Operation control method and device of cloud mobile phone, electronic equipment and medium

Abstract

The invention provides an operation control method and device of a cloud mobile phone, electronic equipment and a medium, and relates to the technical field of computers, in particular to the cloud computing technology. According to the specific implementation scheme, the method comprises the steps of obtaining the utilization rate of a target GPU corresponding to a monitoring cloud mobile phone in real time; when it is determined that the usage rate of the target GPU exceeds a usage rate threshold value, obtaining each associated cloud mobile phone sharing the target GPU with a monitoring cloud mobile phone; and identifying each GPU high-load application in the monitoring cloud mobile phone and each associated cloud mobile phone, and performing operation control on each GPU high-load application according to the operation description information of the cloud mobile phone where each GPU high-load application is located. According to the technical scheme of the embodiment of the invention, real-time abnormity monitoring and abnormity notification of the GPU utilization rate of the cloud mobile phone can be realized, so that a user of a cloud mobile phone platform can timely perceive a problem and perform corresponding processing, and the efficiency and the reliability of operation control of the cloud mobile phone are improved.

Description

Technical Field

[0001] The present disclosure relates to the field of computer technology, in particular to cloud computing technology, and specifically to a cloud phone operation control method, a cloud phone operation control device, an electronic device, and a non-transitory computer-readable storage medium. Background Art

[0002] With the popularization of mobile networks and the advancement of cloud computing technology, cloud phones, as a new type of mobile device, have core advantages in their powerful computing power and unlimited storage space. These advantages make cloud phones feasible and practical in multiple application scenarios.

[0003] In the cloud phone environment, you can use commands to view the GPU (Graphics Processing Unit) usage, so that users can actively call these commands to query the GPU usage of cloud phone instances. For users with a large number of cloud phone instances, manually querying GPU usage is not only inefficient, but also difficult to achieve real-time monitoring, which may result in missing opportunities to adjust resource allocation and optimize performance in a timely manner. Summary of the invention

[0004] The present disclosure provides a cloud phone operation control method, a cloud phone operation control device, an electronic device and a non-transitory computer-readable storage medium.

[0005] According to one aspect of the present disclosure, a method for controlling the operation of a cloud phone is provided, comprising:

[0006] Obtain the usage rate of the target GPU corresponding to the monitored cloud phone in real time;

[0007] When it is determined that the usage rate of the target GPU exceeds the usage rate threshold, obtaining each associated cloud phone that shares the target GPU with the monitoring cloud phone;

[0008] Each GPU high-load application is identified in the monitoring cloud phone and each associated cloud phone, and the operation of each GPU high-load application is controlled according to the operation description information of the cloud phone where each GPU high-load application is located.

[0009] According to another aspect of the present disclosure, there is also provided an operation control device for a cloud phone, comprising:

[0010] GPU usage monitoring module, used to obtain the usage of the target GPU corresponding to the monitored cloud phone in real time;

[0011] An associated cloud phone acquisition module, configured to acquire associated cloud phones that share the target GPU with the monitoring cloud phone when it is determined that the usage rate of the target GPU exceeds a usage rate threshold;

[0012] The GPU high-load application control module is used to identify each GPU high-load application in the monitoring cloud phone and each associated cloud phone, and to control the operation of each GPU high-load application according to the operation description information of the cloud phone where each GPU high-load application is located.

[0013] According to another aspect of the present disclosure, there is also provided an electronic device, including:

[0014] at least one processor; and

[0015] a memory communicatively connected to the at least one processor; wherein,

[0016] The memory stores instructions that can be executed by the at least one processor, and the instructions are executed by the at least one processor so that the at least one processor can execute the operation control method of the cloud phone as described in any one of the embodiments of the present disclosure.

[0017] According to another aspect of the present disclosure, a non-transitory computer-readable storage medium storing computer instructions is also provided, wherein the computer instructions are used to enable a computer to execute the cloud phone operation control method according to any one of the embodiments of the present disclosure.

[0018] According to another aspect of the present disclosure, a computer program product is also provided, including a computer program, which, when executed by a processor, implements the operation control method of the cloud phone of any embodiment of the present invention.

[0019] It should be understood that the content described in this section is not intended to identify the key or important features of the embodiments of the present disclosure, nor is it intended to limit the scope of the present disclosure. Other features of the present disclosure will become easily understood through the following description. BRIEF DESCRIPTION OF THE DRAWINGS

[0020] The accompanying drawings are used to better understand the present solution and do not constitute a limitation of the present disclosure.

[0021] Figure 1 is a schematic diagram of an operation control method of a cloud phone provided according to an embodiment of the present disclosure;

[0022] Figure 2 is a schematic diagram of another cloud phone operation control method provided according to an embodiment of the present disclosure;

[0023] Figure 3 is a schematic diagram of another cloud phone operation control method provided according to an embodiment of the present disclosure;

[0024] Figure 4 It is a schematic diagram of an overall process of a solution applicable to a specific application scenario of the embodiment of the present disclosure;

[0025] Figure 5 It is a structural diagram of an operation control device of a cloud phone provided according to an embodiment of the present disclosure;

[0026] Figure 6 It is a block diagram of an electronic device used to implement a method for controlling the operation of a cloud phone according to an embodiment of the present disclosure. DETAILED DESCRIPTION

[0027] The following is a description of exemplary embodiments of the present disclosure in conjunction with the accompanying drawings, including various details of the embodiments of the present disclosure to facilitate understanding, which should be considered as merely exemplary. Therefore, it should be recognized by those of ordinary skill in the art that various changes and modifications may be made to the embodiments described herein without departing from the scope and spirit of the present disclosure. Similarly, for the sake of clarity and conciseness, the description of well-known functions and structures is omitted in the following description.

[0028] Figure 1 It is a schematic diagram of a cloud phone operation control method provided according to an embodiment of the present disclosure. The embodiment of the present disclosure is applicable to the case where the GPU usage rate of the cloud phone is monitored in real time and the application program running on the cloud phone is controlled based on the numerical value of the GPU usage rate. The method can be executed by the operation control device of the cloud phone, which can be implemented in hardware and / or software, and can generally be integrated in a cloud phone server or a server cluster composed of multiple cloud phone servers. Among them, the cloud phone server (or, it can also be called a cloud phone platform) is used to perform unified scheduling and operation control of multiple cloud phones.

[0029] Correspondingly, such as Figure 1 As shown, the method may specifically include:

[0030] S101. Obtain and monitor the usage rate of the target GPU corresponding to the cloud phone in real time.

[0031] A monitoring cloud phone refers to a cloud phone instance that has a GPU monitoring program deployed. By running the GPU monitoring program on the monitoring cloud phone, the usage rate of a specific GPU (i.e., the target GPU) that provides graphics processing resources to the monitoring cloud phone can be monitored in real time. The GPU monitoring program can directly use the command line tool or monitoring tool of the GPU, or it can be a custom-written GPU monitoring script, which is not limited in this embodiment.

[0032] GPU can be specifically understood as hardware dedicated to graphics processing in a cloud phone environment. In a cloud phone, the GPU is not only responsible for graphics rendering, but can also participate in computationally intensive tasks. Since multiple cloud phone instances share a GPU, when the GPU load is too high, it will also affect other cloud phone users who share the GPU. Therefore, monitoring the GPU usage is crucial to ensure cloud phone performance and resource optimization.

[0033] The usage rate of the target GPU may be monitored and acquired in real time at a set monitoring frequency, for example, 30 seconds, 1 minute, or 5 minutes.

[0034] S102. When it is determined that the usage rate of the target GPU exceeds the usage rate threshold, obtain each associated cloud phone that shares the target GPU with the monitoring cloud phone.

[0035] The GPU usage threshold can be specifically understood as a performance indicator limit set when monitoring GPU performance, which is used to determine whether the GPU usage is in a normal or expected state.

