TWI704494B - Processor performance optimization method and motherboard using the same - Google Patents

Abstract

A processor performance optimization method and a motherboard using the same are provided. The processor performance optimization method includes the steps of: performing a basic input/output system to capture initial setting parameters corresponding to a processor; and comparing an optimized setting model stored in the basic input/output system according to the initial setting parameters; when the optimization setting model corresponds to the initial setting parameters, optimization setting parameters are obtained according to the optimization setting model, and when the optimization setting model does not correspond to the initial setting parameters, a neural network operation is performed to obtain the optimization setting parameters; and operate the processor according to the optimization setting parameters to reduce the operating power consumption of the processor.

Description

Performance optimization method of processor and main board using same

The present invention relates to a functional design of a motherboard (Motherboard), and particularly relates to a performance optimization method of a processor (Processor) and a motherboard using the same.

For a general computer system, when a user purchases a computer host and installs the processor on the motherboard (Motherboard), under normal conditions, the computer system can only use the pre-designed by the processor manufacturer Related initial processor setting parameters to operate the processor. However, if the user uses the adjustment function provided by the computer motherboard to manually modify the relevant parameters of the processor, it is often only the use of greater power consumption in exchange for the processor processing clock, so the processor performance cannot be improved efficiently , And at the same time reduce the operating power consumption of the processor, it can not even ensure that the processor is working in the most stable state. In view of this, the following will propose solutions in several embodiments.

The present invention provides a method for optimizing the performance of a processor and a motherboard using the same, which can effectively automatically optimize the execution performance of the processor provided on the motherboard to effectively reduce the operating power consumption of the processor.

The performance optimization method of the processor of the present invention is suitable for a motherboard with a basic input output system. The performance optimization method includes the following steps: executing a basic input output system to retrieve a plurality of initial setting parameters corresponding to the processor; comparing the plurality of initial setting parameters stored in the basic input output system according to the plurality of initial setting parameters An optimized setting model; when one of the plurality of optimized setting models corresponds to the plurality of initial setting parameters, obtaining a plurality of optimal settings according to one of the plurality of optimized setting models Optimization setting parameters, and when any one of the plurality of optimization setting models does not correspond to the plurality of initial setting parameters, a neural network-like operation is performed to obtain the plurality of optimization setting parameters And operating the processor according to the plurality of optimized setting parameters to reduce the operating power consumption of the processor.

In an embodiment of the present invention, the above-mentioned performance optimization method further includes the following steps: an artificial intelligence engine executes a neural network-like operation to generate the plurality of optimized setting models according to a plurality of initial setting parameter sets ; And writing the plurality of optimized setting models into the basic input output system.

In an embodiment of the present invention, when any one of the plurality of optimized setting models does not correspond to the plurality of initial setting parameters, a neural network-like operation is performed to obtain the plurality of parameters. The step of optimizing setting parameters includes: performing neural network-like operations by an artificial intelligence engine to train at least one new optimized setting model according to the plurality of initial setting parameters; The optimized setting model is written into the basic input output system; and the plurality of optimized setting parameters are obtained according to one of the at least one new optimized setting model.

In an embodiment of the present invention, the step of performing neural network-like operations by an artificial intelligence engine to train the at least one new optimized setting model according to the plurality of initial setting parameters includes: The multiple initial setting parameters are used to train multiple new optimized setting models corresponding to the same processor model and different processor operating frequencies.

In an embodiment of the present invention, the aforementioned processor includes a central processing unit, and the multiple optimized setting models correspond to the same central processing unit model and different central processing unit operating frequencies.

In an embodiment of the present invention, the aforementioned multiple initial setting parameters include the model of the central processing unit, the number of central processing unit cores, the operating frequency of the central processing unit, the operating voltage of the central processing unit, the operating current of the central processing unit, and the central processing unit. At least one of unit power, central processing unit temperature, and central processing unit load line. The weight values corresponding to the model of the central processing unit and the operating frequency of the central processing unit are higher than other initial setting parameters.

In an embodiment of the present invention, the above-mentioned multiple optimization setting parameters include optimizing central processing unit operating voltage, optimizing central processing unit operating current, optimizing central processing unit power, and optimizing At least one of the temperature of the central processing unit, the optimized operating frequency of the central processing unit, and the optimized load line of the central processing unit.

In an embodiment of the present invention, the aforementioned processor includes a central processing unit and a graphics processing unit. The multiple optimized setting models correspond to the same graphics processing unit model and different graphics processing unit operating frequencies.

In an embodiment of the present invention, the above-mentioned multiple initial setting parameters include a graphics processing unit model, a graphics processing unit preset operating frequency, a graphics processing unit operating voltage, a graphics processing unit parameter, a central processing unit operating voltage, and settings At least one of the operating frequencies of the display memory of the display memory in the graphics processing unit. The weight value corresponding to the graphics processing unit model is higher than other initial setting parameters.

In an embodiment of the present invention, the multiple optimization setting parameters described above include optimizing graphics processing unit operating voltage, optimizing graphics processing unit operating frequency, optimizing graphics processing unit parameters, and optimizing At least one of the operating voltages of the central processing unit.

The motherboard of the present invention includes a basic input output system. The basic input output system includes multiple optimized setting models. The basic input output system is used for capturing a plurality of initial setting parameters corresponding to the processor to compare the basic input output system optimization setting model. When one of the basic input output system optimized setting models corresponds to the initial setting parameters of the basic input output system, the basic input output system obtains multiple optimized settings according to one of the basic input output system optimized setting models parameter. When any one of the optimized setting models of the basic input output system does not correspond to the initial setting parameters of the basic input output system, a neural network-like operation is executed to obtain the optimized setting parameters of the basic input output system. The basic input output system operates the processor according to the optimized setting parameters of the basic input output system to reduce the operating power consumption of the processor.

