Semiconductor packaging test optimization method and system

Abstract

The invention relates to the technical field of packaging testing, in particular to a semiconductor packaging testing optimization method and system, and the system comprises an equipment node generation module which is used for collecting packaging testing equipment data, obtaining the processing capability and connection parameters of the testing equipment data, and generating a packaging testing module; and compiling the processing capability and the connection parameters into a structured data entry representing a single device to generate a device node set. According to the invention, equipment functions and resources are integrated through equipment identification and performance evaluation, and the efficiency and accuracy of the packaging test are optimized. Through the accurate extraction of the device identifier and the measurement of the connection parameter, the coordination and resource allocation among the nodes are more reasonable, and the working efficiency of the nodes is improved. In the aspect of data flow path customization, a data transmission path is optimized according to equipment performance and requirements, and resource waste and transmission delay are reduced. Task dependence is defined, a data flow path is optimized, and data exchange between devices is accelerated.

Description

Technical Field

[0001] The present invention relates to the field of packaging and testing technology, and in particular to a semiconductor packaging and testing optimization method and system. Background Art

[0002] With the rapid development of the semiconductor industry, packaging and testing has become a key link in ensuring chip performance and reliability.

[0003] However, existing semiconductor packaging and testing technologies show limitations in processing load balancing and resource optimization. Existing technologies fail to integrate data flow and resource sharing between devices, resulting in uneven resource allocation, with some devices running at high loads for a long time while other devices are inefficient. This unbalanced configuration not only reduces processing efficiency, but may also accelerate equipment wear and shorten its service life. Summary of the invention

[0004] The purpose of the present invention is to solve the shortcomings of the prior art and to propose a semiconductor packaging test optimization method and system.

[0005] In order to achieve the above object, the present invention adopts the following technical solution: A semiconductor packaging test optimization system comprises: A device node generation module, used to collect packaging test device data, obtain processing capabilities and connection parameters of the test device data, compile the processing capabilities and connection parameters into structured data entries representing a single device, and generate a device node set; A test hyperedge establishment module is used to collect a test task set based on the device node set, analyze the data input and output of each test task in the test task set to determine the data flow requirements corresponding to each test task, extract the test task sequence of each test task according to the data flow requirements, extract the dependency relationship of each test in the test task set by comparing the test task sequence with the shared resources corresponding to each test task, map the data flow paths between each device node according to each device node in the device node set to determine the data path, link the device nodes that need to collaborate or share data to create a hyperedge, compile the hyperedge into a structured set, and generate a test hyperedge set; The network structure optimization module is used to measure the bandwidth utilization and delay between each device node in the device node set to calculate the data transmission cost, summarize the task requirements assigned to each of the device nodes, and determine the processing load according to the processing capacity and connection parameters of the test device data. According to the processing load, the hyperedges connecting the high-load device nodes are reallocated to the low-load device nodes to adjust the node connection. After each adjustment of the device node connection, the load of the device node is repeatedly calculated for iterative adjustment, seeking a configuration that minimizes cross-device data transmission and balances the processing load, and obtains an optimized network structure.

[0006] Preferably, it also includes: A collaborative scheduling module, based on the optimized network structure, monitors device load and task priority in real time, checks device availability and workload to compare current processing capabilities, reviews task deadlines and importance levels, compares task urgency, adjusts task allocation by reallocating tasks from overloaded devices to devices with idle capacity, rearranges the order of tasks according to priority to adjust the execution sequence, coordinates start and end time synchronization between devices to avoid conflicts and idle periods, and obtains a dynamic scheduling solution; The feedback adjustment module, based on the dynamic scheduling scheme, records the processing time and resource utilization during the test to collect performance data, aligns similar tasks and devices to compare historical performance indicators and current test results, adjusts processing capabilities and connection parameters according to performance deviations, updates device parameters, and modifies node connections and hyperedges to update network configurations at locations where bottlenecks are detected.

[0007] Preferably, the device node set includes a unique identification code, processing capability indicator, connection parameters and bandwidth for each device, the test hyperedge set includes hyperedges formed based on data flow requirements, task dependencies and shared resource information between devices, each hyperedge represents a group of devices that need to collaborate or share data, and the dynamic scheduling scheme includes task allocation adjusted according to current monitoring, task execution sequence re-ordered by priority, and a schedule for synchronous scheduling between devices.

