Data processing card screening method and device, equipment and medium

Abstract

The invention discloses a data processing card screening method and device, equipment and a medium, and relates to the technical field of computers. The method is applied to a baseboard management controller, and comprises the following steps: adding a screening program for screening data processing cards in a test program of an equipment production line; the equipment production line is used for producing equipment for providing artificial intelligence service; performing scanning judgment on the to-be-scanned address of each slot on the data processing card by using each preset judgment rule in the screening program in sequence to obtain a corresponding scanning judgment result; the to-be-scanned address is an address used for communication on a communication bus realized based on a preset serial communication protocol; and screening bad data processing cards in the data processing cards according to the scanning judgment result, and executing a return authorization process on the bad data processing cards. According to the technical scheme, the bad data processing cards can be eliminated in advance, and abnormal termination of the test program is not caused any more.

Description

Technical Field

[0001] The present invention relates to the field of computer technology, and in particular to a data processing card screening method, device, equipment and medium. Background Art

[0002] BlueField-3 (BF3) is the first DPU (Data Processing Unit) designed for AI (Artificial Intelligence) and accelerated computing, helping enterprises achieve industry-leading performance and data center security for applications of any scale. This DPU is optimized for multi-tenant, cloud-native environments, providing data center-level software-defined and hardware-accelerated networking, storage, security and management services.

[0003] Currently, in the production of AI servers, manual intervention is required only when a BF3 card failure occurs, which reduces the production efficiency of the production line diag (test) automation and wastes labor costs. Therefore, how to provide a solution to the above technical problems is a problem that technicians in this field need to solve. Summary of the invention

[0004] In view of this, the purpose of the present invention is to provide a data processing card screening method, device, equipment and medium, which can eliminate bad BF3 cards in advance, ensure that no bad cards are exposed in the production process, and ultimately improve production efficiency. The specific scheme is as follows:

[0005] In a first aspect, the present application discloses a data processing card screening method, which is applied to a baseboard management controller, comprising:

[0006] Adding a screening program for screening data processing cards to a test program of an equipment production line; the equipment production line is used to produce equipment that provides artificial intelligence services;

[0007] The addresses to be scanned of the slots on the data processing card are scanned and judged in turn using the preset judgment rules in the screening program to obtain corresponding scanning judgment results; the addresses to be scanned are addresses used for communication on a communication bus implemented based on a preset serial communication protocol;

[0008] Defective data processing cards among the data processing cards are screened according to the scanning judgment result, and a return authorization process is performed on the defective data processing cards.

[0009] Optionally, before sequentially using the preset judgment rules in the screening program to scan and judge the to-be-scanned addresses of the slots on the data processing card to obtain corresponding scanning judgment results, the method further includes:

[0010] Responding to the hybrid stress testing instruction in the test program to perform hybrid stress testing on the data server, and monitoring the system status of the current system after the hybrid stress testing instruction is executed;

[0011] When the current system is in an idle state, the step of sequentially using the preset judgment rules in the screening program to scan and judge the to-be-scanned addresses of each slot on the data processing card is triggered.

[0012] Optionally, before sequentially using the preset judgment rules in the screening program to scan and judge the to-be-scanned addresses of the slots on the data processing card to obtain corresponding scanning judgment results, the method further includes:

[0013] Acquire a first control instruction for indicating stopping the device status check by polling, and trigger the first control instruction;

[0014] Correspondingly, after screening out bad data processing cards from the data processing cards according to the scanning judgment result and executing a return authorization process for the bad data processing cards, the method further includes:

[0015] A corresponding second control instruction for indicating resumption of device status checking by polling is obtained for the first control instruction, and the second control instruction is triggered.

[0016] Optionally, the scanning and judging the to-be-scanned addresses of each slot on the data processing card by using the preset judgment rules in the screening program in sequence to obtain corresponding scanning and judgment results includes:

[0017] After the system is started and the system is entered, a first data processing card to be scanned is determined, and the first data processing card to be scanned is used as a current data processing card to be scanned, and then the screening program is run to continuously scan the addresses to be scanned of each slot on the current data processing card to be scanned;

[0018] Using the first judgment rule in the preset judgment rule to judge whether the first address in the to-be-scanned addresses is confirmed to be accessible, and obtaining a first scanning judgment result; wherein the first address is an address that indicates that the current to-be-scanned data processing card is a bad data processing card when it is inaccessible;

[0019] Based on the first scanning judgment result, a second judgment rule in the preset judgment rule is used to judge whether a second address in the to-be-scanned address is confirmed to be accessible, and a second scanning judgment result is obtained; wherein the second address is an address that indicates that the current to-be-scanned data processing card is a bad data processing card when it is accessible;

[0020] Based on the second scanning judgment result, the third judgment rule in the preset judgment rule is used to judge whether the third address in the to-be-scanned address has no response, and the third scanning judgment result is obtained; wherein, the third address is the address that characterizes that the current to-be-scanned data processing card is a bad data processing card when there is no response.