In an embodiment of the present invention, the aforementioned artificial intelligence engine performs neural network-like operations to generate a basic input output system optimization model based on a plurality of initial setting parameter sets, and optimize the basic input output system The standardized setting model is written into the basic input output system.

In an embodiment of the present invention, the above-mentioned artificial intelligence engine performs neural network-like operations to train at least one new optimized setting model according to the initial setting parameters of the basic input output system. The artificial intelligence engine writes the at least one new optimized setting model into the basic input output system, so that the basic input output system obtains the basic input output system according to one of the at least one new optimized setting model Optimize setting parameters.

In an embodiment of the present invention, the aforementioned artificial intelligence engine trains a plurality of new optimized setting models corresponding to the same processor model and different processor operating frequencies according to the initial setting parameters of the basic input output system.

In an embodiment of the present invention, the aforementioned processor includes a central processing unit. The multiple optimized setting models correspond to the same central processing unit model and different central processing unit operating frequencies.

In an embodiment of the present invention, the aforementioned multiple initial setting parameters include the model of the central processing unit, the number of central processing unit cores, the operating frequency of the central processing unit, the operating voltage of the central processing unit, the operating current of the central processing unit, and the central processing unit. At least one of unit power, central processing unit temperature, and central processing unit load line. The weight values corresponding to the model of the central processing unit and the operating frequency of the central processing unit are higher than other initial setting parameters.

In an embodiment of the present invention, the above-mentioned multiple optimization setting parameters include optimizing central processing unit operating voltage, optimizing central processing unit operating current, optimizing central processing unit power, and optimizing At least one of the temperature of the central processing unit, the optimized operating frequency of the central processing unit, and the optimized load line of the central processing unit.

In an embodiment of the present invention, the aforementioned processor includes a central processing unit and a graphics processing unit. The multiple optimized setting models correspond to the same graphics processing unit model and different graphics processing unit operating frequencies.

In an embodiment of the present invention, the above-mentioned multiple initial setting parameters include a graphics processing unit model, a graphics processing unit preset operating frequency, a graphics processing unit operating voltage, a graphics processing unit parameter, a central processing unit operating voltage, and settings At least one of the operating frequencies of the display memory of the display memory in the graphics processing unit. The weight value corresponding to the graphics processing unit model is higher than other initial setting parameters.

In an embodiment of the present invention, the multiple optimization setting parameters described above include optimizing graphics processing unit operating voltage, optimizing graphics processing unit operating frequency, optimizing graphics processing unit parameters, and optimizing At least one of the operating voltages of the central processing unit.

Based on the above, the performance optimization method of the processor of the present invention and the motherboard using the same can compare multiple initial setting parameters of the central processing unit or graphics processing unit with multiple pre-trained neural network operations. The optimization model uses the corresponding optimization model to obtain optimized setting parameters to operate the central processing unit or graphics processing unit, so as to effectively optimize the performance of the central processing unit or graphics processing unit and reduce operating power consumption.

In order to make the above-mentioned features and advantages of the present invention more comprehensible, the following specific embodiments are described in detail in conjunction with the accompanying drawings.

In order to make the content of the present invention more comprehensible, the following embodiments are specifically cited as examples on which the present invention can indeed be implemented. In addition, wherever possible, elements/components/steps with the same reference numbers in the drawings and embodiments represent the same or similar components.

FIG. 1 is a schematic structural diagram of a motherboard according to an embodiment of the invention. 1, a motherboard (Motherboard) 100 includes a basic input/output system (Basic Input/Output System, BIOS) 110 and a central processing unit (Center Processing Unit, CPU) 120. The basic input output system 110 is installed or embedded on the circuit board of the main board 100, and an optimized setting module 111 is stored. The central processing unit 120 is pluggably disposed on the motherboard 100 through a plurality of corresponding slots (Slots) of the motherboard 100. In this embodiment, when the central processing unit 120 is set on the motherboard 100 and the computer system executes the boot procedure, the basic input output system 110 will execute the optimized setting module 111 to obtain information about The relevant optimized operating parameters of the central processing unit 120 are used to optimize the processing performance of the central processing unit 120.

Specifically, the basic input output system 110 captures a plurality of initial setting parameters corresponding to the central processing unit 120 to obtain a corresponding optimized setting model according to the plurality of initial setting parameters. The basic input output system 110 can obtain a plurality of optimized setting parameters according to the corresponding optimized setting model, and operate the central processing unit 120 according to the plurality of optimized setting parameters, so as to reduce the operation of the central processing unit 120 Power consumption. For example, the multiple optimized setting models may correspond to the same central processing unit model and different central processing unit operating frequencies. In other words, the basic input output system 110 pre-establishes one or more sets of multiple optimization setting models corresponding to the same central processing unit model and different central processing unit operating frequencies for the optimization setting module 111 to compare.

It is worth noting that the optimized setting module 111 of this embodiment can be pre-written or burned into the basic input output system 110 by the product manufacturer during the manufacturing process of the motherboard 100, so that the user can set any The central processing unit 120 of the model or specific operating frequency is installed on the motherboard 100, and the computer system can automatically execute the optimized setting module 111 through the basic input output system 110 during the boot process.