[0036] Generally speaking, when the GPU usage exceeds this usage threshold, the GPU resources are overloaded, and the cloud phone may experience freezes when transmitting images back, affecting the user experience of the cloud phone. Therefore, it is necessary to monitor the overload status in real time and take certain measures to reduce screen freezes. The usage threshold can be preset according to actual conditions, for example, 90% or 95%, etc., and this embodiment does not limit this.

[0037] Specifically, in the process of obtaining the usage rate of the target GPU corresponding to the monitored cloud phone in real time, if it is found that the usage rate of the target GPU that provides GPU resources to the monitoring cloud phone exceeds the usage rate threshold, it means that other cloud phones sharing the target GPU with the monitoring cloud phone (that is, associated cloud phones) are also in a state of GPU resource overload. These associated cloud phones may also experience freezes in the screen transmitted back by the cloud phone. At this time, it is also necessary to locate these associated cloud phones, and treat these associated cloud phones together with the monitoring cloud phone as an affected cloud phone set, and control the operation of the application for the affected cloud phone set as a whole.

[0038] S103, identifying each GPU high-load application in the monitoring cloud phone and each associated cloud phone, and controlling the operation of each GPU high-load application according to the operation description information of the cloud phone where each GPU high-load application is located.

[0039] GPU high-load applications can be specifically understood as applications that require the GPU to perform a large number of calculations and processing tasks, such as cloud games, high-definition video conferencing, online learning platforms, high-resolution movie players, animation production applications, movie special effects production applications, and high-definition surveillance cameras and video analysis systems, etc., which require GPU rendering.

[0040] Specifically, after acquiring and monitoring each associated cloud phone that shares a target GPU with the cloud phone, identify the GPU high-load applications in each cloud phone based on the cloud phone's IP and the resource usage information in the target GPU, and make some targeted processing measures based on the operation description information of the cloud phone where each GPU high-load application is located (for example, application name and version, GPU usage, application status, and error log), such as limiting the number of multi-tasking processes and avoiding running multiple high GPU-load applications at the same time, so as to reduce the overall load of the target GPU, realize the operation control of each GPU high-load application, and optimize the use of GPU resources.

[0041] The technical solution of the disclosed embodiment can achieve real-time abnormal monitoring of the GPU usage of the cloud phone. When the GPU usage is abnormal, the high-load applications of each GPU in the monitoring cloud phone and each of the associated cloud phones are identified, and the operation of each high-load application of each GPU is controlled according to the operation description information of the cloud phone where each high-load application of each GPU is located. By real-time monitoring of the GPU usage of the cloud phone, the tedious steps of manual inspection by the user are not only omitted, but also potential problems can be quickly discovered and handled. This automatic operation control mechanism ensures the stable operation of the target GPU and improves the reliability of the overall operation of the cloud phone.

[0042] On the basis of the above embodiments, in addition to automatically controlling the operation of each GPU high-load application according to the operation description information of the cloud phone where the GPU high-load application is located (the user is unaware of the entire process), the embodiments of the present invention can also actively issue abnormal warnings or notifications so that system administrators or cloud phone users can promptly understand the usage of GPU resources and take corresponding measures.

[0043] In a specific example, when the usage rate of the target GPU is monitored to exceed the preset usage rate threshold, it means that the GPU resources are overloaded. At this time, the cloud phone platform can issue a GPU warning to the cloud phone user. After receiving the GPU warning, the cloud phone user can take some targeted measures to deal with this problem, such as manually closing or optimizing the cloud phone application. Alternatively, the cloud phone platform can also issue a GPU warning to the administrator of the merchant to which the cloud phone belongs. After receiving the GPU warning, the administrator can also take some targeted measures to deal with this problem, such as directly choosing to kill the specified GPU high-load application in the background.

[0044] In addition, the various embodiments of the present disclosure mainly implement an operation control scheme when the usage rate of the target GPU exceeds the usage rate threshold. In fact, when the usage rate of the target GPU is less than the preset low usage rate threshold, the cloud phone application can also be effectively controlled. In a specific example, when the usage rate of the target GPU corresponding to the monitored cloud phone is less than the low usage rate threshold (for example, 20% or 30%), at this time, the application currently running in the foreground of the monitored cloud phone can be obtained, and it can be detected whether the application is in the application setting with the best display effect. If not, the application settings of the application can be increased (for example, increasing the resolution or turning on high-consumption graphics effects, etc.) to ensure the effective use of GPU resources when the target GPU is sufficient, thereby improving the application usage experience of cloud phone users.

[0045] Figure 2 It is a schematic diagram of another cloud phone operation control method provided according to an embodiment of the present disclosure. This embodiment is refined based on the above embodiments. In this embodiment, the operation of "identifying each GPU high-load application in the monitoring cloud phone and each of the associated cloud phones" is concretized as: "Identify the application running in the foreground in the monitoring cloud phone and each of the associated cloud phones, respectively, and obtain the resource occupancy rate of each of the applications on the target GPU; determine the application whose resource occupancy rate exceeds the occupancy rate threshold as the GPU high-load application."

[0046] And, the operation of “controlling the operation of each GPU high-load application according to the operation description information of the cloud phone where each GPU high-load application is located” is concretized as: “re-allocating the resources in the target GPU to each GPU high-load application according to the network transmission speed between the cloud phone where each GPU high-load application is located and each entity device currently establishing a communication connection.”

[0047] Correspondingly, such as Figure 2 As shown, the method may specifically include:

[0048] S201. Obtain and monitor the usage rate of the target GPU corresponding to the cloud phone in real time.

[0049] S202. When it is determined that the usage rate of the target GPU exceeds the usage rate threshold, obtain each associated cloud phone that shares the target GPU with the monitoring cloud phone.

[0050] S203, identifying the applications running in the foreground in the monitoring cloud phone and each associated cloud phone respectively, and obtaining the resource occupancy rate of each application on the target GPU.

[0051] S204: Determine an application program whose resource occupancy rate exceeds a occupancy rate threshold as a GPU high-load application.

[0052] Generally speaking, applications that require GPU for graphics rendering, including cloud games, HD video conferencing, and high-resolution movie players, will consume a large amount of GPU resources (referring to various hardware resources used to perform computing tasks, such as processing cores, video memory, and memory bandwidth) when running in the foreground.

[0053] Therefore, when it is determined that the usage rate of the target GPU exceeds the usage threshold, the applications running in the foreground of each cloud phone that currently shares the target GPU are identified, and the resource occupancy rate of each application on the target GPU is obtained, and GPU high-load applications whose resource occupancy rate exceeds the occupancy threshold are screened out.

[0054] S205 , reallocating resources in the target GPU to each GPU high-load application according to the network transmission speed between the cloud phone where each GPU high-load application is located and each physical device currently establishing a communication connection.

[0055] Specifically, the network transmission speed can be understood as the rate at which data is transmitted in the network, measured by the amount of data transmitted per second. In the field of cloud phones, network transmission speed is crucial to user experience, especially in applications that require high-resolution image display, such as cloud games and high-definition video playback. If the network transmission speed is insufficient, it may cause image delay, freeze or degradation of image quality. After determining the GPU high-load applications of each cloud phone, the resource scheduling algorithm is used to reallocate resources in the target GPU for each GPU high-load application according to the GPU load of the application and the network transmission speed. If it is detected that the network transmission speed of a cloud phone is fast, the system may allocate more GPU resources to process higher-resolution images and video streams; if it is detected that the network transmission speed of a cloud phone is slow, in order to ensure the smooth operation and user experience of the cloud phone application, data compression and encoding optimization technology are used to reduce the amount of transmitted data; or, the data or calculation results that have been transmitted are cached to reduce repeated transmission and reduce the occupation of network bandwidth to reduce the overall load of the GPU, realize the operation control of each GPU high-load application, and optimize the use of GPU resources.