[0021] Optionally, the process of screening bad data processing cards among the data processing cards according to the scanning judgment result and executing a return authorization process for the bad data processing cards further includes:

[0022] When at least one of the third addresses is able to respond, determining a second data processing card to be scanned, and using the second data processing card to be scanned as a current data processing card to be scanned;

[0023] Jump to the step of running the screening program to continuously scan the to-be-scanned addresses of each slot on the current to-be-scanned data processing card until all the to-be-scanned data processing cards have completed scanning, and then restart the control power supply of the baseboard management controller;

[0024] Accordingly, screening bad data processing cards from the data processing cards according to the scanning judgment result, and executing a return authorization process for the bad data processing cards, includes:

[0025] After the control power supply is restarted, the bad data processing cards in the data processing cards are re-screened using the screening program according to a preset number of tests, and the test results of each test are obtained;

[0026] A target defective data processing card is determined according to the test result, and a return authorization process is performed on the target defective data processing card.

[0027] Optionally, the data processing card screening method further includes:

[0028] Setting a corresponding continuous scanning time for the screening program according to the number of the data processing cards;

[0029] Correspondingly, the scanning and judging of the to-be-scanned addresses of each slot on the data processing card by using the preset judgment rules in the screening program in sequence includes:

[0030] The preset judgment rules in the screening program are used in sequence to perform continuous scanning judgment on the to-be-scanned addresses of each slot on the data processing card within the continuous scanning time.

[0031] Optionally, the data processing card screening method further includes:

[0032] Using the screening program to monitor the hardware devices on the same communication bus as the data processing card to determine whether the hardware devices are abnormal;

[0033] When an abnormality occurs in the hardware device, corresponding alarm information is generated for the hardware device, and an early warning is issued through the alarm information.

[0034] In a second aspect, the present application discloses a data processing card screening device, which is applied to a baseboard management controller, comprising:

[0035] A program adding module, used to add a screening program for screening data processing cards to a test program of an equipment production line; the equipment production line is used to produce equipment that provides artificial intelligence services;

[0036] A scanning and judging module, used to sequentially use the preset judgment rules in the screening program to scan and judge the to-be-scanned addresses of each slot on the data processing card to obtain corresponding scanning and judging results; the to-be-scanned addresses are addresses used for communication on a communication bus implemented based on a preset serial communication protocol;

[0037] The bad card screening module is used to screen bad data processing cards among the data processing cards according to the scanning judgment result, and execute a return authorization process for the bad data processing cards.

[0038] In a third aspect, the present application discloses an electronic device, comprising a processor and a memory; wherein the memory is used to store a computer program, and the computer program is loaded and executed by the processor to implement the data processing card screening method as described above.

[0039] In a fourth aspect, the present application discloses a computer-readable storage medium for storing a computer program; wherein the computer program, when executed by a processor, implements the data processing card screening method as described above.

[0040] The present application provides a data processing card screening method, which is applied to a baseboard management controller, comprising: adding a screening program for screening data processing cards in a test program of an equipment production line; the equipment production line is used to produce equipment that provides artificial intelligence services; sequentially using each preset judgment rule in the screening program to scan and judge the to-be-scanned address of each slot on the data processing card to obtain a corresponding scanning judgment result; the to-be-scanned address is an address used for communicating on a communication bus implemented based on a preset serial communication protocol; screening out defective data processing cards in the data processing cards according to the scanning judgment result, and executing a return authorization process for the defective data processing cards.

[0041] The beneficial technical effect of the present application is: when the AI ​​server is produced, a screening program for screening data processing cards is integrated into the test program of its production line, and the address of each slot on the data processing card is scanned and judged using the judgment rules in the screening program. The scanning and judgment results can indicate the bad cards in the data processing card and screen them out. In this way, the screening program is run in advance during the server production test to eliminate bad data processing cards in advance, ensuring that bad cards are no longer exposed in the production link, and thus no longer causing abnormal termination of the test program, and ultimately improving production efficiency.

[0042] In addition, the present application provides a data processing card screening device, equipment and storage medium, which correspond to the above-mentioned data processing card screening method and have the same effect as above. BRIEF DESCRIPTION OF THE DRAWINGS

[0043] In order to more clearly illustrate the embodiments of the present invention or the technical solutions in the prior art, the drawings required for use in the embodiments or the description of the prior art will be briefly introduced below. Obviously, the drawings described below are only embodiments of the present invention. For ordinary technicians in this field, other drawings can be obtained based on the provided drawings without paying creative work.