Furthermore, the basic input output system 110 may pre-store the plurality of optimized setting models, and the plurality of optimized setting models may be obtained by the manufacturer or the user based on the previous model training history. Set up a set of initial setting parameters in advance. In detail, each of the plurality of initial setting parameter groups means that a plurality of central processing units of different central processing unit models or different central processing unit operating frequencies are respectively installed on the motherboard by the manufacturer or user. 100, the data collected by the basic input output system 110 successively. In addition, each of the plurality of initial setting parameter groups can obtain the corresponding optimized setting models through neural network operations one by one, and the plurality of optimized setting models will be Write or burn into the basic input output system 110 one by one. In other words, the BIOS 110 can create an optimized setting database in advance.

Therefore, when the user installs the new central processing unit 120 on the motherboard 100 for the first time, and the basic input output system 110 is activated by the computer system, the basic input output system 110 will be based on the current corresponding to the central processing unit 120 The obtained initial setting parameters are compared with the optimized setting models stored in the basic input output system 110 to determine whether the same central processing unit model or the same central processing unit operation has been installed previously The frequency central processing unit can directly read the relevant optimization setting parameters that have been previously obtained through neural network calculations to provide fast optimization effects. In this embodiment, when one of the plurality of optimized setting models corresponds to the plurality of initial setting parameters, the basic input output system 110 according to one of the plurality of optimized setting models Get multiple optimized setting parameters. In contrast, when any one of the plurality of optimized setting models does not correspond to the plurality of initial setting parameters, a neural network-like operation will be executed to obtain the plurality of optimized setting parameters . In other words, even if there is no corresponding optimized setting model in the optimized setting database of the basic input output system 110, so that the corresponding optimized setting parameters cannot be provided, the motherboard 100 of this embodiment can use real-time analog The neural network calculates to obtain a new optimized setting model and the corresponding plurality of optimized setting parameters.

In this embodiment, the above-mentioned neural network-like operation is executed by an artificial intelligence engine (Artificial Intelligence Engine, AI Engine). The artificial intelligence engine can be, for example, a hardware architecture composed of multiple computing units through design or a specially designed algorithm (Algorithm) to implement a computing engine capable of performing related machine learning functions. In addition, the artificial intelligence engine can have the following multiple implementation modes. For example, in one embodiment, the artificial intelligence engine may be provided in the basic input output system 110 to execute the artificial intelligence engine through the basic input output system 110. In another embodiment, the artificial intelligence engine is an application (Application) to execute the artificial intelligence engine through an operating system (OS). In another embodiment, the artificial intelligence engine is set in a cloud system to execute the artificial intelligence engine through communication with the cloud system.

In other words, when any one of the plurality of optimized setting models does not correspond to the plurality of initial setting parameters, in the above-mentioned embodiment, the basic input output system 110 can execute the artificial intelligence engine in real time to Obtain the plurality of optimized setting parameters. In another embodiment described above, after the basic input output system 110 completes the boot process, the artificial intelligence engine can be executed by the operating system of the computer system to obtain the plurality of optimized setting parameters. In another embodiment described above, after the basic input output system 110 completes the boot process, the computer system can communicate with the cloud system through a wired or wireless communication module to execute the artificial intelligence engine and obtain the plurality of Optimize setting parameters. In this regard, the setting method of the artificial intelligence engine can be determined according to different computing requirements or optimized design.

Moreover, in each of the foregoing embodiments, the artificial intelligence engine executes the neural network-like operation to train at least one new optimized setting model according to the plurality of initial setting parameters. Moreover, the artificial intelligence engine can write the at least one new optimized setting model into the basic input output system 110. However, in certain embodiments, the artificial intelligence engine can train multiple new optimized setting models corresponding to the same central processing unit model and different central processing unit operating frequencies according to the multiple initial setting parameters, and combine all The multiple new optimized setting models are written into the basic input output system 110 in a replication manner. For example, the basic input output system 110 can store a fixed number or a limited number of optimized configuration models. Therefore, when the storage space or number is full, a new optimized configuration model can be written to cover the less frequently used models. Other optimized setting models.

2 is a flowchart of a method for optimizing the performance of a central processing unit according to an embodiment of the invention. 1 and FIG. 2, the motherboard 100 of the embodiment of FIG. 1 can perform steps S201 to S211 of the embodiment of FIG. 2. In step S201, the artificial intelligence engine can perform neural network-like operations to generate a plurality of optimized setting models according to a plurality of initial setting parameter sets related to the central processing unit 120. In step S202, the artificial intelligence engine writes the multiple optimized setting models into the basic input output system 110. In step S203, the motherboard 100 executes the booting procedure and starts the basic input output system 110. In step S204, the BIOS 110 can execute the optimization setting module 111. In step S205, the basic input output system 110 retrieves a plurality of initial setting parameters corresponding to the central processing unit 120. In step S206, the optimized setting module 111 compares the plurality of optimized setting models stored in the basic input output system 110 according to the plurality of initial setting parameters. In step S207, the optimization setting module 111 determines whether one of the plurality of optimization setting models corresponds to the plurality of initial setting parameters. When one of the plurality of optimized setting models corresponds to the plurality of initial setting parameters, in step S208, the basic input output system 110 according to one of the plurality of optimized setting models A plurality of optimized setting parameters operate the central processing unit 120. In step S209, when the optimized setting is completed, the BIOS 110 ends the execution of the optimized setting module 111. In contrast, when any one of the plurality of optimized setting models does not correspond to the plurality of initial setting parameters, in step S210, the artificial intelligence engine will be run to perform neural network-like operations to Training at least one new optimized setting model according to the plurality of initial setting parameters. In step S211, the artificial intelligence engine writes the at least one new optimized setting model into the basic input output system 110, and obtains the multiple models according to one of the at least one new optimized setting model. Optimized setting parameters. Therefore, the performance optimization method of this embodiment enables the motherboard 100 to provide an effective automatic performance optimization function of the central processing unit 120.

