Computer expansion card and computer system

Abstract

The invention discloses a computer expansion card and a computer system, which are used in the field of computer data processing, and the computer expansion card comprises an expansion card printed circuit board, a network communication module, a bus interface module, an intelligent analysis module and an expansion card memory, the expansion card printed circuit board is used for installing the network communication module, the bus interface module, the intelligent analysis module and the expansion card memory and providing communication connection among the network communication module, the bus interface module, the intelligent analysis module and the expansion card memory; the network communication module is used for data exchange and communication between a communication protocol and an external network; the bus interface module is used for connecting the expansion card to a mainboard of a computer system through a standard bus protocol; the intelligent analysis module is used for optimizing the performance of the computer system by utilizing parallel computing, task unloading, algorithm loading and data caching mechanisms; and the expansion card memory is used for storing data. According to the method, the data repeated reading time is shortened, and the data access speed is optimized.

Description

Technical Field

[0001] The present invention relates to the field of computer data processing, and in particular to a computer expansion card and a computer system. Background Art

[0002] A computer expansion card, also known as an expansion board or adapter card, is a circuit board that is inserted into the expansion slot of a computer motherboard. It is used to add additional functions to the computer or enhance existing functions. By inserting different expansion cards, the computer can enhance graphics processing capabilities, network connections, sound output, data storage and other functions.

[0003] The expansion card is connected to the expansion slot on the motherboard of the computer system, and then communicates with the computer's internal bus system to achieve hardware-level function expansion. In summary, the computer expansion card is an important hardware component that enables the computer to flexibly expand functions and improve performance according to user needs.

[0004] The defects existing in the existing technology are mainly manifested in that a single-core CPU can easily become a performance bottleneck when processing a large number of single and repetitive computing tasks, resulting in limited operating efficiency of the entire computer system, inability to fully utilize the hardware potential, and affecting the processing speed and energy efficiency of large-scale computing tasks.

[0005] Currently, no effective solution has been proposed for the problems in the related technologies. Summary of the invention

[0006] In order to overcome the above problems, the present invention aims to propose a computer expansion card and a computer system, with the purpose of solving the defects existing in the prior art, which is mainly manifested in that a single-core CPU can easily become a performance bottleneck when processing a large number of single and repetitive computing tasks, resulting in limited operating efficiency of the entire computer system, unable to fully utilize the hardware potential, and affecting the processing speed and energy efficiency of large-scale computing tasks.

[0007] To this end, the specific technical solution adopted by the present invention is as follows:

[0008] According to one aspect of the present invention, there is provided a computer expansion card, the computer expansion card comprising an expansion card printed circuit board, a network communication module, a bus interface module, an intelligent analysis module and an expansion card memory;

[0009] The expansion card printed circuit board is used to install the network communication module, the bus interface module, the intelligent analysis module and the expansion card memory, and provides communication connection between the network communication module, the bus interface module, the intelligent analysis module and the expansion card memory;

[0010] Network communication module, used for data exchange and communication with external network through communication protocol;

[0011] A bus interface module for connecting an expansion card to a computer system's mainboard via a standard bus protocol;

[0012] Intelligent analysis module for optimizing the performance of computer systems using parallel computing, task offloading, algorithm loading, and data caching mechanisms;

[0013] Expansion card memory, used to store data or program code.

[0014] Optionally, the intelligent analysis module includes a task screening module, a task decomposition module, a subtask processing module and a data buffer management module;

[0015] A task screening module, used to detect computationally intensive or repetitive tasks in a computer system using a task scheduling algorithm, and screen tasks that meet preset requirements;

[0016] The task decomposition module is used to decompose the selected tasks into multiple subtasks using the task division algorithm;

[0017] The subtask processing module is used to calculate and process the subtask data based on the decision tree algorithm according to the characteristics and requirements of each subtask;

[0018] The data buffer management module is used to manage the data cache, give priority to accessing frequently used data and dynamically update the cache during the subtask processing.