[0056] The technical solution of the disclosed embodiment obtains the usage rate of the target GPU corresponding to the monitoring cloud phone in real time; when it is determined that the usage rate of the target GPU exceeds the usage rate threshold, obtains each associated cloud phone that shares the target GPU with the monitoring cloud phone; identifies the applications running in the foreground in the monitoring cloud phone and each associated cloud phone respectively, and obtains the resource occupancy rate of each application on the target GPU; determines the application whose resource occupancy rate exceeds the occupancy rate threshold as a GPU high-load application; and re-allocates the resources in the target GPU to each GPU high-load application according to the network transmission speed between the cloud phone where each GPU high-load application is located and each entity device currently establishing a communication connection. The technical means can realize real-time abnormal monitoring of the GPU usage rate of the cloud phone. When the GPU usage rate is abnormal, the high-load foreground applications of each GPU in the monitoring cloud phone and each associated cloud phone are effectively identified according to the resource occupancy rate and the occupancy rate threshold, and the resources in the target GPU are allocated according to the network transmission speed of each application. Under limited network conditions, it is ensured that the foreground application obtains reasonable GPU resources, ensures the performance of the foreground application, avoids resource waste, and ensures the user experience of the monitoring cloud phone and each associated cloud phone user, ensures the stable operation of the target GPU, and improves the reliability of the overall operation of the cloud phone.

[0057] In another optional implementation of this embodiment, according to the network transmission speed between the cloud phone where each GPU high-load application is located and each entity device currently establishing a communication connection, resources in the target GPU are reallocated to each GPU high-load application, including:

[0058] Real-time acquisition of the current network transmission speed between the cloud phone where the current GPU high-load application is located and the physical device currently establishing a communication connection, and acquisition of the current frame rate of the current GPU high-load application;

[0059] If the current frame rate is greater than the current network transmission speed, the current frame rate is adjusted to match the current network transmission speed.

[0060] The current frame rate of the current GPU high-load application can be understood as the number of images rendered per second by the current GPU high-load application using the resources provided by the target GPU. Generally speaking, when the resolution is constant, the higher the frame rate, the higher the GPU resource occupancy rate.

[0061] It is understandable that when the current frame rate is higher than the current network transmission speed, due to the limitation of the current network transmission speed, high-frame rate images cannot be immediately provided to cloud phone users for high-definition display. At this time, even if the target GPU is highly occupied, the desired display effect is not achieved. Based on this, the current frame rate can be adjusted to match the current network transmission speed. At this time, while ensuring the original display effect of the cloud phone user, the resource occupation of the target GPU by the current GPU high-load application can be effectively reduced.

[0062] Through the above settings, the frame rate and network transmission speed are adapted to each other. When network conditions are limited, the pressure of cloud phone image rendering is appropriately reduced, which avoids meaningless waste of GPU and maximizes the performance of GPU.

[0063] Figure 3 It is a schematic diagram of another cloud phone operation control method provided according to an embodiment of the present disclosure. This embodiment is refined based on the above embodiments. In this embodiment, the operation of "operating and controlling each GPU high-load application according to the operation description information of the cloud phone where each GPU high-load application is located" is concretized as follows: "Obtain the priority weight information of the cloud phone where each GPU high-load application is located; select at least one target cloud phone in order from low to high according to the priority weight information; adopt a preset strategy to reduce the resources in the target GPU allocated to each GPU high-load application in each target cloud phone."

[0064] Correspondingly, such as Figure 3 As shown, the method may specifically include:

[0065] S301. Obtain and monitor the usage rate of the target GPU corresponding to the cloud phone in real time.

[0066] S302: When it is determined that the usage rate of the target GPU exceeds the usage rate threshold, obtain each associated cloud phone that shares the target GPU with the monitoring cloud phone.

[0067] S303: Identify high-load applications of each GPU in the monitored cloud phone and each associated cloud phone.

[0068] S304, obtaining priority weight information of the cloud phone where each GPU high-load application is located.

[0069] Priority weight information can be specifically understood as: the priority and weight assigned to different applications or services in resource scheduling and task management. This information helps the system decide how to allocate limited resources, such as GPU, memory, storage space or network bandwidth. Priority determines which applications or services should get resources first when resources are limited. Applications with higher priorities (such as those involving critical business processing or frequent user interaction) will be given priority in resource allocation. Weight is an evaluation indicator when different applications have the same priority, which determines the relative share of applications in resource allocation. For example, applications with higher weights (such as those involving critical business processing or frequent user interaction) may get more resources to ensure stable operation of the application. In cloud phones, this setting of priority and weight allows the system to dynamically adjust resource allocation based on the importance of the application and the current performance requirements. For example, if a cloud phone is running a game with high-quality graphics and visual effects, the system may automatically increase the priority and weight of the game to ensure that it can obtain enough GPU resources to maintain a smooth gaming experience. On the contrary, if the network transmission speed is not enough to support the smooth transmission of high-resolution images, the system may reduce the priority and weight of applications with higher image quality requirements to ensure basic service quality and user experience.

[0070] S305. Filter out at least one target cloud phone according to the priority weight information in descending order.

[0071] S306. Using a preset strategy, reduce the resources in the target GPU allocated to the high-load applications of each GPU in each target cloud phone.

[0072] Specifically, after identifying each GPU high-load application in the monitoring cloud phone and each associated cloud phone, the priority weight information of the cloud phone where each GPU high-load application is located is obtained and sorted, and at least one cloud phone with a low priority weight is screened out as the target cloud phone, and the resources allocated to the target GPU are differentially reduced for the screened cloud phones according to the priority weight. If there are more than one target cloud phones screened out, the allocated GPU resources are differentially reduced according to the priority weight of the target cloud phone (for example, reducing the amount of video memory allocated to a specific application or limiting the number of GPU computing cores used by the application). For high-priority applications, that is, critical applications, it can be ensured that more GPU resources are still held after reducing the allocated GPU resources, and for low-priority applications, that is, non-critical applications, the amount of reduction in resource allocation can be adaptively increased to ensure that the performance of critical applications is not affected. When the priorities of applications are the same, the allocated GPU resources can be reduced differently according to their weights. For high-weight applications, i.e. critical applications, it can be ensured that more GPU resources are still retained after the allocated GPU resources are reduced. For low-weight applications, i.e. non-critical applications, the amount of resource allocation reduction can be adaptively increased to ensure that the performance of critical applications is not affected, ensure the stable operation of the target GPU, and improve the reliability of the overall operation of the cloud phone.

[0073] The technical solution of the disclosed embodiment obtains the usage rate of the target GPU corresponding to the monitored cloud phone in real time; when it is determined that the usage rate of the target GPU exceeds the usage rate threshold, obtains each associated cloud phone that shares the target GPU with the monitored cloud phone; identifies each GPU high-load application in the monitored cloud phone and each associated cloud phone; obtains the priority weight information of the cloud phone where each GPU high-load application is located; and selects at least one target cloud phone in the order of priority weight information from low to high; and adopts a preset strategy to reduce the technical means of resources in the target GPU allocated to each GPU high-load application in each target cloud phone, so as to realize real-time abnormal monitoring of the GPU usage rate of the cloud phone. When the GPU usage rate is abnormal, the high-load foreground applications of each GPU in the monitored cloud phone and each associated cloud phone are identified, and the resources allocated to the target GPU by the application of the target cloud phone are differentially reduced according to the priority weight of each application. In the case of tight GPU resources, it is ensured that high-priority weight applications can obtain sufficient resources to maintain a smooth user experience and application performance, avoid resource waste, prevent system overload, reduce system crashes or performance bottlenecks caused by resource competition, ensure the stable operation of the target GPU, and improve the reliability of the overall operation of the cloud phone.