[0044] Figure 1 A flow chart of a data processing card screening method disclosed in this application;

[0045] Figure 2 This is a structural schematic diagram of a data processing card screening device disclosed in this application;

[0046] Figure 3 This is a structural diagram of an electronic device disclosed in this application. DETAILED DESCRIPTION

[0047] The following will be combined with the drawings in the embodiments of the present invention to clearly and completely describe the technical solutions in the embodiments of the present invention. Obviously, the described embodiments are only part of the embodiments of the present invention, not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without creative work are within the scope of protection of the present invention.

[0048] BlueField-3 is the first DPU designed for AI and accelerated computing. A BlueField-3 DPU can provide data center services equivalent to those provided by up to 300 CPU (Central Processing Unit) cores, freeing up valuable CPU resources to run mission-critical applications.

[0049] The BlueField-3 DPU can offload and isolate the infrastructure services of the data center from business applications, realizing the transformation from traditional infrastructure to a modern "zero trust" environment. It can authenticate each user in the data center, ensuring the security of the enterprise from the cloud to the core data center and then to the edge, while also achieving greater improvements in efficiency and performance. As the industry's first 400G Ethernet and NDR InfiniBand DPU, BlueField-3 has excellent network performance. Compared with the previous generation, it has 10 times the accelerated computing power, 16 Arm A78 CPU cores and 4 times the encryption speed. BlueField-3 is also the first DPU to support the fifth-generation PCIE bus and provide data center time synchronization acceleration.

[0050] Given the initial launch of the BF3 card and the immaturity of some of NVIDIA's processing technologies, the BF3 cards actually obtained by manufacturers have a certain defect rate. A defective BF3 card will cause the entire I2C bus (Inter-Integrated Circuit Bus) monitored by the out-of-band BMC (Baseboard Management Controller) to hang, resulting in the probability of failure of out-of-band monitoring in the stress test aging link during actual production of AI servers, ultimately causing thermal anomalies, which in turn cause the card to drop and ultimately lead to stress test failure. This requires production and R&D personnel to be able to identify defective BF3 cards as soon as possible and ultimately improve production efficiency.

[0051] Currently, when BF3 stress testing fails in AI server production, production line maintenance personnel need to intervene and replace each card one by one to check. This method is very easy to waste time, reduce production efficiency, and consume manpower costs.

[0052] To this end, the present application provides a data processing card screening solution that can eliminate bad BF3 cards in advance, ensure that no bad cards are exposed in the production process, and ultimately improve production efficiency.

[0053] The embodiment of the present invention discloses a data processing card screening method, which is applied to a baseboard management controller, see Figure 1 As shown, the method includes:

[0054] Step S11: Adding a screening program for screening data processing cards to the test program of an equipment production line; the equipment production line is used to produce equipment that provides artificial intelligence services.

[0055] In the embodiment of the present application, the data processing card implements AI tasks and accelerated computing through the DPU design, and accelerates the delivery of AI applications by providing high-performance data processing capabilities. In a specific implementation, the data processing card is illustrated by taking the BlueField-3 card as an example. It can be understood by those of ordinary skill in the art that the BlueField-3 card is only for illustration and does not limit the type of the above-mentioned data processing card, and any DPU that can provide high-performance data processing capabilities can be implemented. The following embodiments are all illustrated by taking the BlueField-3 card as an example, and will not be repeated later.

[0056] It is understandable that since data processing cards are designed for AI tasks and accelerated computing, the process of screening defective cards for equipment production lines is the same as the process for producing equipment used to provide AI, such as the production process for AI servers.

[0057] In order to overcome the situation in which the test program is abnormally terminated only when a BlueField-3 card failure occurs during the equipment production line, a screening program for screening data processing cards is added to the test program of the equipment production line in the embodiment of the present application. The screening program is added to a relatively early position in the test program. After executing the necessary prerequisite processes in the test program, the screening program is executed to screen BF3 defective cards in advance to ensure that no defective cards are exposed in the production process.

[0058] Step S12: using the preset judgment rules in the screening program in turn to scan and judge the to-be-scanned addresses of the slots on the data processing card to obtain corresponding scanning judgment results.

[0059] In the embodiment of the present application, various judgment rules are pre-set in the screening program, and each judgment rule can realize the fault judgment of the data processing card. The judgment rule scans and judges the address to be scanned of each slot on the BlueField-3 card, and then obtains the corresponding scanning judgment result.

[0060] The address to be scanned is an address used for communication on a communication bus implemented based on a preset serial communication protocol. Taking the BlueField-3 card as an example, the address to be scanned is the SMBUS (System Management BUS) of slot 0-7 on the BlueField-3 card. SMBUS is a two-wire serial communication protocol based on the I2C bus specification, mainly used for system management and monitoring.