[0019] Optionally, the task screening module detects computationally intensive or repetitive tasks in the computer system using the task scheduling algorithm and screens tasks that meet preset requirements, including:

[0020] Continuously monitor the task load on computer systems to identify computationally intensive or repetitive tasks;

[0021] For identifying computationally intensive or repetitive tasks, extracting computational resource demand characteristic data;

[0022] Match the extracted computing resource demand feature data with the preset requirements, and filter the tasks that meet the preset conditions based on the matching results;

[0023] Perform a priority assessment on the screened tasks, sort them according to the assessment results, and make task selection decisions based on the results of the priority sorting to determine the tasks to be executed.

[0024] Optionally, matching the extracted computing resource demand feature data with preset requirements, and screening tasks that meet preset conditions based on the matching results include:

[0025] Obtaining data on computing resource requirements characteristics extracted from computing-intensive or repetitive tasks;

[0026] Preset the requirements for computing resource demand characteristic data and establish matching rules;

[0027] According to the matching rules, the computing resource demand characteristic data is compared with the requirements of the preset computing resource demand characteristic data one by one, and the matching degree is recorded;

[0028] Set a matching threshold and filter out tasks with matching degrees higher than the matching threshold.

[0029] Optionally, the selected tasks are evaluated for priority, prioritized according to the evaluation results, and task selection decisions are made based on the priority ranking results to determine that the tasks to be performed include:

[0030] Define the evaluation metrics for priorities;

[0031] For each selected task, collect the computing resource requirement characteristic data of the task on each evaluation indicator;

[0032] Assign weights to each evaluation indicator; calculate the priority score of each task based on the collected computing resource demand feature data and indicator weights;

[0033] Sort the tasks by priority score from high to low to generate a priority queue.

[0034] Optionally, the expression for calculating the priority score of each task is:

[0035]

[0036] In the formula, n represents the total number of evaluation indicators;

[0037] x i，j Represents the characteristic data value of the task on the jth indicator;

[0038] w j represents the weight of the jth indicator;

[0039] S i represents the priority score of each task i;

[0040] j represents the jth evaluation metric.

[0041] Optionally, the task decomposition module decomposes the screened task into multiple subtasks using a task division algorithm, including:

[0042] Decompose the task into multiple subtasks based on heuristic division to ensure that each subtask is relatively independent;

[0043] Verify the generated subtasks to ensure that they meet the predetermined performance requirements when executed independently;

[0044] Evaluate the parallelism of subtasks and establish an effective communication mechanism for communication between subtasks to ensure data synchronization and coordinated execution.

[0045] Optionally, the subtask processing module includes, when calculating and processing the data of the subtask based on the decision tree algorithm according to the characteristics and requirements of each subtask:

[0046] Use resource demand feature data to build a decision tree model and define splitting criteria;

[0047] According to the splitting criterion and based on the branches of the decision tree model, the resource demand feature data is divided into different subsets, where the subsets correspond to the leaf nodes of the decision tree;

[0048] Train the decision tree model based on the training data set, optimize the split of each node layer by layer, and determine the output value or category of each leaf node;

[0049] Evaluate the performance of the decision tree using the validation dataset;

[0050] The trained decision tree model is applied to the resource requirement feature data of the subtask to perform classification or regression prediction, and the processing result of the decision tree model on the subtask data is output.

[0051] Optionally, the trained decision tree model is applied to the resource requirement feature data of the subtask to perform classification or regression prediction, and the processing result of the decision tree model on the subtask data is output, including:

[0052] Standardize the resource requirement characteristic data of subtasks;

[0053] Load the trained decision tree model;

[0054] Input the standardized resource demand feature data into the decision tree model, call the model's prediction function, and generate classification or regression results;

[0055] Extract the classification label or regression value from the output of the model as the final prediction result.

[0056] According to another aspect of the present invention, a computer system is provided, including a mainboard, a central processing unit, a main memory module, a display adapter, a memory, and a computer expansion card.