In addition, regarding the performance optimization method described in this embodiment and other related component features, technical solutions, and implementation details of the motherboard 100, you can refer to the description of the embodiment in FIG. 1 to obtain sufficient teaching, suggestion, and implementation description. Repeat it again.

Fig. 3 is a schematic diagram of a neural network-like operation according to an embodiment of the invention. Referring to FIGS. 1 and 3, the neural network-like operation for optimizing the central processing unit 120 described in the various embodiments of the present invention may be a neural network-like network 300 as shown in FIG. 3. In this embodiment, the basic input output system 110 can retrieve multiple initial setting parameters corresponding to the central processing unit 120. For example, the artificial intelligence model may use the plurality of initial setting parameters as the plurality of input parameters 310_1 to 310_6 and input them to the input layer of the neural network 300. Then, after a plurality of operation neurons (Neurons) 320_1~320_8 in the hidden layer of the neural network 300 are calculated, the output layer of the neural network 300 can generate a plurality of output parameters 330_1~ 330_6. Therefore, the multiple output parameters 330_1 to 330_6 obtained by each operation. The multiple output parameters 330_1 to 330_6 are multiple optimized setting parameters, and the multiple output parameters 330_1 to 330_6 can establish an optimized setting model. In other words, after multiple inputs of different input parameters 310_1 to 310_6 to perform neural network-like operations, the artificial intelligence model can obtain multiple optimized setting models 340_1 to 340_N, and N is a positive integer greater than 1.

In this embodiment, the multiple initial setting parameters may include, for example, the model of the central processing unit, the number of central processing unit cores, the operating frequency of the central processing unit, the operating voltage of the central processing unit, the operating current of the central processing unit, the power of the central processing unit, At least one of the temperature of the central processing unit and the load line of the central processing unit. In addition, the plurality of optimized setting parameters may include, for example, optimized central processing unit operating voltage, optimized central processing unit operating current, optimized central processing unit power, optimized central processing unit temperature, and optimized At least one of the load lines of the central processing unit. In this embodiment, the plurality of computing neurons 320_1~320_8 may be used to respectively perform weighting and accumulation operations according to the plurality of input parameters 310_1~310_6 to obtain the plurality of output parameters 330_1~330_6 . However, the parameter operation modes corresponding to the plurality of operation neurons 320_1 to 320_8 can be determined according to the types of different types of neural networks, and the present invention is not limited.

In a specific embodiment, the weight values corresponding to the above-mentioned central processing unit model and operating frequency of the central processing unit are higher than other initial setting parameters. For example, the weight value corresponding to the model of the central processing unit and the operating frequency of the central processing unit may be 0.2, and other initial setting parameters may be 0.1. However, in one embodiment, the weight values corresponding to the multiple initial setting parameters can also be adjusted correspondingly according to special use requirements, special setting considerations or special operating environment. The weight value corresponding to the initial setting parameter of is adjusted to a higher weight value, and the weight value corresponding to the initial setting parameter of a lower consideration factor is adjusted to a lower weight value.

It is worth noting that the number of the plurality of initial setting parameters of the central processing unit 120 of the present invention is not limited to the number of the plurality of input parameters 310_1 to 310_6 in FIG. 3, and the number of the central processing unit 120 of the present invention The number of the multiple optimization setting parameters is not limited to the number of the multiple output parameters 330_1 to 330_6 in FIG. 3. In addition, the number of computing neurons 320_1 to 320_8 in FIG. 3 is only used for illustration, and the present invention is not limited thereto. Moreover, the neural network 300 in FIG. 3 is only an example of the present invention. The neural network-like operation described in each embodiment of the present invention may be, for example, a deep neural network (DNN) operation, a convolutional neural network (CNN) operation, or a recurrent neural network (Recurrent Neural Network) operation. Neural Networks, RNN) computing and other machine learning computing models.

FIG. 4 is a schematic structural diagram of a motherboard according to another embodiment of the invention. 4, the motherboard 400 includes a basic input output system 410, a central processing unit 420, and a graphics processing unit (GPU) 430. The basic input output system 410 is installed or embedded on the circuit board of the motherboard 400, and an optimized setting module 411 is stored. The central processing unit 420 and the graphics processing unit 430 are pluggably disposed on the motherboard 400 through a plurality of corresponding slots (Slots) of the motherboard 400. In this embodiment, when the central processing unit 420 is set on the motherboard 400 and the computer system executes the boot procedure, the basic input output system 410 will execute the optimized setting module 411 to obtain information about The relevant optimized operating parameters of the central processing unit 420 and the graphics processing unit 430 are used to optimize the processing performance of the graphics processing unit 430.

In this embodiment, the graphics processing unit 430 refers to a processor provided on a display card, and the display card is further configured with a display memory. The display memory may be a random access memory (Random Access Memory, RAM), which is, for example, a Double Data Rate Synchronous Dynamic Random Access Memory (DDR SDRAM). The invention is not limited to this. Moreover, the display memory including the corresponding display memory operating frequency can be one of the multiple initial setting parameters of this embodiment. In addition, compared with the embodiment in FIG. 1, the optimization of the processing performance of the graphics processing unit 430 in this embodiment requires consideration of multiple related initial setting parameters of the central processing unit 420 and the graphics processing unit 430 at the same time.