[0074] In another optional implementation of this embodiment, the adopting of a preset strategy to reduce the resources in the target GPU allocated to each GPU high-load application in each target cloud phone includes at least one of the following:

[0075] Adjust the running mode of the GPU high-load application in the target cloud phone to a minimalist mode;

[0076] Turn off the high-consumption graphics effects in the GPU high-load application in the target cloud phone;

[0077] Reduce the resolution or frame rate of the GPU high-load application in the target cloud phone.

[0078] Specifically, in the case of GPU resource shortage, reducing the resources in the target GPU allocated to each GPU high-load application in each target cloud phone can provide additional available GPU resources for other applications by reducing the GPU resource requirements of the application, thereby realizing GPU resource redistribution. Among them, reducing the GPU resource requirements may include: adjusting the operation mode of the GPU high-load application in the target cloud phone to the minimalist mode, that is, adjusting the graphics settings of the application to the minimum, reducing the complexity of graphics rendering, thereby reducing the GPU load, for example, in the game, it may involve turning off shadows, reducing texture quality or reducing reflection effects, etc.; turning off high-consumption graphics special effects in the GPU high-load application in the target cloud phone, that is, turning off special effects such as ray tracing, anti-aliasing, complex particle effects, etc. that consume a lot of GPU resources to reduce the workload of the GPU; reducing the resolution or frame rate of the GPU high-load application in the target cloud phone, that is, by reducing the output resolution of the game or application, reducing the number of pixels rendered by the GPU, thereby reducing the GPU utilization rate, or by limiting the frame rate, reducing the number of frames that the GPU needs to render per second, thereby reducing the GPU utilization rate, balancing the GPU load, ensuring the stable operation of the target GPU, and improving the reliability of the overall operation of the cloud phone.

[0079] In another optional implementation of this embodiment, the obtaining the priority weight information of the cloud phone where each GPU high-load application is located includes:

[0080] Obtain the cloud phone user to which the cloud phone where each of the GPU high-load applications resides belongs;

[0081] Obtaining historical cloud phone usage frequency information and user level information of each of the cloud phone users;

[0082] Calculate the priority weight information of the cloud phone where each GPU high-load application is located according to the historical cloud phone usage frequency information and the user level information;

[0083] The historical cloud phone usage frequency information includes at least one of the historical usage duration, the historical cloud phone usage frequency, and the historical cloud phone usage high-frequency time period.

[0084] The historical cloud phone usage frequency information can be specifically understood as: recording the historical usage information of users using cloud phones, which can be used to understand the user's behavior pattern, and allocate resources and optimize services accordingly. The historical cloud phone usage frequency information includes at least one of the following: historical usage time, historical cloud phone usage frequency, and historical cloud phone usage high-frequency time period. Among them, the historical usage time reflects the total time the user uses the cloud phone, and the historical usage time will affect the user's priority weight, because users who use it for a long time may need more stable resource allocation; the historical cloud phone usage frequency reflects the frequency of users accessing the cloud phone service, and users with high usage frequency can be given higher priority weights; the historical cloud phone usage high-frequency time period reflects the user's usage peak in a specific time period, and users can be given higher priority weights in a specific time period, which helps to predict and allocate resources to meet peak demand.

[0085] Specifically, the priority weight information of the cloud phone where each GPU high-load application is located can be calculated through the historical cloud phone usage frequency information and user level information (referring to the user's subscription level or membership level, and users with high subscription levels or high membership levels can be given higher priority weights). After obtaining the cloud phone users to which the cloud phones where each GPU high-load application is located belong, the historical cloud phone usage frequency information and user level information of each cloud phone user are obtained, and the above indicators are quantified, such as converting the usage time and frequency into numerical values, and assigning different weights to different indicators. For example, if the user level is high, a higher weight can be given, and then weighted calculation is performed, and the priority weight is dynamically adjusted according to real-time data during the subsequent operation process to adapt to the changing usage pattern. By identifying the user's usage habits and preferences, more personalized services can be provided to them, ensuring that the resource needs of high-value users or high-frequency usage periods are met first, improving resource utilization efficiency, ensuring the stable operation of the target GPU, and improving the reliability of the overall operation of the cloud phone.

[0086] Based on the above embodiments, after obtaining and monitoring the usage rate of the target GPU corresponding to the cloud phone in real time, the method may further include:

[0087] When it is determined that the usage rate of the target GPU does not exceed the usage rate threshold, determining whether the usage rate of the target GPU shows an upward trend according to the historical usage rates of the target GPU obtained from at least two recent monitorings;

[0088] If yes, then obtaining the user historical behavior data of the cloud phone users to which the monitored cloud phone and each of the associated cloud phones belong;

[0089] Predicting the expected time of GPU usage exceeding the limit on the target GPU based on the historical behavior data of each user;

[0090] If the difference between the expected occurrence time and the current system time is less than a preset difference threshold, the expected occurrence time is provided to each of the cloud phone users for autonomous decision-making control.

[0091] Specifically, in the process of real-time acquisition of the usage rate of the target GPU corresponding to the monitored cloud phone, when the usage rate of the target GPU does not exceed the usage rate threshold, that is, it is in a normal working state, the usage rate of the target GPU is determined according to the historical usage rate of the target GPU obtained by at least two recent monitorings. If the usage rate of the target GPU is on a downward trend, it means that the target GPU has no overload risk; if the usage rate of the target GPU is on an upward trend, it means that the load of the target GPU is increasing, and it is also necessary to obtain the user historical behavior data of the cloud phone users to which the monitored cloud phone and each of the associated cloud phones belong to further determine whether the target GPU has an overload risk. According to the acquired user historical behavior data, the user's historical behavior data is analyzed by a machine learning algorithm to establish a prediction model, so as to identify in advance the situation that may cause GPU overload, and predict the expected occurrence time of the target GPU GPU exceeding the usage limit. When the difference between the expected occurrence time and the current system time is less than the preset difference threshold, the expected occurrence time is provided to each cloud phone user in the form of an abnormal warning for autonomous decision-making control. By analyzing the user's historical usage data in the cloud phone to predict the GPU overload time, and issuing a warning to the user before predicting a possible overload, the user experience can be effectively improved and the smoothness of the device performance can be ensured. This approach helps optimize resource allocation, reduce operating costs and system failure risks caused by overload, while improving system stability, effectively managing GPU load, and improving the reliability of cloud phone operation control.

[0092] Typically, if the usage rate of the target GPU is on an upward trend, historical user behavior data of cloud phone users can be obtained, such as the types of applications run by users, especially those with high GPU resource requirements, such as games or video editing software, as well as the frequency and duration of use of these applications, and the user's use of GPU resources when using cloud phones, including usage duration, frequency and time period. Combined with system performance feedback data, such as frame rate, latency and freeze, users are divided into different behavior groups through cluster analysis, and high-risk overload user groups are identified. A decision tree model is constructed to identify the key factors that cause GPU overload, and random forests are used in combination with time series analysis for ensemble learning. By analyzing the time series of user behavior data, future behavior patterns and potential GPU load peaks are predicted, and the estimated time of GPU overload is obtained. For the above-mentioned high-risk overload user groups, the difference between their respective expected occurrence time and the preset difference threshold is determined. When the expected occurrence time is less than the preset difference threshold, the high-risk overload users and their associated cloud phone users are given overload prompts, and the expected occurrence time is displayed to remind users to make independent decision-making controls, such as reminding users to reduce the use of resource-intensive applications to ensure the stable operation of cloud phones.

