TWI768478B - Electronic device and method and for adaptively arranging external hardware resources - Google Patents

Abstract

The disclosure provides an electronic device and a method and for adaptively arranging external hardware resources. The electronic device is built with a virtual environment, and a guest operating system is executed in the virtual environment. The method includes: obtaining a usage rate of a specific hardware device in the electronic device in response to receiving a virtual device command from the guest operating system; in response to determining that the usage rate of the specific hardware device is higher than a usage rate threshold, obtaining a foreground process, background process, and operating system services of the electronic device, and determining an external resource demand accordingly; arranging a hardware device of an external device based on the external resource demand, and using the hardware device of the external device and the specific hardware device to collaboratively execute the foreground process, the background process and the operating system service.

Description

A method for adaptively calling external hardware resources

The present invention relates to a technology for invoking hardware resources, and more particularly, to a method and electronic device for adaptively invoking external hardware resources.

It has been measured that the performance data of electronic devices using the Chrome Operating System (OS), such as Chromebooks, when using a graphics processing unit (GPU) is far inferior to that of electronic devices using Windows ^™ 10.

In the prior art, although there is a technology of invoking the hardware resources of an external device to assist the local device to run programs and software to improve the running performance, there is currently no relevant technology for a virtual environment.

In view of this, the present invention provides a method and electronic device for adaptively calling external hardware resources, which can be used to solve the above technical problems.

The present invention provides a method for adaptively calling external hardware resources, which is suitable for An electronic device, wherein a virtual environment is built in the electronic device, and a guest operating system is running in the virtual environment, comprising: in response to receiving a virtual device command from the guest operating system, obtaining the information of a specific hardware device in the electronic device a usage rate; in response to determining that the usage rate of a specific hardware device is higher than a first usage rate threshold, obtain a foreground program, a background program and at least one operating system service of the electronic device, and obtain a foreground program, a background program and at least one operating system service according to the foreground program, background program and At least one operating system service determines an external resource requirement; invokes a hardware device of at least one external device based on the external resource requirement, and uses the hardware device of the at least one external device and a specific hardware device to cooperatively run foreground programs, background programs and At least one operating system service.

The invention provides an electronic device, wherein a virtual environment is built in the electronic device, and a client operating system runs in the virtual environment. The electronic device includes a storage circuit and a processor. The storage circuit stores a plurality of modules. The processor is coupled to the storage circuit and accesses the plurality of modules to perform the following steps: obtaining a usage rate of a specific hardware device in the electronic device in response to receiving a virtual device command from the client operating system; In response to determining that the usage rate of the specific hardware device is higher than a first usage rate threshold value, obtain a foreground program, a background program and at least one operating system service of the electronic device, and according to the foreground program, the background program and the at least one operating system The service determines an external resource requirement; invokes the hardware device of at least one external device based on the external resource requirement, and uses the hardware device of the at least one external device and a specific hardware device to cooperatively run a foreground program, a background program and at least one operating system Serve.

100,200: Electronic devices

102, 202: Virtual Environments

104,206: Customer Operating System

201:CPU

204: Resource scheduling library

204a: Resource Scheduling Reasoning API

204b: External Resource Scheduling API

299: External Devices

S310~S330, S410~S480: Steps

FIG. 1 is a schematic diagram of a physical environment and a virtual environment of an electronic device according to an embodiment of the present invention.

FIG. 2 is a schematic diagram of the physical environment, the virtual environment and the external device of the electronic device shown in FIG. 1 .

FIG. 3 is a flowchart of a method for adaptively calling external hardware resources according to an embodiment of the present invention.

FIG. 4 is a schematic diagram of a hardware device mechanism for invoking an external device according to an embodiment of the present invention.