[0008] Preferably, the device node generation module includes: The data collection submodule is used to collect packaging test equipment data and test site information, obtain equipment model, serial number and location, select each equipment unique identification code, extract equipment identifier, check data integrity, remove duplicate and erroneous data, organize and store in a structured format, and generate a set of equipment identifiers; A parameter calculation submodule, based on the device identifier set, compares the device processing speed and memory capacity, calculates the processing capacity, measures the network interface type and bandwidth, obtains the connection parameters, records the device input and output port information, summarizes each parameter, and generates a device parameter set; The node generation submodule, based on the device parameter set, compiles each parameter into a structured data entry representing a single device, establishes device node attributes, including identifiers, processing capabilities, and connection parameters, integrates each device information, and generates a device node set.

[0009] Preferably, the test hyperedge establishment module includes: A task collection submodule collects test tasks based on the device node set, retrieves associated requirements, obtains task names, input and output parameters and required resources, and generates a test task set; The requirement analysis submodule analyzes the data input and output of each task based on the test task set, determines the data flow requirements, compares the task sequence and shared resources, extracts the task dependency, and generates a task dependency set; The hyperedge generation submodule maps the data flow path between devices based on the task dependency set, determines the data path, links the nodes that need to collaborate or share data, creates hyperedges, compiles the hyperedges into a structured set, and generates a test hyperedge set.

[0010] Preferably, the network structure optimization module includes: A cost calculation submodule, based on the device node set and the test hyperedge set, measures the bandwidth usage and delay between nodes, calculates the data transmission cost, and generates a transmission cost set; A load analysis submodule, based on the transmission cost set, summarizes the task requirements assigned to each node, determines the processing load, analyzes the internal communication of the device cluster, compares the internal node interactions, and generates a node load set; The structure adjustment submodule, based on the node load set, reallocates the hyperedges connecting high-load nodes to nodes with lower loads, adjusts node connections, changes data flow paths, reduces transmission costs, modifies node links, repeatedly calculates costs and loads after each change, performs iterative adjustments, and obtains an optimized network structure.

[0011] Preferably, the collaborative scheduling module includes: A real-time monitoring submodule, based on the optimized network structure, monitors device load and task priority in real time, checks device availability and workload, compares current processing capabilities, and generates a real-time status set; a task adjustment submodule, based on the real-time status set, reviews task deadlines and importance levels, compares task urgency, reallocates tasks from overloaded devices to devices with spare capacity, adjusts task allocation, rearranges the order of tasks according to priority, adjusts the execution sequence, and generates an adjusted task sequence; The time coordination submodule coordinates the start and end times based on the adjusted task sequence, synchronizes the scheduling between devices, avoids conflicts and idle periods, and obtains a dynamic scheduling solution.

[0012] Preferably, the feedback adjustment module includes: A performance data collection submodule records processing time and resource utilization during the test based on the dynamic scheduling scheme, collects performance data, and generates a performance data set; A parameter update submodule, based on the performance data set, aligns similar tasks and devices, compares historical performance indicators and current test results, adjusts processing capabilities and connection parameters according to performance deviations, updates device parameters, and generates an updated device parameter set; The network configuration submodule modifies node connections and hyperedges and updates network configuration at the location where the bottleneck is detected based on the updated device parameter set.

[0013] A semiconductor packaging test optimization method is implemented based on the semiconductor packaging test optimization system, and comprises the following steps: S1: Extract the unique identification code of each device based on the package test equipment data and test site information, measure the network interface and bandwidth to obtain the connection parameters, and compare the processing speed and memory capacity to determine the processing capacity of each device. Structure the device data and generate a device node set for subsequent data path analysis and load balancing; obtain node integration data; S2: According to the node integration data, collect test tasks and analyze the data input and output of each task to determine the data flow requirements, extract task dependencies by comparing task sequences and shared resources, map the data flow paths between devices and link nodes that need to collaborate or share data to create hyperedges, generate a structured test hyperedge set, and obtain the test task dependency structure; S3: According to the test task dependency structure, measure the bandwidth usage and latency between nodes, calculate the data transmission cost, summarize the task requirements of each node to determine the processing load, analyze the internal communication of the device cluster and adjust the node connection, change the data flow path by reallocating the hyperedges connecting the high-load nodes to the nodes with lower loads, and repeat this process to minimize the cost and balance the load; obtain the optimized network configuration; S4: Based on the optimized network configuration, monitor the device load and task priority in real time, adjust the task allocation according to the current processing capacity of the device and the urgency of the task, synchronize the scheduling between devices to avoid conflicts and idle periods, analyze the performance data based on the processing time and resource utilization collected during the test, adjust the node connections and hyperedges, and optimize the test performance.