[0093] Based on the above embodiments, after obtaining and monitoring the usage rate of the target GPU corresponding to the cloud phone in real time, the method may further include:

[0094] When it is determined that the usage rate of the target GPU does not exceed the usage rate threshold, obtaining user historical behavior data of the cloud phone users to which the monitored cloud phone and each of the associated cloud phones belong, predicting the resources required by the cloud phone users to which the monitored cloud phone and each of the associated cloud phones belong, and using the remaining resources to perform intelligent preloading of multiple tasks simultaneously using multi-threading and asynchronous processing; or

[0095] When it is determined that the usage rate of the target GPU does not exceed the usage rate threshold, the operation mode of the GPU is dynamically adjusted according to the current usage rate of the target GPU and the preset threshold.

[0096] Specifically, in the process of obtaining the usage rate of the target GPU corresponding to the monitored cloud phone in real time, when the usage rate of the target GPU does not exceed the usage rate threshold, that is, it is in a normal working state, the user historical behavior data of the cloud phone users to which the monitored cloud phone and each of the associated cloud phones belong is obtained, and the user historical behavior data is analyzed through a machine learning algorithm to screen out the cloud phone users with high load demand expected to belong to the monitored cloud phone and each of the associated cloud phones, predict the resources required by these users, and use the remaining GPU resources to use multi-threading and asynchronous processing to achieve intelligent preloading of the above resources. In the process of obtaining the usage rate of the target GPU corresponding to the monitored cloud phone in real time, when the usage rate of the target GPU does not exceed the usage rate threshold, that is, it is in a normal working state, the GPU operation mode can also be dynamically adjusted according to the relationship between the current target GPU usage and the preset threshold. For example, according to the load situation, when the GPU usage is less than or equal to the low threshold, between the low threshold and the high threshold, and greater than or equal to the high threshold, the GPU operation mode can be set to low performance, medium performance or high performance mode respectively. Predicting the required resources through historical user behavior data and preloading them according to the prediction results, as well as dynamically adjusting the GPU working mode, can improve the system's response speed and resource utilization, optimize the user experience, ensure smooth performance in resource-intensive tasks, reduce energy consumption, and improve system stability. In addition, it helps achieve load balancing, supports high-concurrency processing, makes cloud phone management more intelligent, reduces the need for manual intervention, reduces operating costs, and improves the reliability of the cloud phone system.

[0097] Typically, when the target GPU is in a normal working state, the user history behavior data of the cloud phone user can be obtained, for example, the user's application usage history, especially the high-load applications with high frequency of use in the same historical time period, such as games or video editing software, etc., to obtain the usage frequency and duration of these applications, as well as the performance mode adopted by the user to use these applications, including but not limited to the operation mode, graphic effects, resolution and frame rate. Users are divided into different behavior groups through cluster analysis to identify users who may need more resources, user behavior is analyzed through time series to predict future resource requirements, and models are built through decision trees and random forests to identify key resources with high resource requirements. GPU asynchronous APIs are called and events are used to record and synchronize GPU tasks to query the completion status of GPU tasks and perform subsequent operations, so as to achieve preloading of predicted key resources for expected high-load users in the normal state of the GPU or low-load operation state, so that the system can respond to user operations faster, GPU resources are fully utilized, user waiting time is reduced, user experience is improved, and the risk of GPU overload is reduced, thereby improving the reliability of the cloud phone system.

[0098] Typically, when the target GPU is in normal working condition, the current utilization rate of the GPU is obtained and compared with the set threshold to adjust the working mode. For example, it can be set to low performance mode when the GPU load is less than 20%, medium performance mode when the load is between 20% and 80%, and high performance mode when the load exceeds 80%. In order to ensure stable operation of the system under sudden high load conditions, a certain proportion of GPU resources can be reserved and not allocated to user tasks to ensure that there are sufficient resources for rapid response under sudden high load conditions, maintain the stability of the cloud phone system in a short period of time, and avoid system crashes or performance degradation caused by overload. Warnings can also be sent to users to remind them that resources are currently tight and suggest that users reduce resource-intensive operations or optimize application configurations.

[0099] To facilitate understanding, specific application scenarios to which each embodiment of the present disclosure is applicable are now described. Figure 4 This is a schematic diagram of the overall process of a solution applicable to a specific application scenario of the disclosed embodiment. In this application scenario, it specifically describes the specific operation steps of real-time monitoring of the GPU usage rate of the cloud phone, and when it is detected that the GPU usage rate of the cloud phone exceeds the set alarm standard, reporting the information and GPU usage rate of the cloud phone to the cloud phone platform, and then the cloud phone platform notifies the cloud phone customer to which the cloud phone instance belongs of the relevant cloud phone information, and the cloud phone customer promptly perceives and handles this problem.

[0100] Specifically, Figure 4 As shown, the operation control method of the cloud phone is specifically divided into 7 steps: deploying a GPU monitoring program and configuring alarm standards, monitoring the GPU usage rate of the current cloud phone in real time, calling back relevant information to the cloud phone platform, querying relevant information based on the cloud phone's unique identifier, returning the query results, notifying the callback of all affected cloud phones to their respective merchants, and taking corresponding measures for cloud phones with abnormal GPU usage rates.

[0101] S401. Deploy a GPU monitoring program and configure alarm standards.

[0102] Optionally, you can deploy a GPU monitoring program on the cloud phone and set the GPU alarm standard and the notification interface address when an exception occurs. For example, you can set the alarm standard to GPU usage exceeding 90% within 3 seconds. The notification interface address is used to call back relevant information to the cloud phone platform through an interface call when the alarm standard is met.

[0103] S402, monitor the GPU usage rate of the current cloud phone in real time.

[0104] After the deployment of the GPU monitoring program is completed, the GPU monitoring program can monitor the GPU usage rate of the current cloud phone in real time according to a preset monitoring frequency.

[0105] S403. Call back relevant information to the cloud phone platform.

[0106] Specifically, when the GPU monitoring program detects that the GPU usage rate of the current cloud phone meets the set alarm standard, it will call back some relevant information to the cloud phone platform through the notification interface address. The relevant information includes the unique identifier of the current cloud phone (instance IP), GPU usage rate, and the name of the application package with higher load on the cloud phone.

[0107] S404. Query relevant information according to the unique identifier of the cloud phone.

[0108] Among them, when the cloud phone platform receives the callback notification of the GPU monitoring program, it can query the database according to the instance IP of the current cloud phone to obtain information such as the merchant (also called cloud phone customer) to which the current cloud phone belongs, the user ID who is using the current cloud phone, other cloud phones sharing the GPU with the current cloud phone and their merchants.

[0109] S405. Return the query result.

[0110] Among them, the database can return the queried relevant information to the cloud phone platform based on the query request of the cloud phone platform.

[0111] S406. Notify the merchants of all affected cloud phone callbacks.

[0112] Specifically, the current cloud phone and other cloud phones that share the GPU with the current cloud phone are all affected cloud phones. The cloud phone platform will notify the merchants of the affected cloud phones of the abnormal GPU usage.

[0113] S407. Take corresponding actions for the cloud phones with abnormal GPU usage.

[0114] After receiving the GPU abnormality notification, the merchant of the cloud phone can take some targeted measures to deal with this problem, such as killing the application process with high GPU usage and notifying its downstream users.

[0115] This technical solution can achieve real-time abnormal monitoring and notification capabilities for the GPU usage of cloud phones. When the GPU usage rate on a cloud phone soars, the merchant to which the cloud phone belongs can be notified in time, so that cloud phone customers on the cloud phone platform can perceive this problem in time and take corresponding measures, avoiding the impact of lag on the actual cloud phone user experience and improving the stability of the cloud phone system.

[0116] As an implementation of the above-mentioned operation control methods of each cloud phone, the present disclosure also provides an optional embodiment of an execution device for implementing the above-mentioned operation control methods of each cloud phone.