[0057] Compared with the prior art, the present application has the following beneficial effects: the present invention continuously monitors the system load through a task scheduling algorithm, identifies computationally intensive or repetitive tasks, extracts and matches their resource demand characteristic data, thereby screening out tasks that meet the requirements, generates a priority queue through priority evaluation, selects high-priority tasks for execution, optimizes resource allocation, reduces resource consumption of invalid tasks, and improves the overall performance of the system; further decomposes the tasks into relatively independent subtasks through a task partitioning algorithm, verifies and evaluates the parallelism of the subtasks, and sets up an effective communication mechanism to ensure data synchronization between subtasks. By reasonably decomposing the tasks, parallel processing is achieved, the execution speed of the tasks is accelerated, and the responsiveness of the system is improved; based on the resource requirements of each subtask, a decision tree algorithm is used for calculation and processing. By constructing and training a decision tree model, the data of the subtask is segmented, classified, or regressed, and finally a prediction result is generated and extracted; priority is given to accessing commonly used data and dynamically updating the cache to reduce the time for repeated data reading, optimize data access speed, and thereby improve the overall efficiency of subtask processing. BRIEF DESCRIPTION OF THE DRAWINGS

[0058] The above characteristics, features and advantages of the present invention and their implementation methods and methods will become more clearly understood in conjunction with the following description of the embodiments, which are described in detail in conjunction with the accompanying drawings. Herein, a schematic diagram is shown:

[0059] Figure 1 is a principle block diagram of a computer expansion card according to an embodiment of the present invention;

[0060] Figure 2 The invention is a principle block diagram of an intelligent analysis module in a computer expansion card according to an embodiment of the invention.

[0061] In the figure:

[0062] 1. Expansion card printed circuit board; 2. Network communication module; 3. Bus interface module; 4. Intelligent analysis module; 401. Task screening module; 402. Task disassembly module; 403. Subtask processing module; 404. Data buffer management module; 5. Expansion card memory. DETAILED DESCRIPTION

[0063] In order to enable those skilled in the art to better understand the solution of the present application, the technical solution in the embodiments of the present application will be clearly and completely described below in conjunction with the drawings in the embodiments of the present application. Obviously, the described embodiments are only part of the embodiments of the present application, not all of the embodiments. Based on the embodiments in the present application, all other embodiments obtained by ordinary technicians in this field without creative work are within the scope of protection of this application.

[0064] According to an embodiment of the present invention, a computer expansion card and a computer system are provided.

[0065] The present invention is further described with reference to the accompanying drawings and specific embodiments. Figure 1-Figure 2 As shown, according to one embodiment of the present invention, a computer expansion card is provided, the computer expansion card comprising an expansion card printed circuit board 1, a network communication module 2, a bus interface module 3, an intelligent analysis module 4 and an expansion card memory 5;

[0066] The expansion card printed circuit board 1 is used to install the network communication module, the bus interface module, the intelligent analysis module and the expansion card memory, and provide communication connection between the network communication module, the bus interface module, the intelligent analysis module and the expansion card memory;

[0067] Network communication module 2, used for data exchange and communication with external network through communication protocol;

[0068] A bus interface module 3, used to connect the expansion card to the mainboard of the computer system through a standard bus protocol;

[0069] Intelligent analysis module 4, used to optimize the performance of the computer system by utilizing parallel computing, task offloading, algorithm loading and data caching mechanisms;

[0070] The expansion card memory 5 is used to store data or program codes.

[0071] It needs to be explained that the expansion card printed circuit board (PCB) 1 is a circuit board used to install and connect various modules. There are various modules on it, including network communication module, bus interface module, intelligent analysis module and expansion card memory. The printed circuit board provides communication connections between these modules to ensure that they can exchange data with each other and complete their respective tasks.

[0072] Network communication module 2: This module is used to realize data exchange and communication with the external network. It will connect the expansion card to the external network according to a specific communication protocol, so that data can be obtained from the network or sent to the network, realizing the network communication function.

[0073] Bus interface module 3: This module is used to connect the expansion card to the motherboard of the computer system. It connects to the motherboard through a standard bus protocol (such as PCIe, USB, etc.), so that the expansion card can transfer data with other components of the computer and achieve efficient exchange of data and instructions.

[0074] Intelligent analysis module 4: This is the core functional module of the expansion card, which is used to optimize the performance of the computer system. It can accelerate task processing through parallel computing, distribute the computing load to other resources through task offloading, dynamically load and update algorithms through algorithm loading, and use data caching mechanism to improve data access efficiency. These functions help the system handle computing-intensive tasks and data management more efficiently.

[0075] Expansion card memory 5: This module is used to store data or program code. It can store the algorithms required by the intelligent analysis module, the data to be processed, and the calculation results, etc., thereby reducing the system's dependence on external storage devices and improving data access speed and system response efficiency.

