The chip according to claim 14, characterized in that: The IO controller includes a first management configuration unit and a data processing unit, and a first scan chain is set corresponding to the first management configuration unit and the data processing unit, and the first scan chain is used to scan and test the execution logic of the first management configuration unit and the data processing unit; the first management configuration unit includes a first configuration domain and a first core domain, the first core domain is used to configure the core function of the IO controller, and a second scan chain is set corresponding to the first core domain, and the second scan chain is used to scan and test the execution logic of the first core domain; the first configuration domain is used to configure the communication parameters of the IO controller.

16. The chip according to claim 14, characterized in that: The IO physical layer includes a second management configuration unit and a physical layer unit, and a third scan chain is set corresponding to the second management configuration unit and the physical layer unit, and the third scan chain is used to scan and test the execution logic of the second management configuration unit and the physical layer unit; the second management configuration unit includes a second configuration domain and a second core domain, the second core domain is used to configure the core function of the IO physical layer, and a fourth scan chain is set corresponding to the second core domain, and the fourth scan chain is used to scan and test the execution logic of the second core domain; the second configuration domain is used to configure component information of the IO physical layer.

17. A computing device, characterized in that It comprises a bad film recovery device as described in any one of claims 11 to 12.

18. A storage medium, characterized in that: The storage medium stores one or more computer executable instructions, and when the one or more computer executable instructions are executed, the bad chip recycling method according to any one of claims 1 to 10 is implemented.

19. A computer program product, characterized in that The method comprises one or more computer executable instructions, and when the one or more computer executable instructions are executed, the bad chip recycling method as claimed in any one of claims 1 to 10 is implemented.