Please refer to FIG. 1 , which is a schematic diagram of a physical environment and a virtual environment of an electronic device according to an embodiment of the present invention. In different embodiments, the electronic device 100 is, for example, various computer devices and smart devices, but not limited thereto. In FIG. 1 , the electronic device 100 may include various processors (eg, central processing unit (CPU), GPU), various storage circuits (eg, system random access memory (RAM), image RAM), screens, screens A hardware device such as a controller, but not limited to this.

In addition, a virtual environment 102 may be built on the electronic device 100, and a guest operating system (guest OS) 104 may be executed therein. In different embodiments, the virtual environment 102 can be implemented by, for example, a virtual machine (VM), a container (container), a Docker image, Wine, etc., and the operating system 104 is, for example, Chrome OS, Windows ^™ , or other similar ones.

As shown in FIG. 1 , a graphics library (GL) renderer may run in the virtual environment 102 , and VirGL, Vulkan, Virtio-GPU and applications may run in the operating system 104 . . In addition, the electronic device 100 itself may run a host OS, and the host OS may run programs such as OpenGL, Vulkan, and system call interface (SCI), but not limited thereto.

For the various software/programs shown in FIG. 1 , reference may be made to the descriptions in the related art, so the details thereof will not be repeated here.

Please refer to FIG. 2 , which is a schematic diagram of the physical environment, the virtual environment and the external device of the electronic device shown in FIG. 1 . In FIG. 2, the electronic device 200 can be connected to one or more external devices 299 (such as various computer devices and/or smart devices) in various wired/wireless ways, and the electronic device 200 can be based on the method proposed in the present invention Available resources provided by one or more of these external devices 299 are adaptively invoked, details of which will be described later.

As shown in FIG. 2 , the hardware structure and implementation of the electronic device 200 are substantially similar to the electronic device 100 in FIG. 1 , but the virtual environment 202 of the electronic device 200 is further provided with a resource arrangement library 204 , which may include Resource scheduling reasoning application programming interface (API) 204a and external resource scheduling API 204b.

In an embodiment of the present invention, the CPU 201 in the electronic device 200 can be used to execute the resource arrangement reasoning API 204a and the external resource arrangement API 204b, so as to execute the method for adaptively calling external hardware resources proposed by the present invention. Further explanation.

Please refer to FIG. 3 , which is a flowchart of a method for adaptively calling external hardware resources according to an embodiment of the present invention. The method of this embodiment can be performed by the electronic device 200 in FIG. 2 , and the details of each step in FIG. 3 will be described below in conjunction with the elements shown in FIG. 2 .

First, in step S310 , in response to receiving the virtual device command from the guest operating system 206 , the CPU 201 executes the resource allocation reasoning API 204 a to obtain the usage rate of a specific hardware device in the electronic device 200 . In one embodiment, the above-mentioned virtual device command is, for example, a Virtio device command, but it is not limited thereto. In addition, in different embodiments, the above-mentioned specific hardware devices are, for example, the CPU 201 , the GPU, the system RAM and/or the image RAM of the electronic device 200 , or other hardware devices in the electronic device 200 .

For the convenience of description, the CPU 201 will be used as an example of the specific hardware device below, but it is not intended to limit possible implementations of the present invention. In this case, the CPU 201 can obtain its own usage rate in step S310, and can then determine whether the obtained usage rate is higher than the first load threshold value.

In different embodiments, the first load threshold value can be understood as a basis for judging whether a specific hardware device (such as the CPU 201 ) is in an overload state, and the designer can set the value of the first load threshold value ( For example, 95%), but not limited to this.

In one embodiment, in response to determining that the usage rate of the specific hardware device is not higher than the first usage rate threshold, the CPU 201 can execute the resource allocation reasoning API 204a to run the foreground program of the electronic device 200 only with the specific hardware device , background programs and Operating system services.

On the other hand, in step S320 , in response to determining that the usage rate of the specific hardware device is higher than the first usage rate threshold, the CPU 201 executes the resource allocation reasoning API 204a to obtain the foreground program, background program and operation of the electronic device 200 System service (OS service), and determines the external resource demand according to foreground programs, background programs and operating system services. In different embodiments, the aforementioned foreground/background programs are, for example, various foreground/background applications/software running on the electronic device 200, and the operating system services are, for example, various software services recorded in the job administrator , but not limited to this.