Specifically, the basic input output system 410 captures a plurality of initial setting parameters corresponding to the central processing unit 420 and the graphics processing unit 430 to obtain the corresponding optimized setting model according to the plurality of initial setting parameters. The basic input output system 410 can obtain a plurality of optimized setting parameters according to the corresponding optimized setting model, and operate the central processing unit 420 and the graphics processing unit 430 according to the plurality of optimized setting parameters to reduce graphics The operating power consumption of the processing unit 430. For example, the multiple optimized setting models may correspond to the same graphics processing unit model and different graphics processing unit operating frequencies. In other words, the basic input output system 410 pre-establishes one or more sets of multiple optimized setting models corresponding to the same graphics processing unit model and different graphics processing unit operating frequencies for the optimization setting module 411 to compare.

It is worth noting that the optimized setting module 411 of this embodiment can be pre-written or burned into the basic input output system 410 by the product manufacturer during the manufacturing process of the motherboard 400, so that the user can set any The graphics processing unit 430 of the model or specific operating frequency is installed on the motherboard 400, and the computer system can automatically execute the optimization setting module 411 through the basic input output system 410 during the boot process.

Further, the basic input output system 410 may pre-store the plurality of optimized setting models, and the plurality of optimized setting models may be obtained by the manufacturer or the user based on the previous model training history. Set up a set of initial setting parameters in advance. In detail, each of the plurality of initial setting parameter groups refers to the successive installation of a plurality of graphics processing units of different graphics processing unit models or different graphics processing unit operating frequencies on the motherboard by the manufacturer or user. 400, the data collected by the basic input output system 410 successively. In addition, each of the plurality of initial setting parameter groups can obtain the corresponding optimized setting models through neural network operations one by one, and the plurality of optimized setting models will be Write or burn into the basic input output system 410 one by one. In other words, the basic input output system 410 can create an optimized setting database in advance.

Therefore, when the user installs the new graphics processing unit 430 on the motherboard 100 for the first time, and the basic input output system 410 is activated by the computer system, the basic input output system 410 will be based on the current graphics processing unit 430 The obtained initial setting parameters are compared with the optimized setting models stored in the basic input output system 410 to determine whether the same graphics processing unit model or the same graphics processing unit operation has been installed previously The frequency graphics processing unit can directly read the relevant optimization setting parameters obtained through neural network-like calculations in advance to provide fast optimization effects. In this embodiment, when one of the plurality of optimized setting models corresponds to the plurality of initial setting parameters, the basic input output system 410 according to one of the plurality of optimized setting models Get multiple optimized setting parameters. In contrast, when any one of the plurality of optimized setting models does not correspond to the plurality of initial setting parameters, a neural network-like operation will be executed to obtain the plurality of optimized setting parameters . In other words, even if there is no corresponding optimized setting model in the optimized setting database of the basic input output system 410, so that the corresponding optimized setting parameters cannot be provided, the motherboard 400 of this embodiment can use real-time analog The neural network calculates to obtain a new optimized setting model and the corresponding plurality of optimized setting parameters.

In this embodiment, the above-mentioned neural network-like operations are performed by an artificial intelligence engine. The artificial intelligence engine may be, for example, a hardware architecture composed of multiple computing units through design or a specially designed algorithm to implement a computing engine capable of performing related machine learning functions. In addition, the artificial intelligence engine can have the following multiple implementation modes. For example, in one embodiment, the artificial intelligence engine may be provided in the basic input output system 410 to execute the artificial intelligence engine through the basic input output system 410. In another embodiment, the artificial intelligence engine is an application program to execute the artificial intelligence engine through the operating system. In yet another embodiment, the artificial intelligence engine is set in the cloud system to execute the artificial intelligence engine through communication with the cloud system.

In other words, when any one of the plurality of optimized setting models does not correspond to the plurality of initial setting parameters, in the above-mentioned embodiment, the basic input output system 410 can execute the artificial intelligence engine in real time to Obtain the plurality of optimized setting parameters. In another embodiment described above, after the basic input output system 410 completes the boot process, the artificial intelligence engine can be executed by the operating system of the computer system to obtain the plurality of optimized setting parameters. In another embodiment described above, after the basic input output system 410 completes the boot process, the computer system can communicate with the cloud system through a wired or wireless communication module to execute the artificial intelligence engine and obtain the plurality of Optimize setting parameters. In this regard, the setting method of the artificial intelligence engine can be determined according to different computing requirements or optimized design.

Moreover, in each of the foregoing embodiments, the artificial intelligence engine executes the neural network-like operation to train at least one new optimized setting model according to the plurality of initial setting parameters. And, the artificial intelligence engine can write the at least one new optimized setting model into the basic input output system 410. However, in certain embodiments, the artificial intelligence engine can train multiple new optimized setting models corresponding to the same graphics processing unit model and different graphics processing unit operating frequencies according to the multiple initial setting parameters, and combine all The multiple new optimized setting models are written into the basic input output system 410 in a replication manner. For example, the basic input output system 410 can store a fixed number or a limited number of optimized configuration models. Therefore, when the storage space or number is full, a new optimized configuration model can be written to cover the less frequently used models. Other optimized setting models.