[0014] Compared with the prior art, the advantages and positive effects of the present invention are: In the present invention, device functions and resources are integrated through device identification and performance evaluation to optimize the efficiency and accuracy of packaging testing. The precise extraction of device identifiers and the determination of connection parameters make the coordination and resource allocation between nodes more reasonable, and improve the working efficiency of nodes. In terms of data flow path customization, the data transmission route is optimized according to device performance and requirements to reduce resource waste and transmission delay. Clarifying task dependencies and optimizing data flow paths accelerates data exchange between devices and improves the parallel processing capability and flexibility of jobs. Real-time dynamic monitoring and task scheduling further achieve the balance of device loads, avoid device overload, and improve response speed and overall processing capabilities. Through continuous feedback adjustment of performance data, device performance and test accuracy are enhanced. Optimal allocation and utilization of resources are achieved, data transmission costs are reduced, and efficiency and reliability are improved. BRIEF DESCRIPTION OF THE DRAWINGS

[0015] Figure 1 It is a system flow chart of the present invention. DETAILED DESCRIPTION

[0016] In order to make the purpose, technical solution and advantages of the present invention more clearly understood, the present invention is further described in detail below in conjunction with the accompanying drawings and embodiments. It should be understood that the specific embodiments described herein are only used to explain the present invention and are not intended to limit the present invention.

[0017] See also Figure 1 The present invention provides a technical solution: a semiconductor packaging test optimization system comprising: A device node generation module is used to collect packaging test device data, obtain the processing capability and connection parameters of the test device data, compile the processing capability and connection parameters into a structured data entry representing a single device, and generate a device node set; A test hyperedge establishment module is used to collect a test task set based on a device node set, analyze the data input and output of each test task in the test task set to determine the data flow requirements corresponding to each test task, extract the test task sequence of each test task according to the data flow requirements, extract the dependencies of each test in the test task set by comparing the test task sequence and the shared resources corresponding to each test task, map the data flow paths between each device node according to each device node in the device node set to determine the data path, link the device nodes that need to collaborate or share data to create hyperedges, compile the hyperedges into a structured set, and generate a test hyperedge set; The network structure optimization module is used to measure the bandwidth utilization and delay between each device node in the device node set to calculate the data transmission cost, summarize the task requirements assigned to each device node, and determine the processing load according to the processing capacity and connection parameters of the test device data. According to the processing load, the hyperedges connecting the high-load device nodes are reallocated to the low-load device nodes to adjust the node connection. After each adjustment of the device node connection, the load of the device node is repeatedly calculated for iterative adjustment, seeking the configuration that minimizes cross-device data transmission and balances the processing load, and obtains the optimized network structure; The collaborative scheduling module, based on the optimized network structure, monitors the equipment load and task priority in real time, checks the equipment availability and workload to compare the current processing capacity, reviews the task deadlines and importance levels, compares the task urgency, adjusts the task allocation by reallocating tasks from overloaded equipment to equipment with idle capacity, rearranges the order of tasks according to priority, adjusts the execution sequence, coordinates the start and end time synchronization, avoids conflicts and idle periods between equipment scheduling, and obtains a dynamic scheduling solution; The feedback adjustment module, based on the dynamic scheduling scheme, records the processing time and resource utilization during the test to collect performance data, aligns similar tasks and devices to compare historical performance indicators and current test results, adjusts the processing capacity and connection parameters according to performance deviations, updates device parameters, and modifies node connections and hyperedges to update network configurations at the location where bottlenecks are detected.