[0061] It should be pointed out that the judgment logic for the address to be scanned corresponding to each judgment rule can be provided by the manufacturer corresponding to the data processing card. Different manufacturers may have different judgment logics for the address to be scanned of the data processing card. The specific situation is based on the address to be scanned provided by the manufacturer. The address information corresponding to the address to be scanned provided by the manufacturer is used to replace the parameter information of the address to be scanned in each preset judgment rule in the screening program to implement the judgment logic for the current data processing card.

[0062] Step S13: screening out defective data processing cards among the data processing cards according to the scanning judgment result, and executing a return authorization process for the defective data processing cards.

[0063] In the embodiment of the present application, each preset judgment rule in the screening program scans and judges the to-be-scanned addresses of each slot on the data processing card, and obtains a corresponding scanning judgment result. According to the scanning judgment result, the bad data processing cards in the data processing cards can be screened, so that the bad data processing cards can be excluded in advance.

[0064] Furthermore, after filtering out the bad data processing cards, the return material authorization (RMA) process is performed on the bad data processing cards. The RMA process is particularly important for high-tech manufacturing industries, which can effectively control and track product returns, avoid economic losses and improve customer satisfaction.

[0065] The present application provides a data processing card screening method, which is applied to a baseboard management controller, comprising: adding a screening program for screening data processing cards in a test program of an equipment production line; the equipment production line is used to produce equipment that provides artificial intelligence services; sequentially using each preset judgment rule in the screening program to scan and judge the to-be-scanned address of each slot on the data processing card to obtain a corresponding scanning judgment result; the to-be-scanned address is an address used for communicating on a communication bus implemented based on a preset serial communication protocol; screening out defective data processing cards in the data processing cards according to the scanning judgment result, and executing a return authorization process for the defective data processing cards.

[0066] The beneficial technical effect of the present application is: when the AI ​​server is produced, a screening program for screening data processing cards is integrated into the test program of its production line, and the address of each slot on the data processing card is scanned and judged using the judgment rules in the screening program. The scanning and judgment results can indicate the bad cards in the data processing card and screen them out. In this way, the screening program is run in advance during the server production test to eliminate bad data processing cards in advance, ensuring that bad cards are no longer exposed in the production link, and thus no longer causing abnormal termination of the test program, and ultimately improving production efficiency.

[0067] In a specific implementation manner, before using the screening program to screen out bad data processing cards, the following steps are also included:

[0068] Responding to the hybrid stress testing instruction in the test program to perform hybrid stress testing on the data server, and monitoring the system status of the current system after the hybrid stress testing instruction is executed;

[0069] When the current system is in an idle state, the step of sequentially using the preset judgment rules in the screening program to scan and judge the to-be-scanned addresses of each slot on the data processing card is triggered.

[0070] In the embodiment of the present application, before executing the screening program, the mixed stress test is first run through the test program to put the system in an idle state, that is, an idle state. In the idle state, the program or system is in a period of time when no tasks or activities are executed. In this state, the preset judgment rules in the screening program are continued to be used in sequence to scan and judge the addresses to be scanned of each slot on the data processing card.

[0071] In a specific implementation, since the screening program is an external program added later than the BMC, a conflict error will be generated with the screening program when the system is monitored through the BMC. Therefore, it is necessary to stop polling the BMC before executing the screening program. Specifically, the method of sequentially using the preset judgment rules in the screening program to scan and judge the to-be-scanned addresses of each slot on the data processing card to obtain the corresponding scanning judgment results also includes the following steps:

[0072] A first control instruction for indicating stopping the device status check in a polling manner is obtained, and the first control instruction is triggered.

[0073] It is understandable that after the bad data processing cards are screened out by the screening program, it is necessary to continue to monitor the system through the BMC, so it is necessary to confirm the recovery of polling. Specifically, the corresponding second control instruction for characterizing the recovery of the device status check by polling is obtained for the first control instruction, and the second control instruction is triggered.

[0074] In a specific implementation, various judgment rules are pre-set in the screening program, and each judgment rule can realize the fault judgment of the data processing card. The judgment rule scans and judges the to-be-scanned addresses of each slot on the BlueField-3 card, and then obtains the corresponding scanning judgment result. Specifically, the scanning and judging of the to-be-scanned addresses of each slot on the data processing card by using the preset judgment rules in the screening program in turn to obtain the corresponding scanning judgment result includes the following steps:

[0075] After the system is started and the system is entered, a first data processing card to be scanned is determined, and the first data processing card to be scanned is used as a current data processing card to be scanned, and then the screening program is run to continuously scan the addresses to be scanned of each slot on the current data processing card to be scanned;

[0076] Using a first judgment rule in the preset judgment rules to judge whether a first address in the to-be-scanned addresses is confirmed to be accessible, and obtaining a first scanning judgment result;

[0077] Based on the first scanning judgment result, using a second judgment rule in the preset judgment rule to judge whether a second address in the to-be-scanned address is confirmed to be accessible, and obtaining a second scanning judgment result;

[0078] Based on the second scanning judgment result, a third judgment rule in the preset judgment rule is used to judge whether a third address in the to-be-scanned addresses has no response, and a third scanning judgment result is obtained.