[0117] Figure 5 1 is a structural diagram of a cloud phone operation control device provided according to an embodiment of the present disclosure. Figure 5 The device comprises: a GPU usage monitoring module 501, an associated cloud phone acquisition module 502 and a GPU high-load application control module 503, wherein:

[0118] GPU usage monitoring module, used to obtain the usage of the target GPU corresponding to the monitored cloud phone in real time;

[0119] An associated cloud phone acquisition module, configured to acquire associated cloud phones that share the target GPU with the monitoring cloud phone when it is determined that the usage rate of the target GPU exceeds a usage rate threshold;

[0120] The GPU high-load application control module is used to identify each GPU high-load application in the monitoring cloud phone and each associated cloud phone, and to control the operation of each GPU high-load application according to the operation description information of the cloud phone where each GPU high-load application is located.

[0121] The technical solution of the disclosed embodiment can achieve real-time abnormal monitoring of the GPU usage of the cloud phone, identify the high-load applications of the GPU in the monitoring cloud phone and the cloud phone where the high-load applications are located, and control the operation of the high-load applications of the GPU in accordance with the operation description information of the cloud phone where the high-load applications are located. When the GPU usage is abnormal, identify the high-load applications of the GPU in the monitoring cloud phone and the cloud phone where the high-load applications are located, and control the operation of the high-load applications of the GPU. By monitoring the GPU usage of the cloud phone in real time, not only the tedious steps of manual inspection by the user are omitted, but also potential problems can be quickly discovered and handled. This active and automatic operation control mechanism ensures the stable operation of the target GPU and improves the reliability of the overall operation of the cloud phone.

[0122] On the basis of the above embodiments, the GPU high-load application control module is specifically used to:

[0123] Identify the applications running in the foreground in the monitoring cloud phone and each of the associated cloud phones respectively, and obtain the resource occupancy rate of each of the applications on the target GPU;

[0124] An application program whose resource occupancy rate exceeds an occupancy rate threshold is determined as the GPU high-load application.

[0125] On the basis of the above embodiments, the GPU high-load application control module is further used to:

[0126] According to the network transmission speed between the cloud phone where each GPU high-load application is located and each physical device currently establishing a communication connection, the resources in the target GPU are reallocated to each GPU high-load application.

[0127] On the basis of the above embodiments, the GPU high-load application control module is further used to:

[0128] Real-time acquisition of the current network transmission speed between the cloud phone where the current GPU high-load application is located and the physical device currently establishing a communication connection, and acquisition of the current frame rate of the current GPU high-load application;

[0129] If the current frame rate is greater than the current network transmission speed, the current frame rate is adjusted to match the current network transmission speed.

[0130] On the basis of the above embodiments, the GPU high-load application control module further includes:

[0131] A priority weight information acquisition unit, used to acquire the priority weight information of the cloud phone where each GPU high-load application is located;

[0132] A target cloud phone screening unit, configured to screen out at least one target cloud phone according to the priority weight information in descending order;

[0133] The resource reduction unit is used to adopt a preset strategy to reduce the resources in the target GPU allocated to each GPU high-load application in each target cloud phone.

[0134] Based on the above embodiments, the resource reduction unit is specifically used to implement at least one of the following functions:

[0135] Adjust the running mode of the GPU high-load application in the target cloud phone to a minimalist mode;

[0136] Turn off the high-consumption graphics effects in the GPU high-load application in the target cloud phone;

[0137] Reduce the resolution or frame rate of the GPU high-load application in the target cloud phone.

[0138] On the basis of the above embodiments, the priority weight information acquisition unit is specifically used to:

[0139] Obtain the cloud phone user to which the cloud phone where each of the GPU high-load applications resides belongs;

[0140] Obtaining historical cloud phone usage frequency information and user level information of each of the cloud phone users;

[0141] Calculate the priority weight information of the cloud phone where each GPU high-load application is located according to the historical cloud phone usage frequency information and the user level information;

[0142] The historical cloud phone usage frequency information includes at least one of the historical usage duration, the historical cloud phone usage frequency, and the historical cloud phone usage high-frequency time period.

[0143] Based on the above embodiments, the device further includes:

[0144] A usage trend determination module, configured to determine whether the usage rate of the target GPU shows an upward trend based on the historical usage rates of the target GPU obtained from at least two recent monitorings when determining that the usage rate of the target GPU does not exceed the usage rate threshold;

[0145] A historical behavior data acquisition module, used to acquire user historical behavior data of the cloud phone users to which the monitored cloud phone and each of the associated cloud phones belong if the usage rate of the target GPU shows an upward trend;

[0146] An over-limit time prediction module, used to predict the expected occurrence time of GPU usage over-limit of the target GPU according to the historical behavior data of each user;

[0147] The estimated occurrence time providing module is used to provide the estimated occurrence time to each cloud phone user for autonomous decision-making control if the difference between the estimated occurrence time and the current system time is less than a preset difference threshold.

[0148] Based on the above embodiments, the device further includes:

[0149] An intelligent preloading module is used to obtain user historical behavior data of cloud phone users to which the monitored cloud phone and each of the associated cloud phones belong when it is determined that the usage rate of the target GPU does not exceed the usage rate threshold, predict the resources required by the cloud phone users to which the monitored cloud phone and each of the associated cloud phones belong, and use the remaining resources to perform intelligent preloading of multiple tasks simultaneously using multi-threading and asynchronous processing; or

[0150] The operation mode adjustment module is used to dynamically adjust the operation mode of the GPU according to the current usage rate of the target GPU and a preset threshold when it is determined that the usage rate of the target GPU does not exceed the usage rate threshold.

[0151] The above-mentioned product can execute the method provided by any embodiment of the present disclosure, and has the corresponding functional modules and beneficial effects of the execution method.

[0152] In the technical solution of the present disclosure, the collection, storage, use, processing, transmission, provision and disclosure of user personal information involved are in compliance with the provisions of relevant laws and regulations and do not violate public order and good morals.

[0153] According to an embodiment of the present disclosure, the present disclosure also provides an electronic device, a readable storage medium and a computer program product.

[0154] Figure 6 A schematic block diagram of an example electronic device 600 that can be used to implement an embodiment of the present disclosure is shown. The electronic device is intended to represent various forms of digital computers, such as laptop computers, desktop computers, workstations, personal digital assistants, servers, blade servers, mainframe computers, and other suitable computers. The electronic device can also represent various forms of mobile devices, such as personal digital processing, cellular phones, smart phones, wearable devices, and other similar computing devices. The components shown herein, their connections and relationships, and their functions are merely examples and are not intended to limit the implementation of the present disclosure described and / or required herein.

[0155] like Figure 6 As shown, the device 600 includes a computing unit 601, which can perform various appropriate actions and processes according to a computer program stored in a read-only memory (ROM) 602 or a computer program loaded from a storage unit 608 into a random access memory (RAM) 603. In the RAM 603, various programs and data required for the operation of the device 600 can also be stored. The computing unit 601, the ROM 602, and the RAM 603 are connected to each other via a bus 604. An input / output (I / O) interface 605 is also connected to the bus 604.

[0156] A number of components in the device 600 are connected to the I / O interface 605, including: an input unit 606, such as a keyboard, a mouse, etc.; an output unit 607, such as various types of displays, speakers, etc.; a storage unit 608, such as a disk, an optical disk, etc.; and a communication unit 609, such as a network card, a modem, a wireless communication transceiver, etc. The communication unit 609 allows the device 600 to exchange information / data with other devices through a computer network such as the Internet and / or various telecommunication networks.