[0018] The device node set includes the unique identification code, processing capability indicators, connection parameters and bandwidth of each device. The test hyperedge set is a hyperedge formed based on the data flow requirements, task dependencies and shared resource information between devices. Each hyperedge represents a group of devices that need to collaborate or share data. The dynamic scheduling scheme includes task allocation adjusted according to current monitoring, task execution sequence re-ordered by priority, and a schedule for synchronous scheduling between devices.

[0019] In this embodiment, the device node generation module includes: The data collection submodule collects packaging test equipment data and test site information, obtains equipment model, serial number and location, selects each equipment unique identification code, extracts equipment identifiers, checks data integrity, removes duplicate and erroneous data, organizes and stores in a structured format, and generates equipment identifier sets; The parameter calculation submodule compares the device processing speed and memory capacity based on the device identifier set, calculates the processing capacity, measures the network interface type and bandwidth, obtains the connection parameters, records the device input and output port information, summarizes each parameter, and generates a device parameter set; The node generation submodule, based on the device parameter set, compiles each parameter into a structured data entry representing a single device, establishes device node attributes, including identifiers, processing capabilities, and connection parameters, integrates each device information, and generates a device node set.

[0020] Specifically, the data acquisition submodule collects packaging test equipment data and test site information. The specific operations are as follows: First, the system collects real-time data of each packaging test equipment through the integrated sensor network and interface module, including the equipment model, serial number and specific installation location. Next, the collected data is preliminarily screened to eliminate incomplete or duplicate data items. After that, the system standardizes the unique identification code of each device to generate a set of device identifiers. This process includes standardizing the format of the device serial number and ensuring the uniqueness and consistency of all device identifiers, ultimately forming a clearly defined device identifier database to provide an accurate input source for subsequent parameter calculations and node generation.

[0021] The parameter calculation submodule calculates device performance parameters based on the device identifier set. First, the system locates the detailed configuration information of each device through the device identifier set, including processor speed and memory capacity. This data is analyzed to calculate the processing capacity index of each device. This calculation involves synthesizing the processing speed and memory capacity into a comprehensive performance index. At the same time, the system also measures the network interface type and bandwidth of each device and records the configuration information of its input and output ports. These connection parameters are then aggregated to form a device parameter set, and each parameter precisely corresponds to a device identifier to ensure data consistency and traceability.

[0022] The node generation submodule builds device nodes based on the device parameter set. During this process, the system compiles the core data such as the device's identifier, processing power, connection parameters, etc. into the structured data entry of a single device. In this way, each device is given a clear node attribute label, including its performance indicators and network configuration. These device nodes are then integrated into a device node set to support further analysis and optimization operations in the system. The generation of the device node set is based on ensuring that each node completely and uniquely represents a physical device. Through this precise mapping, the system can effectively perform device management and performance monitoring to support the overall optimization of semiconductor packaging testing.

[0023] In this embodiment, the test hyperedge establishment module includes: The task collection submodule collects test tasks based on the device node set, retrieves associated requirements, obtains task names, input and output parameters, and required resources, and generates a test task set; The requirement analysis submodule analyzes the data input and output of each task based on the test task set, determines the data flow requirements, compares the task sequence and shared resources, extracts the task dependencies, and generates the task dependency set; The hyperedge generation submodule, based on the task dependency set, maps the data flow path between devices, determines the data path, links the nodes that need to collaborate or share data, creates hyperedges, compiles the hyperedges into structured sets, and generates a test hyperedge set.

[0024] Specifically, the task collection submodule collects test tasks based on the device node set. The operation process is detailed as follows: the system first communicates with the interface of each package test device to automatically obtain the current and scheduled test task information, including the task name, required input and output parameters, and related resource requirements. During this information collection process, the system also retrieves the requirements related to each task, such as specific device compatibility and resource coordination requirements, to ensure the integrity and practicality of the task information. After completing these steps, all relevant task information is organized and stored in the test task set, providing a data basis for the next step of demand analysis.