FIG. 5 is a flowchart of a performance optimization method of a graphics processing unit according to an embodiment of the invention. Referring to FIG. 4 and FIG. 5, the motherboard 400 of the embodiment of FIG. 4 can perform steps S501 to S511 of the embodiment of FIG. 5. In step S501, the artificial intelligence engine can perform neural network-like operations to generate a plurality of optimized setting models according to a plurality of initial setting parameter sets related to the central processing unit 420 and the graphics processing unit 430. In step S502, the artificial intelligence engine writes the multiple optimized setting models into the basic input output system 410. In step S503, the motherboard 400 executes the booting procedure and starts the basic input output system 410. In step S504, the BIOS 410 can execute the optimization setting module 411. In step S505, the basic input output system 410 retrieves a plurality of initial setting parameters corresponding to the central processing unit 420 and the graphics processing unit 430. In step S506, the optimization setting module 411 compares the plurality of optimization setting models stored in the BIOS 410 according to the plurality of initial setting parameters. In step S507, the optimization setting module 411 determines whether one of the plurality of optimization setting models corresponds to the plurality of initial setting parameters. When one of the plurality of optimized setting models corresponds to the plurality of initial setting parameters, in step S508, the basic input output system 410 according to one of the plurality of optimized setting models A plurality of optimized setting parameters operate the central processing unit 420 and the graphics processing unit 430. In step S509, when the optimized setting is completed, the BIOS 410 ends the execution of the optimized setting module 411. In contrast, when any one of the plurality of optimized setting models does not correspond to the plurality of initial setting parameters, in step S510, the artificial intelligence engine will be run to perform neural network-like operations to Training at least one new optimized setting model according to the plurality of initial setting parameters. In step S511, the artificial intelligence engine writes the at least one new optimized setting model into the basic input output system 410, and obtains the multiple models according to one of the at least one new optimized setting model. Optimized setting parameters. Therefore, the performance optimization method of this embodiment enables the motherboard 400 to provide an effective automatic performance optimization function of the graphics processing unit 430.

In addition, regarding the performance optimization method described in this embodiment and other related component features, technical solutions, and implementation details of the motherboard 400, you can refer to the description of the embodiment in FIG. 4 to obtain sufficient teaching, suggestion, and implementation description. Repeat it again.

Referring again to FIGS. 3 and 4, the neural network-like operation related to the optimization of the graphics processing unit 420 described in the various embodiments of the present invention may be the neural network-like network 300 shown in FIG. 3. In this embodiment, the basic input output system 410 can capture a plurality of initial setting parameters corresponding to the central processing unit 420 and the graphics processing unit 430. For example, the artificial intelligence model may use the multiple initial setting parameters as multiple input parameters 310_1 to 310_6, and input them into the input layer of the neural network 300. Then, after the multiple operation neurons 320_1~320_8 in the hidden layer of the neural network 300 are calculated, the output layer of the neural network 300 can generate multiple output parameters 330_1~330_6. Therefore, the multiple output parameters 330_1 to 330_6 obtained by each operation. The multiple output parameters 330_1 to 330_6 are multiple optimized setting parameters, and the multiple output parameters 330_1 to 330_6 can establish an optimized setting model. In other words, after multiple inputs of different input parameters 310_1 to 310_6 to perform neural network-like operations, the artificial intelligence model can obtain multiple optimized setting models 340_1 to 340_N, and N is a positive integer greater than 1.

In this embodiment, the multiple initial setting parameters may include, for example, the model of the graphics processing unit, the preset operating frequency of the graphics processing unit, the operating voltage of the graphics processing unit, the parameters of the graphics processing unit, the operating voltage of the central processing unit, and the graphics At least one of the display memory operating frequencies of the display memory in the processing unit. Moreover, the plurality of optimized setting parameters may include, for example, at least one of optimized graphics processing unit operating voltage, optimized graphics processing unit operating frequency, optimized graphics processing unit parameters, and optimized central processing unit operating voltage. one of them. In this embodiment, the plurality of computing neurons 320_1~320_8 may be used to respectively perform weighting and accumulation operations according to the plurality of input parameters 310_1~310_6 to obtain the plurality of output parameters 330_1~330_6 . However, the parameter operation modes corresponding to the plurality of operation neurons 320_1 to 320_8 can be determined according to the types of different types of neural networks, and the present invention is not limited.

In a specific embodiment, the weight value corresponding to the aforementioned graphics processing unit model is higher than other initial setting parameters. For example, the weight value corresponding to the graphics processing unit model may be 0.2, and other initial setting parameters may be 0.1. However, in one embodiment, the weight values corresponding to the multiple initial setting parameters can also be adjusted correspondingly according to special use requirements, special setting considerations or special operating environment. The weight value corresponding to the initial setting parameter of is adjusted to a higher weight value, and the weight value corresponding to the initial setting parameter of a lower consideration factor is adjusted to a lower weight value.

It should be noted that the number of the plurality of initial setting parameters of the central processing unit 420 and the graphics processing unit 430 of the present invention is not limited to the number of the plurality of input parameters 310_1 to 310_6 in FIG. 3, and the central processing unit of the present invention The number of the multiple optimization setting parameters of the unit 420 and the graphics processing unit 430 is not limited to the number of the multiple output parameters 330_1 to 330_6 in FIG. 3. In addition, the number of computing neurons 320_1 to 320_8 in FIG. 3 is only used for illustration, and the present invention is not limited thereto.