[0157] The computing unit 601 may be a variety of general and / or special processing components with processing and computing capabilities. Some examples of the computing unit 601 include, but are not limited to, a central processing unit (CPU), a graphics processing unit (GPU), various dedicated artificial intelligence (AI) computing chips, various computing units running machine learning model algorithms, digital signal processors (DSPs), and any appropriate processors, controllers, microcontrollers, etc. The computing unit 601 performs the various methods and processes described above, such as the operation control method of the cloud phone. For example, in some embodiments, the operation control method of the cloud phone may be implemented as a computer software program, which is tangibly contained in a machine-readable medium, such as a storage unit 608. In some embodiments, part or all of the computer program may be loaded and / or installed on the device 600 via ROM 602 and / or communication unit 609. When the computer program is loaded into RAM 603 and executed by the computing unit 601, one or more steps of the operation control method of the cloud phone described above may be performed. Alternatively, in other embodiments, the computing unit 601 may be configured to execute the operation control method of the cloud phone in any other appropriate manner (e.g., by means of firmware).

[0158] That is, the usage rate of the target GPU corresponding to the monitored cloud phone is obtained in real time;

[0159] When it is determined that the usage rate of the target GPU exceeds the usage rate threshold, obtaining each associated cloud phone that shares the target GPU with the monitoring cloud phone;

[0160] Each GPU high-load application is identified in the monitoring cloud phone and each associated cloud phone, and the operation of each GPU high-load application is controlled according to the operation description information of the cloud phone where each GPU high-load application is located.

[0161] Various implementations of the systems and techniques described above herein can be implemented in digital electronic circuit systems, integrated circuit systems, field programmable gate arrays (FPGAs), application specific integrated circuits (ASICs), application specific standard products (ASSPs), systems on chips (SOCs), complex programmable logic devices (CPLDs), computer hardware, firmware, software, and / or combinations thereof. These various implementations can include: being implemented in one or more computer programs that can be executed and / or interpreted on a programmable system including at least one programmable processor, which can be a special purpose or general purpose programmable processor that can receive data and instructions from a storage system, at least one input device, and at least one output device, and transmit data and instructions to the storage system, the at least one input device, and the at least one output device.

[0162] The program code for implementing the method of the present disclosure may be written in any combination of one or more programming languages. These program codes may be provided to a processor or controller of a general-purpose computer, a special-purpose computer, or other programmable data processing device, so that the program code, when executed by the processor or controller, enables the functions / operations specified in the flow chart and / or block diagram to be implemented. The program code may be executed entirely on the machine, partially on the machine, partially on the machine and partially on a remote machine as a stand-alone software package, or entirely on a remote machine or server.

[0163] In the context of the present disclosure, a machine-readable medium may be a tangible medium that may contain or store a program for use by or in conjunction with an instruction execution system, device, or equipment. A machine-readable medium may be a machine-readable signal medium or a machine-readable storage medium. A machine-readable medium may include, but is not limited to, an electronic, magnetic, optical, electromagnetic, infrared, or semiconductor system, device, or equipment, or any suitable combination of the foregoing. A more specific example of a machine-readable storage medium may include an electrical connection based on one or more lines, a portable computer disk, a hard disk, a random access memory (RAM), a read-only memory (ROM), an erasable programmable read-only memory (EPROM or flash memory), an optical fiber, a portable compact disk read-only memory (CD-ROM), an optical storage device, a magnetic storage device, or any suitable combination of the foregoing.

[0164] To provide interaction with a user, the systems and techniques described herein can be implemented on a computer having: a display device (e.g., a CRT (cathode ray tube) or LCD (liquid crystal display) monitor) for displaying information to the user; and a keyboard and pointing device (e.g., a mouse or trackball) through which the user can provide input to the computer. Other types of devices can also be used to provide interaction with the user; for example, the feedback provided to the user can be any form of sensory feedback (e.g., visual feedback, auditory feedback, or tactile feedback); and input from the user can be received in any form (including acoustic input, voice input, or tactile input).

[0165] The systems and techniques described herein may be implemented in a computing system that includes back-end components (e.g., as a data server), or a computing system that includes middleware components (e.g., an application server), or a computing system that includes front-end components (e.g., a user computer with a graphical user interface or a web browser through which a user can interact with implementations of the systems and techniques described herein), or a computing system that includes any combination of such back-end components, middleware components, or front-end components. The components of the system may be interconnected by any form or medium of digital data communication (e.g., a communication network). Examples of communication networks include: a local area network (LAN), a wide area network (WAN), and the Internet.

[0166] A computer system may include a client and a server. The client and the server are generally remote from each other and usually interact through a communication network. The relationship between the client and the server is generated by computer programs running on the corresponding computers and having a client-server relationship with each other. The server may be a cloud server, also known as a cloud computing server or a cloud host, which is a host product in the cloud computing service system to solve the defects of difficult management and weak business scalability in traditional physical hosts and VPS services. The server may also be a server of a distributed system, or a server combined with a blockchain.

[0167] Artificial intelligence is a discipline that studies how to use computers to simulate certain human thought processes and intelligent behaviors (such as learning, reasoning, thinking, planning, etc.). It includes both hardware-level and software-level technologies. Artificial intelligence hardware technologies generally include technologies such as sensors, dedicated artificial intelligence chips, cloud computing, distributed storage, and big data processing; artificial intelligence software technologies mainly include computer vision technology, speech recognition technology, natural language processing technology, machine learning / deep learning technology, big data processing technology, knowledge graph technology, and other major directions.

[0168] Cloud computing refers to a technology system that uses network access to elastically scalable shared physical or virtual resource pools. Resources can include servers, operating systems, networks, software, applications, and storage devices, and can be deployed and managed on demand and in a self-service manner. Cloud computing technology can provide efficient and powerful data processing capabilities for technical applications such as artificial intelligence and blockchain, as well as model training.

[0169] It should be understood that the various forms of processes shown above can be used to reorder, add or delete steps. For example, the steps recorded in this disclosure can be executed in parallel, sequentially or in different orders, as long as the desired results of the technical solutions provided by this disclosure can be achieved, and this document does not limit this.

[0170] The above specific implementations do not constitute a limitation on the protection scope of the present disclosure. It should be understood by those skilled in the art that various modifications, combinations, sub-combinations and substitutions can be made according to design requirements and other factors. Any modification, equivalent substitution and improvement made within the spirit and principle of the present disclosure shall be included in the protection scope of the present disclosure.

Claims

1. A method for controlling the operation of a cloud phone, comprising: Obtain the usage rate of the target graphics processor GPU corresponding to the monitored cloud phone in real time; When it is determined that the usage rate of the target GPU exceeds the usage rate threshold, obtaining each associated cloud phone that shares the target GPU with the monitoring cloud phone; Each GPU high-load application is identified in the monitoring cloud phone and each associated cloud phone, and the operation of each GPU high-load application is controlled according to the operation description information of the cloud phone where each GPU high-load application is located.

2. The method according to claim 1, wherein: The step of identifying each GPU high-load application in the monitoring cloud phone and each associated cloud phone includes: Identify the applications running in the foreground in the monitoring cloud phone and each of the associated cloud phones respectively, and obtain the resource occupancy rate of each of the applications on the target GPU; An application program whose resource occupancy rate exceeds an occupancy rate threshold is determined as the GPU high-load application.

3. The method according to claim 1, wherein: The step of controlling the operation of each of the GPU high-load applications according to the operation description information of the cloud phone where each of the GPU high-load applications is located includes: According to the network transmission speed between the cloud phone where each GPU high-load application is located and each physical device currently establishing a communication connection, the resources in the target GPU are reallocated to each GPU high-load application.

4. The method according to claim 3, wherein: The re-allocating resources in the target GPU for each GPU high-load application according to the network transmission speed between the cloud phone where each GPU high-load application is located and each entity device currently establishing a communication connection comprises: Real-time acquisition of the current network transmission speed between the cloud phone where the current GPU high-load application is located and the physical device currently establishing a communication connection, and acquisition of the current frame rate of the current GPU high-load application; If the current frame rate is greater than the current network transmission speed, the current frame rate is adjusted to match the current network transmission speed.