[0025] The requirements analysis submodule performs in-depth analysis of task data based on the test task set. The system first determines the data input and output requirements of each task, which includes a detailed analysis of the amount of data processed by the task, the type of data, and the processing rate. Then, the system compares the task sequences and their dependencies on shared resources to extract the dependencies between tasks. For example, some tasks may need to be executed sequentially, or multiple tasks may rely on the same data source. These dependencies are organized by the system and stored in the task dependency set to support subsequent hyperedge generation.

[0026] The hyperedge generation submodule maps the data flow paths between devices based on the task dependency set. During this process, the system analyzes how to share and transmit data most efficiently between devices, including determining the best data path to reduce latency and bandwidth consumption. Through a comprehensive evaluation of task dependencies and device capabilities, the system selects those device nodes that require collaboration or data sharing and creates so-called "hyperedges" between these nodes. These hyperedges represent the data flow and dependencies between devices and are the key to optimizing data processing and resource allocation. After the creation of these hyperedges, they are compiled into a structured collection to form a test hyperedge set, which provides the infrastructure for the system's network structure optimization and collaborative scheduling.

[0027] In this embodiment, the network structure optimization module includes: The cost calculation submodule measures the bandwidth usage and delay between nodes based on the device node set and the test hyperedge set, calculates the data transmission cost, and generates a transmission cost set; The load analysis submodule summarizes the task requirements assigned to each node based on the transmission cost set, determines the processing load, analyzes the internal communication of the device cluster, compares the internal node interactions, and generates the node load set; The structure adjustment submodule, based on the node load set, reallocates the hyperedges connecting high-load nodes to nodes with lower loads, adjusts node connections, changes data flow paths, reduces transmission costs, modifies node links, repeatedly calculates costs and loads after each change, performs iterative adjustments, and obtains an optimized network structure.

[0028] Specifically, the cost calculation submodule performs the calculation of data transmission cost based on the device node set and the test hyperedge set. The specific process includes: the system first monitors the bandwidth usage and communication delay between all device nodes, and uses these data to evaluate the data transmission cost between each node. To this end, a cost evaluation algorithm is adopted, which takes into account bandwidth consumption, delay time and possible error rate to calculate the cost of transmitting unit data from one node to another. These cost data are sorted by the system and stored in the transmission cost set, which provides basic data for subsequent load analysis and structural adjustment.

[0029] The load analysis submodule performs node load analysis based on the transmission cost set. This submodule summarizes the task requirements assigned to each node and determines the processing load of each node by analyzing these task requirements. In addition, the system analyzes the communication mode within the device cluster and compares the frequency and amount of data interaction between nodes to identify possible communication bottlenecks or inefficient nodes. In this way, a node load set is generated, recording the load level and communication activity of each node, providing decision support information for the structural adjustment of the network.

[0030] The structure adjustment submodule optimizes and adjusts the network structure based on the node load set. The operation steps are as follows: First, the system identifies those nodes with higher loads and analyzes the hyperedges connected to these nodes. Then, the system tries to redistribute some connections from high-load nodes to lower-load nodes to balance the processing load in the network and reduce the overall data transmission cost. After each structural adjustment, the system recalculates the cost and load of each node, and this iterative process continues until the best network structure configuration is found. In this way, the system finally obtains an optimized network structure, which is characterized by the lowest data transmission cost across devices and balanced processing load of each node, thereby improving the efficiency and responsiveness of the entire semiconductor packaging and testing system.

[0031] In this embodiment, the collaborative scheduling module includes: The real-time monitoring submodule monitors the device load and task priority in real time based on the optimized network structure, checks the device availability and workload, compares the current processing capacity, and generates a real-time status set; The task adjustment submodule reviews the task deadlines and importance levels based on the real-time status set, compares the task urgency, reallocates tasks from overloaded devices to devices with spare capacity, adjusts task allocation, rearranges the order of tasks according to priority, adjusts the execution sequence, and generates an adjusted task sequence; The time coordination submodule coordinates the start and end times based on the adjusted task sequence, synchronizes the scheduling between devices, avoids conflicts and idle periods, and obtains a dynamic scheduling solution.

[0032] Specifically, the real-time monitoring submodule implements real-time monitoring of device load and task priority based on the optimized network structure. The operation steps are detailed as follows: First, the system uses the optimized network structure to continuously monitor the device load and collect the workload data and current task priority information of each device. At the same time, the system checks the availability status of each device, identifies any possible operating failures or performance degradation, and ensures the accuracy and timeliness of the monitoring data. Based on this data, the system generates a real-time status set that contains the current processing capacity and task status of all devices, providing a decision-making basis for the task adjustment submodule.