In one embodiment, the CPU 201, for example, can execute the resource scheduling reasoning API 204a to input foreground programs, background programs, and operating system services into a trained artificial intelligence model, wherein the artificial intelligence model can correspond to foreground programs, background programs, and operating systems service and output a specific resource demand.

In one embodiment, the above-mentioned artificial intelligence model is, for example, a long short term memory (LSTM) model, and in the related pre-training process, the designer may, for example, run a certain device with some foreground programs first. , the hardware device usage rate (such as CPU usage rate), device specifications, capacity, and device information displayed by the operating system during background programs and operating system services are converted into two text files, which are used as the training data for the above artificial intelligence model set. In the process of training the artificial intelligence model, each set of foreground programs, background programs and operating system services and the resources (such as page files) of the called hardware devices (including CPU/GPU/RAM) can be created ), Swap archives, Swap splits, etc.). By this, After completing the training of the artificial intelligence model, when the artificial intelligence model receives a certain set of foreground programs, background programs and operating system services represented by text files, the artificial intelligence model can accordingly provide the required amount of The hardware resources (ie, the specific resource requirements) can allow the hardware device to run the set of foreground programs, background programs, and operating system services at a low utilization rate (eg, 50%).

After obtaining the specific resource requirements, the CPU 201 can execute the resource scheduling inference API 204a to determine the external resource requirements based on the specific resource requirements and the available resources of a specific hardware device (eg, the CPU 201 ). For example, the CPU 201 may use the difference between the specific resource demand and the available resource of the specific hardware device as the above-mentioned external resource demand, but it is not limited thereto.

Next, in step S330, the CPU 201 can execute the external resource arrangement API 204b to call the hardware device of the external device based on the external resource demand, and use the hardware device of the external device and the specific hardware device to cooperatively run the foreground program, the background program and the Operating system services. For the convenience of description, the details of step S330 are described below with reference to FIG. 4 .

Please refer to FIG. 4 , which is a schematic diagram of a hardware device mechanism for calling an external device according to an embodiment of the present invention. The various steps of FIG. 4 may be implemented by the CPU 201 by executing the external resource scheduling API 204b. Roughly speaking, assuming that all the external devices 299 have virtual environments, for the i-th external device in the external devices 299, the CPU 201 can determine the first external device in the external devices 299 (hereinafter referred to as the first external device) Whether all available resources that can be provided by the i-th external device can meet the external resource demand. If so, the CPU 201 can accordingly call the The hardware devices from 1 to the i-th external device are used to assist the electronic device 200 to run foreground programs, background programs and operating system services. until the external resource demand is satisfied, but not limited to this.

Specifically, in step S410, the CPU 201 may determine whether the first external device has an operating system. If so, the CPU 201 can execute step S420, otherwise, execute step S460 to ignore the first external device. And, after step S460, the CPU 201 may continue to execute step S480, the details of which will be described later.

In step S420, the CPU 201 may determine whether the first external device has a virtual environment. If not, the CPU 201 may execute step S470 to control the first external device to establish a virtual environment, and then execute step S430. On the other hand, if the CPU 201 determines in step S420 that the first external device has a virtual environment, the CPU 201 may then execute step S430 to obtain the first hardware device of the first external device corresponding to the specific hardware device. - Available hardware resources. Following the above example, in the case where the specific hardware device is assumed to be the CPU 201 of the electronic device 200, the CPU 201 can accordingly obtain the available hardware resources of the CPU in the first external device as the above-mentioned first available hardware resources, but But not limited to this.