5. The method according to claim 1, wherein: The step of controlling the operation of each of the GPU high-load applications according to the operation description information of the cloud phone where each of the GPU high-load applications is located includes: Obtaining priority weight information of the cloud phones where the GPU high-load applications are located; Filtering out at least one target cloud phone according to the priority weight information in descending order; A preset strategy is adopted to reduce the resources in the target GPU allocated to the high-load application of each GPU in each target cloud phone.

6. The method according to claim 5, wherein: The adopting of a preset strategy to reduce the resources in the target GPU allocated to each GPU high-load application in each target cloud phone includes at least one of the following: Adjust the running mode of the GPU high-load application in the target cloud phone to a minimalist mode; Turn off the high-consumption graphics effects in the GPU high-load application in the target cloud phone; Reduce the resolution or frame rate of the GPU high-load application in the target cloud phone.

7. The method according to claim 5, wherein: The obtaining of priority weight information of the cloud phone where each GPU high-load application is located includes: Obtain the cloud phone user to which the cloud phone where each of the GPU high-load applications resides belongs; Obtaining historical cloud phone usage frequency information and user level information of each of the cloud phone users; Calculate the priority weight information of the cloud phone where each GPU high-load application is located according to the historical cloud phone usage frequency information and the user level information; The historical cloud phone usage frequency information includes at least one of the historical usage duration, the historical cloud phone usage frequency, and the historical cloud phone usage high-frequency time period.

8. The method according to any one of claims 1 to 7, after obtaining and monitoring the usage rate of the target GPU corresponding to the cloud phone in real time, further comprising: When it is determined that the usage rate of the target GPU does not exceed the usage rate threshold, determining whether the usage rate of the target GPU shows an upward trend according to the historical usage rates of the target GPU obtained from at least two recent monitorings; If yes, then obtaining the user historical behavior data of the cloud phone users to which the monitored cloud phone and each of the associated cloud phones belong; Predicting the expected time of GPU usage exceeding the limit on the target GPU based on the historical behavior data of each user; If the difference between the expected occurrence time and the current system time is less than a preset difference threshold, the expected occurrence time is provided to each of the cloud phone users for autonomous decision-making control.

9. The method according to any one of claims 1 to 7, after acquiring and monitoring the usage rate of the target GPU corresponding to the cloud phone in real time, further comprising: When it is determined that the usage rate of the target GPU does not exceed the usage rate threshold, the user historical behavior data of the cloud phone users to which the monitored cloud phone and each of the associated cloud phones belong is obtained, the resources required by the cloud phone users to which the monitored cloud phone and each of the associated cloud phones belong are predicted, and the remaining resources are used to perform intelligent preloading of multiple tasks simultaneously using multi-threading and asynchronous processing; or When it is determined that the usage rate of the target GPU does not exceed the usage rate threshold, the operation mode of the GPU is dynamically adjusted according to the current usage rate of the target GPU and the preset threshold.

10. A cloud phone operation control device, comprising: The GPU usage monitoring module is used to obtain the usage rate of the target graphics processor GPU corresponding to the monitored cloud phone in real time; An associated cloud phone acquisition module, configured to acquire associated cloud phones that share the target GPU with the monitoring cloud phone when it is determined that the usage rate of the target GPU exceeds a usage rate threshold; The GPU high-load application control module is used to identify each GPU high-load application in the monitoring cloud phone and each associated cloud phone, and to control the operation of each GPU high-load application according to the operation description information of the cloud phone where each GPU high-load application is located.

11. The device according to claim 10, wherein: The GPU high-load application control module is specifically used for: Identify the applications running in the foreground in the monitoring cloud phone and each of the associated cloud phones respectively, and obtain the resource occupancy rate of each of the applications on the target GPU; An application program whose resource occupancy rate exceeds an occupancy rate threshold is determined as the GPU high-load application.

12. The device according to claim 10, wherein: The GPU high-load application control module is further used to: According to the network transmission speed between the cloud phone where each GPU high-load application is located and each physical device currently establishing a communication connection, the resources in the target GPU are reallocated to each GPU high-load application.

13. The device according to claim 12, wherein: The GPU high-load application control module is further used to: Real-time acquisition of the current network transmission speed between the cloud phone where the current GPU high-load application is located and the physical device currently establishing a communication connection, and acquisition of the current frame rate of the current GPU high-load application; If the current frame rate is greater than the current network transmission speed, the current frame rate is adjusted to match the current network transmission speed.

14. The device according to claim 10, wherein: The GPU high-load application control module further includes: A priority weight information acquisition unit, used to acquire the priority weight information of the cloud phone where each GPU high-load application is located; A target cloud phone screening unit, configured to screen out at least one target cloud phone according to the priority weight information in descending order; The resource reduction unit is used to adopt a preset strategy to reduce the resources in the target GPU allocated to each GPU high-load application in each target cloud phone.

15. The device according to claim 14, wherein: The resource reduction unit is specifically configured to implement at least one of the following functions: Adjust the running mode of the GPU high-load application in the target cloud phone to a minimalist mode; Turn off the high-consumption graphics effects in the GPU high-load application in the target cloud phone; Reduce the resolution or frame rate of the GPU high-load application in the target cloud phone.

16. The device according to claim 14, wherein: The priority weight information acquisition unit is specifically used to: Obtain the cloud phone user to which the cloud phone where each of the GPU high-load applications resides belongs; Obtaining historical cloud phone usage frequency information and user level information of each of the cloud phone users; Calculate the priority weight information of the cloud phone where each GPU high-load application is located according to the historical cloud phone usage frequency information and the user level information; The historical cloud phone usage frequency information includes at least one of the historical usage duration, the historical cloud phone usage frequency, and the historical cloud phone usage high-frequency time period.

17. The device according to any one of claims 10 to 16, further comprising: A usage trend determination module, configured to determine whether the usage rate of the target GPU shows an upward trend based on the historical usage rates of the target GPU obtained from at least two recent monitorings when determining that the usage rate of the target GPU does not exceed the usage rate threshold; A historical behavior data acquisition module, used to acquire user historical behavior data of the cloud phone users to which the monitored cloud phone and each of the associated cloud phones belong if the usage rate of the target GPU shows an upward trend; An over-limit time prediction module, used to predict the expected time when the target GPU will exceed the GPU usage limit based on the historical behavior data of each user; The estimated occurrence time providing module is used to provide the estimated occurrence time to each cloud phone user for autonomous decision-making control if the difference between the estimated occurrence time and the current system time is less than a preset difference threshold.

18. The device according to any one of claims 10 to 16, further comprising: The intelligent preloading module is used to obtain the user historical behavior data of the cloud phone users to which the monitored cloud phone and each of the associated cloud phones belong when it is determined that the usage rate of the target GPU does not exceed the usage rate threshold, predict the resources required by the cloud phone users to which the monitored cloud phone and each of the associated cloud phones belong, and use the remaining resources to perform intelligent preloading of multiple tasks simultaneously using multi-threading and asynchronous processing; or The operation mode adjustment module is used to dynamically adjust the operation mode of the GPU according to the current usage rate of the target GPU and a preset threshold when it is determined that the usage rate of the target GPU does not exceed the usage rate threshold.

19. An electronic device comprising: at least one processor; as well as a memory communicatively connected to the at least one processor; wherein, The memory stores instructions that can be executed by the at least one processor, and the instructions are executed by the at least one processor to enable the at least one processor to perform the method according to any one of claims 1 to 9.

20. A non-transitory computer-readable storage medium storing computer instructions, wherein: The computer instructions are used to make a computer execute the method according to any one of claims 1-9.

21. A computer program product, comprising computer instructions, which, when executed by a processor, implement the steps of the method according to any one of claims 1 to 9.

Images