[0079] In the embodiment of the present application, the data processing card is also described as a BlueField-3 card. After the entire system is booted up, the system runs the script and continuously scans the SMBUS of slots 0-7 of the BF3 card. Among them, PCA9546 is at the address 0x70, which is used to expand the I2C bus, allowing multiple devices with the same I2C address to be connected to a microcontroller.

[0080] When the addresses to be scanned of each slot are continuously scanned, the slot number to be scanned is first printed on the screen. Further, the first judgment rule in the preset judgment rules is used to judge whether the first address in the addresses to be scanned is confirmed to be accessible.

[0081] It should be pointed out that the judgment logic for the address to be scanned corresponding to each judgment rule in the preset judgment rules can be provided by the manufacturer corresponding to the data processing card. Different manufacturers may have different judgment logics for the address to be scanned of the data processing card. The specific situation shall be subject to the address to be scanned provided by the manufacturer. The address information corresponding to the address to be scanned provided by the manufacturer is used to replace the parameter information of the address to be scanned in each preset judgment rule in the screening program to realize the judgment logic for the current data processing card.

[0082] Taking the first address provided by a specific manufacturer as an example, first, use the first judgment rule to determine whether 0x52 (CEC), 0x1F (BF3) - 7 bit addressing can be accessed. If it is confirmed that it can be accessed, continue to scan using other rules in the preset judgment rules, otherwise stop the script running, and go through the RMA process for the BF3 of the slot being scanned. It can be seen that the first address is the address that characterizes the current data processing card to be scanned as a bad data processing card when it cannot be accessed. That is, if the first address of the current scanning slot of the BF3 card cannot be accessed, then the BlueField-3 card is a bad BF3 card. The address information of the first address provided by different manufacturers may be different, and the specific situation shall be based on the information provided by the manufacturer corresponding to the actual data processing card to be scanned.

[0083] It is understandable that after the judgment is completed using the first judgment rule, the subsequent process is executed based on the generated first scan judgment result. If the first address is accessible, the second judgment rule in the preset judgment rule is continued to be used to judge whether the second address in the address to be scanned is confirmed to be accessible. Exemplarily, taking the data provided by the above-mentioned manufacturer as an example, the second judgment rule is used to judge whether 0x53 can be accessed. If it is confirmed that it can be accessed, then stop the script and take the BF3 of the slot being scanned through the RMA process. It can be seen that the second address is the address that characterizes the current data processing card to be scanned as a bad data processing card when it can be accessed. That is, if the second address of the current scanning slot of the BF3 card can be accessed, then the BlueField-3 card is a bad BF3 card. The address information of the second address provided by different manufacturers may be different, and the specific situation depends on the information provided by the manufacturer corresponding to the actual data processing card to be scanned.

[0084] Similarly, after the judgment is completed using the second judgment rule, the subsequent process is executed based on the generated second scan judgment result. If the second address cannot be accessed, the third judgment rule in the preset judgment rule is continued to be used to judge whether the third address in the address to be scanned has no response. It can be understood that the address can be accessed but not necessarily all will respond. Therefore, the third address is continued to be judged. For example, if none of the following addresses respond: 0x52, 0x1F, 0x32, 0x70 (I2C switch), then the script is stopped and the BF3 of the slot being scanned goes through the RMA process. It can be seen that the third address is the address that characterizes the current data processing card to be scanned as a bad data processing card when there is no response. That is, if the third address of the current scan slot of the BF3 card has no response, then the BlueField-3 card is a bad BF3 card. The address information of the third address provided by different manufacturers may be different. The specific situation shall be subject to the information provided by the manufacturer corresponding to the actual data processing card to be scanned.

[0085] It should be noted that if there is an address that can respond in the third address, the next process is continued to be executed, and other BF3 cards are scanned. Specifically, when there is at least one address that can respond in the third address, the second data processing card to be scanned is determined, and the second data processing card to be scanned is used as the current data processing card to be scanned; jump to the step of running the screening program to continuously scan the addresses to be scanned of each slot on the current data processing card to be scanned until all the data processing cards to be scanned are scanned, and then restart the control power supply of the baseboard management controller.