[0033] The task adjustment submodule dynamically adjusts task allocation based on the real-time status set. The specific process is as follows: the system first reviews the deadline and importance level of each task to determine the urgency and priority of the task. For those tasks assigned to overloaded devices, the system will reallocate them to devices with idle processing capacity to balance the workload and avoid delays. In addition, the execution sequence of tasks is reordered according to priority to ensure that more urgent tasks can be processed first. This adjustment process ultimately generates an adjusted task sequence that optimizes the efficiency and timeliness of task execution.

[0034] The time coordination submodule performs time coordination between devices based on the adjusted task sequence. The operation steps include: the system coordinates the start and end time of each task according to the adjusted task sequence to synchronize the work progress between devices and avoid time conflicts and unnecessary idle periods during execution. This time coordination ensures that all devices can work together efficiently, minimize waiting time and improve overall work efficiency. Through these measures, the system ultimately generates a dynamic scheduling plan that responds to changes in network conditions and task requirements in real time, achieving optimal resource utilization and task execution in the semiconductor packaging and testing process.

[0035] In this embodiment, the feedback adjustment module includes: The performance data collection submodule records the processing time and resource utilization during the test based on the dynamic scheduling scheme, collects performance data, and generates a performance data set; The parameter update submodule aligns similar tasks and devices based on the performance data set, compares historical performance indicators and current test results, adjusts processing capabilities and connection parameters according to performance deviations, updates device parameters, and generates an updated device parameter set; The network configuration submodule, based on the updated set of device parameters, modifies node connections and hyperedges at the locations where bottlenecks are detected and updates the network configuration.

[0036] Specifically, the performance data collection submodule performs detailed recording of performance data during the test based on the dynamic scheduling scheme. The operation process of this module includes monitoring and recording the processing time and resource utilization of each device, as well as the efficiency of task execution. The system regularly collects these performance data and analyzes the data to evaluate the performance of each task and device. All collected data is organized and stored in the performance data set, which provides real-time and historical performance information for subsequent parameter updates and network configuration optimization.

[0037] The parameter update submodule adjusts and updates device parameters based on the performance data set. The specific operations are as follows: First, the system compares the current test results of similar tasks and devices with historical performance indicators to identify devices with poor performance or large performance deviations. For these devices, the system adjusts their processing power and connection parameters to optimize their performance. For example, if the processing power of a device is not sufficient to meet the current task requirements, the system may increase its resource allocation or optimize its input and output configuration. After completing these adjustments, the system generates an updated device parameter set, which includes all adjusted device parameters and provides basic data for the network configuration submodule.

[0038] The network configuration submodule optimizes and adjusts the network configuration based on the updated device parameter set. The operation steps include: the system first identifies the bottleneck locations in the network, such as nodes with high data transmission delays or connections with heavy loads. Then, based on the latest data in the device parameter set, the system modifies the connections and related hyperedges of these nodes to eliminate bottlenecks and optimize data flow paths. This includes reconfiguring the connection methods between nodes and adjusting the data traffic distribution to improve the overall performance and efficiency of the network.

[0039] The present invention provides a semiconductor packaging test optimization method, comprising the following steps: S1, extracts the unique identification code of each device based on the package test equipment data and test site information, measures the network interface and bandwidth to obtain connection parameters, compares the processing speed and memory capacity to determine the processing capacity of each device, structures the device data, generates a device node set for subsequent data path analysis and load balancing, and obtains node integration data; S2, based on the node integration data, collects test tasks and analyzes the data input and output of each task to determine the data flow requirements, extracts task dependencies by comparing task sequences and shared resources, maps the data flow paths between devices and links nodes that need to collaborate or share data to create hyperedges, and generates a structured test hyperedge set; obtains the test task dependency structure; S3, based on the test task dependency structure, measures the bandwidth usage and latency between nodes, calculates the data transmission cost, and summarizes the task requirements of each node to determine the processing load. It analyzes the internal communication of the device cluster and adjusts the node connection. It changes the data flow path by reallocating the hyperedges connecting high-load nodes to nodes with lower loads. It repeats this process to minimize costs and balance loads. It obtains the optimized network configuration. S4, based on the optimized network configuration, monitors device load and task priority in real time, adjusts task allocation according to the current processing capacity of the device and the urgency of the task, synchronizes scheduling between devices to avoid conflicts and idle periods, analyzes performance data based on the processing time and resource utilization collected during the test, adjusts node connections and hyperedges, and optimizes test performance.