Afterwards, in step S440, the CPU 201 may determine whether the first available hardware resource meets the external resource requirement. If so, the CPU 201 can execute step S450 to call the first hardware device (eg CPU) of the first external device, and use the first hardware device to cooperate with the specific hardware device to run the foreground program, background program and operating system service. In one embodiment, the CPU 201 may, for example, store relevant information of the first hardware device Report back to the client operating system 206 in FIG. 2 , so that the client operating system 206 can use the CPU 201 of the electronic device 200 and the first hardware device (eg CPU) of the first external device at the same time to run the prospect obtained in step S320 Programs, background programs, and operating system services, but not limited thereto.

On the other hand, if the CPU 201 determines in step S440 that the first available hardware resource does not meet the external resource requirement, the CPU 201 may continue to perform step S480 to call other external devices.

Roughly speaking, the CPU 201 can then perform the content similar to steps S410 to S480 on the second external device in the external device 299 (hereinafter referred to as the second external device), but the details of performing step S440 on the second external device are omitted. There are different.

Specifically, for the second external device, the CPU 201 may be configured to: determine whether the second external device has an operating system (corresponding to step S410 ); in response to determining that the second external device has an operating system, determine whether the second external device has an operating system virtual environment (corresponding to step S420 ); in response to determining that the second external device has a virtual environment, obtain a first hardware device (for example, a CPU of the second external device) corresponding to a second hardware device in the second external device 2. Available hardware resources (corresponding to step S430).

After acquiring the second available hardware resource of the second hardware device, the CPU 201 may determine whether the sum of the first available hardware resource and the second available hardware resource satisfies the external resource requirement. If so, the CPU 201 can accordingly call the first hardware device and the second hardware device, and use the first hardware device, the second hardware device and the specific hardware device to cooperatively run the foreground program, the background program and the operating system service .

On the other hand, if the CPU 201 still determines that the sum of the first available hardware resource and the second available hardware resource does not meet the external resource requirement, the CPU 201 can execute the process similar to the third external device in the external device 299 again. The content of steps S410 to S480 is until the sum of the available hardware resources of the N external devices under consideration satisfies the external resource requirement.

In some embodiments, after step S330 is executed, the CPU 201 may determine whether the usage rate of any one of the specific hardware device and the hardware devices of each external device is higher than a second usage rate threshold. If so, the CPU 201 can retrain the above-mentioned artificial intelligence model, otherwise, it can maintain the artificial intelligence model.

In detail, after step S330 is executed, the specific hardware device and the hardware devices of each external device should run at a lower usage rate. In other words, if the specific hardware device and the hardware devices of each external device are still running at a state higher than the second usage rate threshold (for example, 50%), it means that the external resource demand inferred by the above artificial intelligence model may be Hardware devices that cannot effectively degrade specific hardware devices and various external devices due to poor quality. Therefore, the CPU 201 can retrain the above-mentioned artificial intelligence model accordingly.

In one embodiment, the CPU 201 may maintain an observation agent, and the observation agent may be used to determine whether the usage rate of any one of the specific hardware device and the hardware devices of each external device is higher than the second usage rate Threshold value. If so, the observation program can provide a punishment signal to trigger the CPU 201 to retrain the above-mentioned artificial intelligence model. On the other hand, if the observation program determines that the usage rate of the specific hardware device and the hardware device of each external device is not higher than the second usage rate If the threshold value is set, the observation program can provide a reward signal to trigger the CPU 201 to maintain the above-mentioned artificial intelligence model, but it is not limited thereto.

To sum up, the method provided by the embodiments of the present invention can be based on the current foreground program, background program and operating system when the electronic device runs the guest operating system in the virtual environment and causes a high usage rate of a specific hardware device The service determines the corresponding external resource demand, and then invokes the hardware device of the external device to assist the electronic device to execute the aforementioned foreground program, background program and operating system service. In this way, the utilization rate of the specific hardware device of the electronic device can be reduced, thereby obtaining the effect of increasing the system performance. Moreover, by properly training the artificial intelligence model for inferring the demand of external resources, the hardware device of the external device can be invoked more intelligently and reasonably.