[0086] In the embodiment of the present application, each time the preset judgment rules in the screening program are used to continuously scan and judge the addresses to be scanned in each slot on the data processing card, one execution time corresponds to the execution time. Therefore, the corresponding continuous scanning time is set for the screening program according to the number of the data processing cards, and the script is used to continuously scan within the continuous scanning time until all the data processing cards to be scanned are scanned. For example, if there are 8 BF3 cards + GPU modules to be scanned, the script can be set to scan continuously for 5 minutes.

[0087] In addition, during the test using the screening program, as long as the screening conditions of the bad data processing card are found, an error will be reported (Unspecified error). Each error will be recorded and a unified error will be reported after the scanning is completed within the continuous scanning time.

[0088] In an embodiment of the present application, after completing a test, restart the control power supply of the baseboard management controller, Power cycle (DC Cycle), and retest. The accuracy of the test is ensured by multiple tests. Specifically, after the control power supply is restarted, the bad data processing cards in the data processing cards are rescreened using the screening program according to a preset number of tests, and the test results of each time are obtained; the target bad data processing card is determined according to the test results, and a return authorization process is executed for the target bad data processing card. For example, the process can be incorporated into AC (Alternating Current) or DC (Direct Current), and tested 10 times to ensure the accuracy of the test.

[0089] In a specific implementation, the present invention can also be used to evaluate other out-of-band monitoring hardware devices on the same i2cbus, such as PCIE cards, and predict in advance when an abnormality is detected. Specifically, the following steps may also be included:

[0090] Using the screening program to monitor the hardware devices on the same communication bus as the data processing card to determine whether the hardware devices are abnormal;

[0091] When an abnormality occurs in the hardware device, corresponding alarm information is generated for the hardware device, and an early warning is issued through the alarm information.

[0092] In the embodiment of the present application, the execution logic in the screening program is used to monitor other hardware devices on the same i2c bus. If there is a hardware device abnormality, it is screened out in the same way, and the bad hardware device is excluded in advance, ensuring that the production link will no longer expose bad hardware devices, and thus no longer cause abnormal termination of the test program, and ultimately improve production efficiency.

[0093] Correspondingly, the embodiment of the present application also discloses a data processing card screening device, which is applied to a baseboard management controller, see Figure 2 As shown, the device comprises:

[0094] A program adding module 11 is used to add a screening program for screening data processing cards to a test program of an equipment production line; the equipment production line is used to produce equipment that provides artificial intelligence services;

[0095] The scanning judgment module 12 is used to sequentially use the preset judgment rules in the screening program to scan and judge the to-be-scanned addresses of each slot on the data processing card to obtain corresponding scanning judgment results; the to-be-scanned addresses are addresses used for communication on a communication bus implemented based on a preset serial communication protocol;

[0096] The bad card screening module 13 is used to screen bad data processing cards among the data processing cards according to the scanning judgment result, and execute a return authorization process for the bad data processing cards.

[0097] Among them, for more specific working processes of the above-mentioned modules, please refer to the corresponding contents disclosed in the aforementioned embodiments, which will not be repeated here.

[0098] It can be seen that the above-mentioned scheme of this embodiment is applied to a baseboard management controller, including: adding a screening program for screening data processing cards in the test program of the equipment production line; the equipment production line is used to produce equipment that provides artificial intelligence services; the addresses to be scanned of each slot on the data processing card are scanned and judged in turn using the preset judgment rules in the screening program to obtain corresponding scanning judgment results; the addresses to be scanned are addresses used for communication on a communication bus implemented based on a preset serial communication protocol; bad data processing cards in the data processing cards are screened according to the scanning judgment results, and a return authorization process is executed for the bad data processing cards.

[0099] The beneficial technical effect of the present application is: when the AI ​​server is produced, a screening program for screening data processing cards is integrated into the test program of its production line, and the address of each slot on the data processing card is scanned and judged using the judgment rules in the screening program. The scanning and judgment results can indicate the bad cards in the data processing card and screen them out. In this way, the screening program is run in advance during the server production test to eliminate bad data processing cards in advance, ensuring that bad cards are no longer exposed in the production link, and thus no longer causing abnormal termination of the test program, and ultimately improving production efficiency.

[0100] Furthermore, the present application also discloses an electronic device. Figure 3 This is a structural diagram of an electronic device 20 according to an exemplary embodiment, and the content in the diagram cannot be regarded as any limitation on the scope of use of the present application.

[0101] Figure 3 The present invention provides a schematic diagram of the structure of an electronic device 20 provided in an embodiment of the present application. The electronic device 20 may specifically include: at least one processor 21, at least one memory 22, a power supply 23, a communication interface 24, an input / output interface 25, and a communication bus 26. The memory 22 is used to store a computer program, which is loaded and executed by the processor 21 to implement the relevant steps in the data processing card screening method disclosed in any of the above embodiments. In addition, the electronic device 20 in this embodiment may specifically be a server.