FIG. 6 is a flowchart of a method for optimizing the performance of a processor according to an embodiment of the invention. The performance optimization method of this embodiment can be applied to the motherboards 100 and 400 of FIGS. 1 and 4. Taking the central processing unit 120 of FIG. 1 as an optimization object as an example, referring to FIG. 1 and FIG. 6, when the motherboard 100 of the embodiment of FIG. 1 executes the boot program, the computer system corresponding to the motherboard 100 can execute the embodiment of FIG.的 steps S610~S640. In step S610, the basic input output system 110 is executed to retrieve a plurality of initial setting parameters corresponding to the processor (central processing unit 120). In step S620, the basic input output system 110 compares a plurality of optimized setting models stored in the basic input output system 110 according to the plurality of initial setting parameters. In step S630, when one of the plurality of optimized setting models corresponds to the plurality of initial setting parameters, obtain a plurality of optimizations according to one of the plurality of optimized setting models Setting parameters, and when any one of the plurality of optimized setting models does not correspond to the plurality of initial setting parameters, a neural network-like operation is executed to obtain the plurality of optimized setting parameters . In step S640, the basic input output system 110 operates the processor (central processing unit 120) according to the plurality of optimized setting parameters to reduce the operating power consumption of the processor (central processing unit 120). Therefore, the performance optimization method of this embodiment enables the motherboard 100 to provide an effective automatic performance optimization function of the central processing unit 120.

Taking the graphics processing unit 430 of FIG. 4 as an optimization object as an example, referring to FIGS. 4 and 6, when the motherboard 400 of the embodiment of FIG. 4 executes the boot program, the computer system corresponding to the motherboard 400 can execute the embodiment of FIG. 6的 steps S610~S640. In step S610, the basic input output system 610 is executed to retrieve a plurality of initial setting parameters corresponding to the processors (central processing unit 420 and graphics processing unit 430). In step S620, the basic input output system 610 compares a plurality of optimized setting models stored in the basic input output system 610 according to the plurality of initial setting parameters. In step S630, when one of the plurality of optimized setting models corresponds to the plurality of initial setting parameters, obtain a plurality of optimizations according to one of the plurality of optimized setting models Setting parameters, and when any one of the plurality of optimized setting models does not correspond to the plurality of initial setting parameters, a neural network-like operation is executed to obtain the plurality of optimized setting parameters . In step S640, the basic input output system 610 operates the processor (graphic processing unit 430) according to the plurality of optimized setting parameters to reduce the operating power consumption of the processor (graphic processing unit 430). Therefore, the performance optimization method of this embodiment enables the motherboard 100 to provide an effective automatic performance optimization function of the graphics processing unit 430.

In addition, regarding the performance optimization method described in this embodiment and other related component features, technical solutions, and implementation details of the motherboards 100 and 400, please refer to the description of the above-mentioned embodiments of FIGS. 1 to 5 to obtain sufficient teachings, suggestions, and Implementation instructions, so I will not repeat them.

In summary, the performance optimization method of the processor and the motherboard using it of the present invention can be performed by comparing multiple initial setting parameters of the processor (central processing unit or graphics processing unit) through a neural network in advance Operate a plurality of optimized models trained, or use an artificial intelligence engine to execute neural network-like operations in real time according to the plurality of initial setting parameters to obtain a new optimized model. Therefore, the basic input and output system of the motherboard can obtain optimized setting parameters to operate the processor (central processing unit or graphics processing unit) according to the corresponding optimization model to effectively optimize the processor (central processing unit or graphics processing unit). Processing unit) performance and reduce operating power consumption.

Although the present invention has been disclosed in the above embodiments, it is not intended to limit the present invention. Anyone with ordinary knowledge in the technical field can make some changes and modifications without departing from the spirit and scope of the present invention. The scope of protection of the present invention shall be determined by the scope of the attached patent application.

100, 400: Motherboard 110, 410: Basic input and output system 111, 411: Optimization setting module 120, 420: Central processing unit 300: Neural network 310_1~310_6: Input parameter 320_1~320_8: Operation neuron 330_1~330_6: output parameters 340_1~340_N: optimized setting model 430: graphics processing unit S201~S211, S501~S511, S610~S640: steps

FIG. 1 is a schematic structural diagram of a motherboard according to an embodiment of the invention. 2 is a flowchart of a method for optimizing the performance of a central processing unit according to an embodiment of the invention. Fig. 3 is a schematic diagram of a neural network-like operation according to an embodiment of the invention. FIG. 4 is a schematic structural diagram of a motherboard according to another embodiment of the invention. FIG. 5 is a flowchart of a performance optimization method of a graphics processing unit according to an embodiment of the invention. FIG. 6 is a flowchart of a method for optimizing the performance of a processor according to an embodiment of the invention.

S610~S640: steps

Claims (20)