Although the present invention has been disclosed above by the 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. Therefore, The protection scope of the present invention shall be determined by the scope of the appended patent application.

S310~S330: Steps

Claims (10)

A method for adaptively calling external hardware resources, suitable for an electronic device, wherein the electronic device is built with a virtual environment, and a client operating system runs in the virtual environment, comprising: obtaining a usage rate of a specific hardware device in the electronic device in response to receiving a virtual device command from the client operating system; In response to determining that the usage rate of the specific hardware device is higher than a first usage rate threshold value, obtain a foreground program, a background program and at least one operating system service of the electronic device, and according to the foreground program and the background program and the at least one operating system service determines an external resource requirement; The hardware device of at least one external device is invoked based on the external resource requirement, and the foreground program, the background program and the at least one operating system service are cooperatively run by the hardware device of the at least one external device and the specific hardware device . The method of claim 1, wherein the step of determining the external resource requirement according to the foreground program, the background program and the at least one operating system service comprises: inputting the foreground program, the background program and the at least one operating system service into a trained artificial intelligence model, wherein the artificial intelligence model outputs a specific resource in response to the foreground program, the background program and the at least one operating system service demand; and The external resource requirement is determined based on the specific resource requirement and an available resource of the specific hardware device. The method of claim 2, wherein after the step of cooperatively running the foreground program, the background program and the at least one operating system service with the hardware device of the at least one external device and the specific hardware device, further comprising: : In response to determining that the usage rate of any one of the specific hardware device and the hardware device of each of the external devices is higher than a second usage rate threshold, the artificial intelligence model is retrained. The method of claim 1, wherein the at least one external device includes a first external device, and the step of invoking the hardware resource of the at least one external device based on the external resource requirement comprises: determining whether the first external device has an operating system; In response to determining that the first external device has the operating system, determining whether the first external device has a virtual environment; In response to determining that the first external device has the virtual environment, obtaining a first available hardware resource of a first hardware device in the first external device corresponding to the specific hardware device; In response to determining that the first available hardware resource meets the external resource requirement, the first hardware device is invoked. The method of claim 4, wherein in response to determining that the first external device does not have any operating system, the method further comprises ignoring the first external device. The method of claim 4, wherein in response to determining that the first external device does not have any virtual environment, the method further comprises controlling the first external device to establish the virtual environment. The method of claim 4, wherein the at least one external device further comprises a second external device, and in response to determining that the first available hardware resource does not meet the external resource requirement, the method further comprises: determining whether the second external device has the operating system; In response to determining that the second external device has the operating system, determining whether the second external device has the virtual environment; In response to determining that the second external device has the virtual environment, obtaining a second available hardware resource of a second hardware device in the second external device corresponding to the specific hardware device; In response to determining that the sum of the first available hardware resource and the second available hardware resource satisfies the external resource requirement, the first hardware device and the second hardware device are called. The method of claim 1, wherein the virtual device command comprises a Virtio device command from the guest operating system. The method of claim 1, wherein in response to determining that the usage rate of the specific hardware device is not higher than the first usage rate threshold value, further comprising running the foreground program and the background program only on the specific hardware device and the at least one operating system service. An electronic device, wherein the electronic device is built with a virtual environment, and a client operating system runs in the virtual environment, comprising: a storage circuit storing a plurality of modules; and A processor, coupled to the storage circuit, accesses the modules to perform the following steps: obtaining a usage rate of a specific hardware device in the electronic device in response to receiving a virtual device command from the client operating system; In response to determining that the usage rate of the specific hardware device is higher than a first usage rate threshold value, obtain a foreground program, a background program and at least one operating system service of the electronic device, and according to the foreground program and the background program and the at least one operating system service determines an external resource requirement; The hardware device of at least one external device is invoked based on the external resource requirement, and the foreground program, the background program and the at least one operating system service are cooperatively run by the hardware device of the at least one external device and the specific hardware device .

Images