Claims

1. A semiconductor packaging test optimization system, characterized in that: The system comprises: A device node generation module, used to collect packaging test device data, obtain processing capabilities and connection parameters of the test device data, compile the processing capabilities and connection parameters into structured data entries representing a single device, and generate a device node set; A test hyperedge establishment module is used to collect a test task set based on the device node set, analyze the data input and output of each test task in the test task set to determine the data flow requirements corresponding to each test task, extract the test task sequence of each test task according to the data flow requirements, extract the dependency relationship of each test in the test task set by comparing the test task sequence with the shared resources corresponding to each test task, map the data flow paths between each device node according to each device node in the device node set to determine the data path, link the device nodes that need to collaborate or share data to create a hyperedge, compile the hyperedge into a structured set, and generate a test hyperedge set; The network structure optimization module is used to measure the bandwidth utilization and delay between each device node in the device node set to calculate the data transmission cost, summarize the task requirements assigned to each of the device nodes, and determine the processing load according to the processing capacity and connection parameters of the test device data. According to the processing load, the hyperedges connecting the high-load device nodes are reallocated to the low-load device nodes to adjust the node connection. After each adjustment of the device node connection, the load of the device node is repeatedly calculated for iterative adjustment, seeking a configuration that minimizes cross-device data transmission and balances the processing load, and obtains an optimized network structure.

2. The semiconductor packaging test optimization system according to claim 1, characterized in that: Also includes: A collaborative scheduling module, based on the optimized network structure, monitors device load and task priority in real time, checks device availability and workload to compare current processing capabilities, reviews task deadlines and importance levels, compares task urgency, adjusts task allocation by reallocating tasks from overloaded devices to devices with idle capacity, rearranges the order of tasks according to priority to adjust the execution sequence, coordinates start and end time synchronization between devices to avoid conflicts and idle periods, and obtains a dynamic scheduling solution; The feedback adjustment module, based on the dynamic scheduling scheme, records the processing time and resource utilization during the test to collect performance data, aligns similar tasks and devices to compare historical performance indicators and current test results, adjusts processing capabilities and connection parameters according to performance deviations, updates device parameters, and modifies node connections and hyperedges to update network configurations at locations where bottlenecks are detected.

3. The semiconductor packaging test optimization system according to claim 2, characterized in that: The device node set includes a unique identification code, processing capability indicator, connection parameters and bandwidth for each device. The test hyperedge set includes hyperedges formed based on data flow requirements, task dependencies and shared resource information between devices. Each hyperedge represents a group of devices that need to collaborate or share data. The dynamic scheduling scheme includes task allocation adjusted according to current monitoring, task execution sequence re-ordered by priority, and a schedule for synchronous scheduling between devices.

4. The semiconductor packaging test optimization system according to claim 1, characterized in that: The device node generation module includes: A data collection submodule is used to collect packaging test equipment data and test site information, obtain the model, serial number and location of the packaging test equipment, select the unique identification code of each packaging test equipment, extract the equipment identifier, check the data integrity, remove duplicate and erroneous data, organize and store in a structured format, and generate a set of equipment identifiers; A parameter calculation submodule, based on the device identifier set, compares the processing speed and memory capacity of the packaging and testing device, calculates the processing capacity, measures the network interface type and bandwidth, obtains the connection parameters, records the input and output port information of the packaging and testing device, summarizes each parameter, and generates a device parameter set; The node generation submodule, based on the device parameter set, compiles each parameter into a structured data entry representing a single device, establishes device node attributes, including identifiers, processing capabilities and connection parameters, integrates each device information, and generates a device node set.