[0076] Preferably, the intelligent analysis module 4 includes a task screening module 401, a task decomposition module 402, a subtask processing module 403 and a data buffer management module 404;

[0077] The task screening module 401 is used to detect computationally intensive or repetitive tasks in the computer system using a task scheduling algorithm, and screen tasks that meet preset requirements.

[0078] Preferably, the task screening module 401 detects computationally intensive or repetitive tasks in the computer system using the task scheduling algorithm and screens tasks that meet preset requirements, including:

[0079] Continuously monitor the task load on computer systems to identify computationally intensive or repetitive tasks;

[0080] For identifying computationally intensive or repetitive tasks, extracting computational resource demand characteristic data;

[0081] Match the extracted computing resource demand feature data with the preset requirements, and filter the tasks that meet the preset conditions based on the matching results;

[0082] Perform a priority assessment on the screened tasks, sort them according to the assessment results, and make task selection decisions based on the results of the priority sorting to determine the tasks to be executed.

[0083] Preferably, matching the extracted computing resource demand characteristic data with preset requirements, and screening tasks that meet preset conditions based on the matching results includes:

[0084] Obtaining data on computing resource requirements characteristics extracted from computing-intensive or repetitive tasks;

[0085] Preset the requirements for computing resource demand characteristic data and establish matching rules;

[0086] According to the matching rules, the computing resource demand characteristic data is compared with the requirements of the preset computing resource demand characteristic data one by one, and the matching degree is recorded;

[0087] Set a matching threshold and filter out tasks with matching degrees higher than the matching threshold.

[0088] Preferably, priority evaluation is performed on the screened tasks, priority is sorted according to the evaluation results, and task selection decisions are made based on the priority sorting results. The tasks to be performed include:

[0089] Define the evaluation metrics for priorities;

[0090] For each selected task, collect the computing resource requirement characteristic data of the task on each evaluation indicator;

[0091] Assign weights to each evaluation indicator; calculate the priority score of each task based on the collected computing resource demand feature data and indicator weights;

[0092] Sort the tasks by priority score from high to low to generate a priority queue.

[0093] Preferably, the expression for calculating the priority score of each task is:

[0094]

[0095] In the formula, n represents the total number of evaluation indicators;

[0096] x i，j Represents the characteristic data value of the task on the jth indicator;

[0097] w j represents the weight of the jth indicator;

[0098] S i represents the priority score of each task i;

[0099] j represents the jth evaluation metric.

[0100] It should be explained that the task screening module monitors the system load through the task scheduling algorithm, identifies computationally intensive or repetitive tasks, extracts and matches resource demand feature data to screen out tasks that meet the requirements. Then, the screened tasks are evaluated for priority, a priority queue is generated based on weights and scores, and the tasks that need to be executed first are determined, thereby optimizing resource allocation and system performance.

[0101] The task decomposition module 402 is used to decompose the screened task into multiple subtasks using a task division algorithm.

[0102] Preferably, the task decomposition module 402 comprises:

[0103] Decompose the task into multiple subtasks based on heuristic division to ensure that each subtask is relatively independent;

[0104] Verify the generated subtasks to ensure that they meet the predetermined performance requirements when executed independently;

[0105] Evaluate the parallelism of subtasks and establish an effective communication mechanism for communication between subtasks to ensure data synchronization and coordinated execution.

[0106] It should be explained that the task decomposition module decomposes the selected complex tasks into multiple subtasks through the task division algorithm to ensure the relative independence between the subtasks. Then, the generated subtasks are verified to ensure that they meet the performance requirements when executed independently and their parallelism is evaluated. At the same time, an effective communication mechanism is established between subtasks to ensure data synchronization and coordinated execution, thereby achieving efficient parallel processing.

[0107] The subtask processing module 403 is used to perform calculation processing on the subtask data based on the decision tree algorithm according to the characteristics and requirements of each subtask.