[0102] In this embodiment, the power supply 23 is used to provide working voltage for each hardware device on the electronic device 20; the communication interface 24 can create a data transmission channel between the electronic device 20 and the external device, and the communication protocol it follows is any communication protocol that can be applied to the technical solution of the present application, and is not specifically limited here; the input and output interface 25 is used to obtain external input data or output data to the outside world, and its specific interface type can be selected according to specific application needs and is not specifically limited here.

[0103] In addition, the memory 22 as a carrier for resource storage may be a read-only memory, a random access memory, a disk or an optical disk, etc. The resources stored thereon may include an operating system 221, a computer program 222 and data 223, etc. The data 223 may include various data. The storage method may be temporary storage or permanent storage.

[0104] The operating system 221 is used to manage and control the hardware devices and computer programs 222 on the electronic device 20, and can be Windows Server, Netware, Unix, Linux, etc. In addition to the computer program that can be used to complete the data processing card screening method performed by the electronic device 20 disclosed in any of the aforementioned embodiments, the computer program 222 can further include computer programs that can be used to complete other specific tasks.

[0105] Furthermore, the embodiment of the present application also discloses a computer-readable storage medium, wherein the computer-readable storage medium mentioned here includes a random access memory (RAM), a memory, a read-only memory (ROM), an electrically programmable ROM, an electrically erasable programmable ROM, a register, a hard disk, a magnetic disk or an optical disk, or any other form of storage medium known in the technical field. Wherein, when the computer program is executed by the processor, the aforementioned data processing card screening method is implemented. For the specific steps of the method, reference can be made to the corresponding contents disclosed in the aforementioned embodiment, and no further description will be given here.

[0106] Furthermore, an embodiment of the present application also provides a computer program product, including a computer program / instruction, which implements any one of the above-mentioned data processing card screening methods when executed by a processor.

[0107] In this specification, each embodiment is described in a progressive manner, and each embodiment focuses on the differences from other embodiments. The same or similar parts between the embodiments can be referred to each other. For the device disclosed in the embodiment, since it corresponds to the method disclosed in the embodiment, the description is relatively simple, and the relevant parts can be referred to the method part.

[0108] The steps of the data processing card screening method or algorithm described in conjunction with the embodiments disclosed herein can be implemented directly using hardware, a software module executed by a processor, or a combination of the two. The software module can be placed in a random access memory (RAM), a memory, a read-only memory (ROM), an electrically programmable ROM, an electrically erasable programmable ROM, a register, a hard disk, a removable disk, a CD-ROM, or any other form of storage medium known in the technical field.

[0109] Finally, it should be noted that, in this article, relational terms such as first and second, etc. are only used to distinguish one entity or operation from another entity or operation, and do not necessarily require or imply any such actual relationship or order between these entities or operations. Moreover, the terms "include", "comprise" or any other variants thereof are intended to cover non-exclusive inclusion, so that a process, method, article or device including a series of elements includes not only those elements, but also other elements not explicitly listed, or also includes elements inherent to such process, method, article or device. In the absence of further restrictions, the elements defined by the sentence "comprise a ..." do not exclude the presence of other identical elements in the process, method, article or device including the elements.

[0110] The above is a detailed introduction to a data processing card screening method, device, equipment and medium provided by the present invention. Specific examples are used in this article to illustrate the principles and implementation methods of the present invention. The description of the above embodiments is only used to help understand the method of the present invention and its core idea; at the same time, for a general technician in this field, according to the idea of ​​the present invention, there will be changes in the specific implementation method and application scope. In summary, the content of this specification should not be understood as a limitation on the present invention.

Claims

1. A data processing card screening method, characterized in that: Applicable to baseboard management controllers, including: Adding a screening program for screening data processing cards to a test program of an equipment production line; the equipment production line is used to produce equipment that provides artificial intelligence services; The addresses to be scanned of the slots on the data processing card are scanned and judged in turn using the preset judgment rules in the screening program to obtain corresponding scanning judgment results; the addresses to be scanned are addresses used for communication on a communication bus implemented based on a preset serial communication protocol; Defective data processing cards among the data processing cards are screened according to the scanning judgment result, and a return authorization process is performed on the defective data processing cards.

2. The data processing card screening method according to claim 1, characterized in that: Before sequentially using the preset judgment rules in the screening program to scan and judge the to-be-scanned addresses of each slot on the data processing card to obtain the corresponding scanning judgment results, the method further includes: Responding to the hybrid stress testing instruction in the test program to perform hybrid stress testing on the data server, and monitoring the system status of the current system after the hybrid stress testing instruction is executed; When the current system is in an idle state, the step of sequentially using the preset judgment rules in the screening program to scan and judge the to-be-scanned addresses of each slot on the data processing card is triggered.