A method for optimizing the performance of a processor is suitable for a motherboard with a basic input output system, and the method includes: executing the basic input output system to retrieve a plurality of initial setting parameters corresponding to a processor; The initial setting parameters are compared with a plurality of optimized setting models stored in the basic input output system; when one of the optimized setting models corresponds to the initial setting parameters, according to the best One of the optimized setting models obtains multiple optimized setting parameters, and when any one of the optimized setting models does not correspond to the initial setting parameters, a type of neural network operation is performed to obtain the Some optimized setting parameters; and operating the processor according to the optimized setting parameters to reduce the operating power consumption of the processor. For example, the performance optimization method described in claim 1 further includes: executing this type of neural network operation by an artificial intelligence engine to generate the optimized setting models according to a plurality of initial setting parameter sets; and Write these optimized setting models into the basic input output system. In the performance optimization method described in the first item of the patent application, when any one of the optimized setting models does not correspond to the initial setting parameters, the neural network operation is executed to obtain the The step of optimizing the setting parameters includes: executing the neural network operation by an artificial intelligence engine to train at least one new optimized setting model according to the initial setting parameters; and the at least one new optimized setting model The optimized setting model is written into the basic input output system; and the optimized setting parameters are obtained according to one of the at least one new optimized setting model. The performance optimization method described in item 3 of the scope of patent application, wherein the artificial intelligence engine executes the neural network operation to train the at least one new optimized setting model according to the initial setting parameters It includes: training multiple new optimized setting models corresponding to the same processor model and different processor operating frequencies according to the initial setting parameters. According to the performance optimization method described in claim 1, wherein the processor includes a central processing unit, and the optimized setting models correspond to the same central processing unit model and different central processing unit operating frequencies. The performance optimization method described in item 5 of the scope of patent application, wherein the initial setting parameters include a central processing unit model, a central processing unit core number, a central processing unit operating frequency, a central processing unit operating voltage, and a central processing unit. At least one of the processing unit operating current, a central processing unit power, a central processing unit temperature, and a central processing unit load line, and the central processing unit model and the central processing unit operating frequency have respective corresponding weight values higher than Other initial setting parameters. The performance optimization method described in item 5 of the scope of patent application, wherein the optimized setting parameters include an optimized central processing unit operating voltage, an optimized central processing unit operating current, and an optimized central processing unit At least one of power, an optimized central processing unit temperature, an optimized central processing unit operating frequency, and an optimized central processing unit load line. According to the performance optimization method described in claim 1, wherein the processor includes a central processing unit and a graphics processing unit, and the optimized setting models correspond to the same graphics processing unit model and different graphics processing unit operations frequency. As described in item 8 of the scope of patent application, the initial setting parameters include a graphics processing unit model, a graphics processing unit preset operating frequency, a graphics processing unit operating voltage, a graphics processing unit parameter, and At least one of the operating voltage of the central processing unit and the operating frequency of a display memory of a display memory provided in the graphics processing unit, and the weight value corresponding to the model of the graphics processing unit is higher than other initial setting parameters. The performance optimization method described in item 8 of the scope of patent application, wherein the optimized setting parameters include an optimized graphics processing unit operating voltage, an optimized graphics processing unit operating frequency, and an optimized graphics processing unit At least one of parameters and an optimized operating voltage of the central processing unit. A motherboard includes: a basic input output system, including a plurality of optimized setting models, and the basic input output system is used to retrieve a plurality of initial setting parameters corresponding to a processor to compare the optimal ones Optimized setting model, wherein when one of the optimized setting models corresponds to the initial setting parameters, the basic input output system obtains a plurality of optimized settings according to one of the optimized setting models Parameters, and when any one of the optimized setting models does not correspond to the initial setting parameters, a type of neural network operation is executed to obtain the optimized setting parameters, wherein the basic input output system is based on The optimized setting parameters operate the processor to reduce the operating power consumption of the processor. For the motherboard described in claim 11, an artificial intelligence engine executes this type of neural network operation to generate the optimized setting models according to a plurality of initial setting parameter sets, and the most optimized setting models The optimized setting model is written into the basic input output system. For the motherboard described in claim 11, an artificial intelligence engine executes this type of neural network operation to train at least one new optimal setting model according to the initial setting parameters, and the artificial intelligence The engine writes the at least one new optimized setting model into the basic input output system, so that the basic input output system obtains the optimized settings according to one of the at least one new optimized setting model parameter. For the motherboard described in claim 13, wherein the artificial intelligence engine trains a plurality of new optimized setting models corresponding to the same processor model and different processor operating frequencies according to the initial setting parameters. As described in item 11 of the scope of patent application, the processor includes a central processing unit, and the optimized setting models correspond to the same central processing unit model and different central processing unit operating frequencies. For the motherboard described in item 15 of the scope of patent application, the initial setting parameters include a central processing unit model, a central processing unit core number, a central processing unit operating frequency, a central processing unit operating voltage, and a central processing unit. At least one of the unit operating current, a central processing unit power, a central processing unit temperature, and a central processing unit load line, and the weight value of the central processing unit model and the central processing unit operating frequency is higher than the others Initial setting parameters. For the motherboard described in item 15 of the scope of patent application, the optimized setting parameters include an optimized central processing unit operating voltage, an optimized central processing unit operating current, and an optimized central processing unit power , At least one of an optimized central processing unit temperature, an optimized central processing unit operating frequency, and an optimized central processing unit load line. The motherboard described in item 11 of the scope of patent application, wherein the processor includes a central processing unit and a graphics processing unit, and the optimized setting models correspond to the same graphics processing unit model and different graphics processing unit operating frequencies . For the motherboard described in item 18 of the scope of patent application, the initial setting parameters include a graphics processing unit model, a graphics processing unit preset operating frequency, a graphics processing unit operating voltage, a graphics processing unit parameter, and a central At least one of the operating voltage of the processing unit and the operating frequency of a display memory of a display memory provided in the graphics processing unit, and the weight value corresponding to the model of the graphics processing unit is higher than other initial setting parameters. The motherboard described in item 18 of the scope of patent application, wherein the optimized setting parameters include an optimized graphics processing unit operating voltage, an optimized graphics processing unit operating frequency, and an optimized graphics processing unit parameter And at least one of an optimized operating voltage of the central processing unit.

Images