[0108] Preferably, the subtask processing module 403 includes the following steps when calculating and processing the subtask data based on the decision tree algorithm according to the characteristics and requirements of each subtask:

[0109] Use resource demand feature data to build a decision tree model and define splitting criteria;

[0110] According to the splitting criterion and based on the branches of the decision tree model, the resource demand feature data is divided into different subsets, where the subsets correspond to the leaf nodes of the decision tree;

[0111] Train the decision tree model based on the training data set, optimize the split of each node layer by layer, and determine the output value or category of each leaf node;

[0112] Evaluate the performance of the decision tree using the validation dataset;

[0113] The trained decision tree model is applied to the resource requirement feature data of the subtask to perform classification or regression prediction, and the processing result of the decision tree model on the subtask data is output.

[0114] Preferably, the trained decision tree model is applied to the resource demand feature data of the subtask to perform classification or regression prediction, and the processing result of the decision tree model on the subtask data is output, including:

[0115] Standardize the resource requirement characteristic data of subtasks;

[0116] Load the trained decision tree model;

[0117] Input the standardized resource demand feature data into the decision tree model, call the model's prediction function, and generate classification or regression results;

[0118] Extract the classification label or regression value from the output of the model as the final prediction result.

[0119] It is important to explain that the decision tree model is constructed using the resource demand feature data, and the splitting criteria are defined to divide the data into different subsets. After the model is trained and verified, it is applied to the subtask data for classification or regression prediction. Finally, the subtask data is standardized and input into the trained decision tree model to generate and extract prediction results for subsequent execution or optimization of the subtask. This module is designed to efficiently and accurately meet the resource demand and computing requirements of different subtasks.

[0120] The data buffer management module 404 is used to manage the data cache, give priority to accessing frequently used data and dynamically update the cache during the subtask processing.

[0121] According to another embodiment of the present invention, a computer system is provided, including a mainboard, a central processing unit, a main memory module, a display adapter, a memory, and a computer expansion card.

[0122] In addition, it should be supplemented that the mainboard also includes an expansion card slot, and the bus interface module 3 of the computer expansion card is plugged into the expansion card slot. This is a prior art and will not be described in detail here.

[0123] To sum up, with the help of the above technical scheme of the present invention, the present invention continuously monitors the system load through the task scheduling algorithm, identifies computing-intensive or repetitive tasks, extracts and matches their resource demand characteristic data, thereby screening out tasks that meet the requirements, generates a priority queue through priority evaluation, selects high-priority tasks for execution, optimizes resource allocation, reduces resource consumption of invalid tasks, and improves the overall performance of the system; further decomposes into relatively independent subtasks through the task division algorithm, verifies and evaluates the parallelism of the subtasks, and sets an effective communication mechanism to ensure data synchronization between subtasks. By reasonably decomposing tasks, parallel processing is achieved, the execution speed of tasks is accelerated, and the responsiveness of the system is improved; based on the resource requirements of each subtask, a decision tree algorithm is used for calculation and processing. By constructing and training a decision tree model, the data of the subtask is segmented, classified or regressed, and finally the prediction results are generated and extracted; priority is given to accessing commonly used data and dynamically updating the cache to reduce the time for repeated data reading, optimize the data access speed, and thereby improve the overall efficiency of subtask processing.

[0124] Although the present invention has been disclosed as above with preferred embodiments, the embodiments are merely examples for the convenience of description and are not intended to limit the present invention. Those skilled in the art may make several changes and modifications without departing from the spirit and scope of the present invention. The scope of protection claimed by the present invention shall be based on the claims.

Claims

1. A computer expansion card, characterized in that: The computer expansion card includes an expansion card printed circuit board, a network communication module, a bus interface module, an intelligent analysis module and an expansion card memory; The expansion card printed circuit board is used to install the network communication module, the bus interface module, the intelligent analysis module and the expansion card memory, and provides communication connection between the network communication module, the bus interface module, the intelligent analysis module and the expansion card memory; The network communication module is used for data exchange and communication with the external network through the communication protocol; The bus interface module is used to connect the expansion card to the mainboard of the computer system through a standard bus protocol; The intelligent analysis module is used to optimize the performance of the computer system by utilizing parallel computing, task offloading, algorithm loading and data caching mechanisms; The expansion card memory is used to store data or program codes.