3. The data processing card screening method according to claim 1, characterized in that: Before sequentially using the preset judgment rules in the screening program to scan and judge the to-be-scanned addresses of each slot on the data processing card to obtain the corresponding scanning judgment results, the method further includes: Acquire a first control instruction for indicating stopping the device status check by polling, and trigger the first control instruction; Correspondingly, after screening out bad data processing cards from the data processing cards according to the scanning judgment result and executing a return authorization process for the bad data processing cards, the method further includes: A corresponding second control instruction for indicating resumption of device status checking by polling is obtained for the first control instruction, and the second control instruction is triggered.

4. The data processing card screening method according to claim 1, characterized in that: The method of sequentially using the preset judgment rules in the screening program to scan and judge the to-be-scanned addresses of each slot on the data processing card to obtain corresponding scanning judgment results includes: After the system is started and the system is entered, a first data processing card to be scanned is determined, and the first data processing card to be scanned is used as a current data processing card to be scanned, and then the screening program is run to continuously scan the addresses to be scanned of each slot on the current data processing card to be scanned; Using the first judgment rule in the preset judgment rule to judge whether the first address in the to-be-scanned addresses is confirmed to be accessible, and obtaining a first scanning judgment result; wherein the first address is an address that indicates that the current to-be-scanned data processing card is a bad data processing card when it is inaccessible; Based on the first scanning judgment result, a second judgment rule in the preset judgment rule is used to judge whether a second address in the to-be-scanned address is confirmed to be accessible, and a second scanning judgment result is obtained; wherein the second address is an address that indicates that the current to-be-scanned data processing card is a bad data processing card when it is accessible; Based on the second scanning judgment result, the third judgment rule in the preset judgment rule is used to judge whether the third address in the to-be-scanned address has no response, and the third scanning judgment result is obtained; wherein, the third address is the address that characterizes that the current to-be-scanned data processing card is a bad data processing card when there is no response.

5. The data processing card screening method according to claim 4, characterized in that: The process of screening bad data processing cards among the data processing cards according to the scanning judgment result and executing a return authorization process for the bad data processing cards also includes: When at least one of the third addresses is able to respond, determining a second data processing card to be scanned, and using the second data processing card to be scanned as a current data processing card to be scanned; Jump to the step of running the screening program to continuously scan the to-be-scanned addresses of each slot on the current to-be-scanned data processing card until all the to-be-scanned data processing cards have completed scanning, and then restart the control power supply of the baseboard management controller; Accordingly, screening bad data processing cards from the data processing cards according to the scanning judgment result, and executing a return authorization process for the bad data processing cards, includes: After the control power supply is restarted, the bad data processing cards in the data processing cards are re-screened using the screening program according to a preset number of tests, and the test results of each test are obtained; A target defective data processing card is determined according to the test result, and a return authorization process is performed on the target defective data processing card.

6. The data processing card screening method according to claim 1, characterized in that: Also includes: Setting a corresponding continuous scanning time for the screening program according to the number of the data processing cards; Correspondingly, the scanning and judging of the to-be-scanned addresses of each slot on the data processing card by using the preset judgment rules in the screening program in sequence includes: The preset judgment rules in the screening program are used in sequence to perform continuous scanning judgment on the to-be-scanned addresses of each slot on the data processing card within the continuous scanning time.

7. The data processing card screening method according to any one of claims 1 to 6, characterized in that: Also includes: Using the screening program to monitor the hardware devices on the same communication bus as the data processing card to determine whether the hardware devices are abnormal; When an abnormality occurs in the hardware device, corresponding alarm information is generated for the hardware device, and an early warning is issued through the alarm information.

8. A data processing card screening device, characterized in that: Applicable to baseboard management controllers, including: A program adding module, used to add a screening program for screening data processing cards to a test program of an equipment production line; the equipment production line is used to produce equipment that provides artificial intelligence services; A scanning and judging module, used to sequentially use the preset judgment rules in the screening program to scan and judge the to-be-scanned addresses of each slot on the data processing card to obtain corresponding scanning and judging results; the to-be-scanned addresses are addresses used for communication on a communication bus implemented based on a preset serial communication protocol; The bad card screening module is used to screen bad data processing cards among the data processing cards according to the scanning judgment result, and execute a return authorization process for the bad data processing cards.

9. An electronic device, characterized in that: The electronic device comprises a processor and a memory; wherein the memory is used to store a computer program, and the computer program is loaded and executed by the processor to implement the data processing card screening method as claimed in any one of claims 1 to 7.

10. A computer-readable storage medium, characterized in that: Used to store computer programs; wherein when the computer program is executed by a processor, the data processing card screening method according to any one of claims 1 to 7 is implemented.

Images