5. The semiconductor packaging test optimization system according to claim 1, characterized in that: The test hyperedge establishment module includes: A task collection submodule collects test tasks based on the device node set, retrieves associated requirements, obtains task names, input and output parameters and required resources, and generates a test task set; The requirement analysis submodule analyzes the data input and output of each task based on the test task set, determines the data flow requirements, compares the task sequence and shared resources, extracts the task dependency, and generates a task dependency set; The hyperedge generation submodule maps the data flow path between devices based on the task dependency set, determines the data path, links the nodes that need to collaborate or share data, creates hyperedges, compiles the hyperedges into a structured set, and generates a test hyperedge set.

6. The semiconductor packaging test optimization system according to claim 1, characterized in that: The network structure optimization module includes: A cost calculation submodule, based on the device node set and the test hyperedge set, measures the bandwidth usage and delay between nodes, calculates the data transmission cost, and generates a transmission cost set; A load analysis submodule, based on the transmission cost set, summarizes the task requirements assigned to each node, determines the processing load, analyzes the internal communication of the device cluster, compares the internal node interactions, and generates a node load set; The structure adjustment submodule reallocates the hyperedges connecting high-load nodes to nodes with lower loads based on the node load set, adjusts the node connections, changes the data flow path, reduces the transmission cost, repeatedly calculates the cost and load after each adjustment, performs iterative adjustment, and obtains an optimized network structure.

7. The semiconductor packaging test optimization system according to claim 2, characterized in that: The collaborative scheduling module includes: A real-time monitoring submodule, based on the optimized network structure, monitors device load and task priority in real time, checks device availability and workload, compares current processing capabilities, and generates a real-time status set; a task adjustment submodule, based on the real-time status set, reviews task deadlines and importance levels, compares task urgency, reallocates tasks from overloaded devices to devices with spare capacity, adjusts task allocation, rearranges the order of tasks according to priority, adjusts the execution sequence, and generates an adjusted task sequence; The time coordination submodule coordinates the start and end times based on the adjusted task sequence, synchronizes the scheduling between devices, avoids conflicts and idle periods, and obtains a dynamic scheduling solution.

8. The semiconductor packaging test optimization system according to claim 2, characterized in that: The feedback adjustment module comprises: A performance data collection submodule records processing time and resource utilization during the test based on the dynamic scheduling scheme, collects performance data, and generates a performance data set; A parameter update submodule, based on the performance data set, aligns similar tasks and devices, compares historical performance indicators and current test results, adjusts processing capabilities and connection parameters according to performance deviations, updates device parameters, and generates an updated device parameter set; The network configuration submodule modifies the node connections and hyperedges and updates the network configuration at the location where the bottleneck is detected based on the updated device parameter set.

9. A semiconductor packaging test optimization method, characterized in that: The semiconductor packaging test optimization system according to any one of claims 1 to 8 is implemented, comprising the following steps: Based on the packaged test equipment data and test site information, the unique identification code of each device is extracted, and the network interface and bandwidth are measured to obtain the connection parameters. At the same time, the processing speed and memory capacity are compared to determine the processing capacity of each device. The device data is structured and encoded to generate a device node set for subsequent data path analysis and load balancing; node integration data is obtained; According to the node integration data, test tasks are collected and the data input and output of each task are analyzed to determine the data flow requirements, task dependencies are extracted by comparing task sequences and shared resources, data flow paths between devices are mapped and nodes that need to collaborate or share data are linked to create hyperedges, and a structured test hyperedge set is generated; and a test task dependency structure is obtained; According to the test task dependency structure, measure the bandwidth usage and latency between nodes, calculate the data transmission cost, summarize the task requirements of each node to determine the processing load, analyze the internal communication of the device cluster and adjust the node connection, change the data flow path by reallocating the hyperedges connecting the high-load nodes to the nodes with lower loads, and repeat this process to minimize the cost and balance the load; obtain the optimized network configuration; According to the optimized network configuration, the device load and task priority are monitored in real time, the task allocation is adjusted according to the current processing capacity of the device and the urgency of the task, the scheduling between devices is synchronized to avoid conflicts and idle periods, performance data analysis is performed based on the processing time and resource utilization collected during the test, node connections and hyperedges are adjusted, and the test performance is optimized.

Images