2. A computer expansion card according to claim 1, characterized in that: The intelligent analysis module includes a task screening module, a task disassembly module, a subtask processing module and a data buffer management module; The task screening module is used to detect computationally intensive or repetitive tasks in the computer system using a task scheduling algorithm, and screen tasks that meet preset requirements; The task decomposition module is used to decompose the screened task into multiple subtasks using a task division algorithm; The subtask processing module is used to calculate and process the subtask data based on the decision tree algorithm according to the characteristics and requirements of each subtask; The data buffer management module is used to manage the data cache, give priority to accessing frequently used data and dynamically update the cache during the subtask processing.

3. A computer expansion card according to claim 2, characterized in that: The task screening module detects computationally intensive or repetitive tasks in the computer system using the task scheduling algorithm and screens tasks that meet preset requirements, including: Continuously monitor the task load on computer systems to identify computationally intensive or repetitive tasks; For identifying computationally intensive or repetitive tasks, extracting computational resource demand characteristic data; Match the extracted computing resource demand feature data with the preset requirements, and filter the tasks that meet the preset conditions based on the matching results; Perform a priority assessment on the screened tasks, sort them according to the assessment results, and make task selection decisions based on the results of the priority sorting to determine the tasks to be executed.

4. A computer expansion card according to claim 3, characterized in that: The step of matching the extracted computing resource demand feature data with preset requirements and screening tasks that meet preset conditions based on the matching results includes: Obtaining data on computing resource requirements characteristics extracted from computing-intensive or repetitive tasks; Preset the requirements for computing resource demand characteristic data and establish matching rules; According to the matching rules, the computing resource demand characteristic data is compared with the requirements of the preset computing resource demand characteristic data one by one, and the matching degree is recorded; Set a matching threshold and filter out tasks with matching degrees higher than the matching threshold.

5. A computer expansion card according to claim 4, characterized in that: The screened tasks are evaluated for priority, prioritized according to the evaluation results, and task selection decisions are made based on the priority ranking results. The tasks to be performed include: Define the evaluation metrics for priorities; For each selected task, collect the computing resource requirement characteristic data of the task on each evaluation indicator; Assign weights to each evaluation indicator; calculate the priority score of each task based on the collected computing resource demand feature data and indicator weights; Sort the tasks by priority score from high to low to generate a priority queue.

6. A computer expansion card according to claim 5, characterized in that: The expression for calculating the priority score of each task is: In the formula, n represents the total number of evaluation indicators; x i，j Represents the characteristic data value of the task on the jth indicator; w j represents the weight of the jth indicator; S i represents the priority score of each task i; j represents the jth evaluation metric.

7. A computer expansion card according to claim 6, characterized in that: The task decomposition module decomposes the screened task into multiple subtasks using the task division algorithm, including: Decompose the task into multiple subtasks based on heuristic division to ensure that each subtask is relatively independent; Verify the generated subtasks to ensure that they meet the predetermined performance requirements when executed independently; Evaluate the parallelism of subtasks and establish an effective communication mechanism for communication between subtasks to ensure data synchronization and coordinated execution.

8. A computer expansion card according to claim 7, characterized in that: The subtask processing module includes the following steps when calculating and processing the subtask data based on the decision tree algorithm according to the characteristics and requirements of each subtask: Use resource demand feature data to build a decision tree model and define splitting criteria; According to the splitting criterion and based on the branches of the decision tree model, the resource demand feature data is divided into different subsets, where the subsets correspond to the leaf nodes of the decision tree; Train the decision tree model based on the training data set, optimize the split of each node layer by layer, and determine the output value or category of each leaf node; Evaluate the performance of the decision tree using the validation dataset; The trained decision tree model is applied to the resource requirement feature data of the subtask to perform classification or regression prediction, and the processing result of the decision tree model on the subtask data is output.

9. A computer expansion card according to claim 8, characterized in that: The method of applying the trained decision tree model to the resource requirement feature data of the subtask to perform classification or regression prediction, and outputting the processing result of the decision tree model on the subtask data includes: Standardize the resource requirement characteristic data of subtasks; Load the trained decision tree model; Input the standardized resource demand feature data into the decision tree model, call the model's prediction function, and generate classification or regression results; Extract the classification label or regression value from the output of the model as the final prediction result.

10. A computer system, characterized in that: The computer comprises a mainboard, a central processing unit, a main memory module, a display adapter, a memory and at least one computer expansion card as claimed in any one of claims 1 to 9.

Images