Cloud application running method and device based on GPU cloud desktop platform, equipment and medium

Abstract

The application relates to a GPU cloud desktop platform-based cloud application running method, device, equipment, medium and computer program product. The method comprises the following steps: in response to a starting event of a target cloud application requiring a graphics processor on a CPU cloud desktop connected by a terminal, acquiring a preset graphics computing power requirement value of the target cloud application; according to the preset graphics computing power requirement value, determining a target graphics processor from a preconfigured graphics processor sharing pool, so that a remaining graphics computing power value of the target graphics processor is greater than or equal to the preset graphics computing power requirement value; starting the target cloud application on a GPU cloud desktop platform, connecting the target cloud application with the terminal directly, and responding to a graphics computing power request of the target cloud application through the target graphics processor. The method can improve the resource utilization rate of a graphics manager.

Description

TECHNICAL FIELD

[0001] The present application relates to the technical field of cloud desktop, in particular to a cloud application running method and device based on a GPU cloud desktop platform, equipment, medium and computer program product. BACKGROUND

[0002] With the development of computer technology, cloud desktop appears. The cloud desktop is a new mode to replace the traditional computer. In the running mode of the cloud desktop, the user does not need to purchase a computer host, and the components such as CPU (Central Processing Unit), memory and hard disk contained in the host of the cloud desktop are all virtually generated in the server in the back end. The user only needs to access the virtual host on the back-end server through the client on the thin terminal to realize interactive operation, and achieve the same experience effect as the computer.

[0003] The traditional cloud desktop only carries CPU for operation processing due to the limitation of early technology. With the gradual popularization of the cloud desktop, users have higher requirements for the graphical presentation and rendering performance of the cloud desktop. Therefore, the traditional cloud desktop carrying only CPU for operation processing is insufficient to meet the needs of users, and the cloud desktop carrying GPU (Graphics Processing Unit) (i.e. GPU cloud desktop) emerges as the times require.

[0004] At present, the implementation of the GPU cloud desktop needs to bind the GPU for the cloud desktop. In order to support various cloud applications, it is necessary to prepare enough GPUs for running most cloud applications for binding. However, the actual performance consumed in actual operation is limited, resulting in resource waste. SUMMARY

[0005] Therefore, it is necessary to provide a cloud application running method, device, equipment, medium and computer program product based on a GPU cloud desktop platform, which can improve the resource utilization rate of the graphics manager.

[0006] In a first aspect, the present application provides a cloud application running method based on a GPU cloud desktop platform. The method comprises:

[0007] In response to a start event of a target cloud application requiring a graphics processor on a CPU cloud desktop connected by a terminal, obtaining a preset graphics computing power requirement value of the target cloud application;

[0008] According to the preset graphics computing power requirement value, determining a target graphics processor from a pre-configured graphics processor shared pool, so that the remaining graphics computing power value of the target graphics processor is greater than or equal to the preset graphics computing power requirement value;

[0009] start the target cloud application on the GPU cloud desktop platform, directly connect the target cloud application with the terminal, and respond to a graphics computing power request of the target cloud application through the target graphics processor.

[0010] In one of the embodiments, the start event of the target cloud application is triggered in response to a triggering operation on identification information of the target cloud application in an application panel, the application panel being provided by the CPU cloud desktop to the terminal for display, and the application panel displaying identification information of at least one cloud application requiring a graphics processor.

[0011] In one of the embodiments, the determining of the target graphics processor from the preconfigured graphics processor sharing pool according to the preset graphics computing power requirement value, so that the remaining graphics computing power value of the target graphics processor is greater than or equal to the preset graphics computing power requirement value, comprises:

[0012] obtaining the remaining graphics computing power values of the plurality of graphics processors in the preconfigured graphics processor sharing pool;

[0013] screening out the graphics processors with the remaining graphics computing power values greater than or equal to the preset graphics computing power requirement value;

[0014] selecting the target graphics processor from the screened graphics processors according to a preset selection condition.

[0015] In one of the embodiments, the selecting of the target graphics processor from the screened graphics processors according to the preset selection condition comprises:

[0016] sorting the screened graphics processors according to a preset sorting condition based on the remaining graphics computing power values of the screened graphics processors;

[0017] determining the graphics processor with the first position in the sorting result as the target graphics processor.

[0018] In one of the embodiments, the starting of the target cloud application on the GPU cloud desktop platform, the directly connecting of the target cloud application with the terminal, and the responding to the graphics computing power request of the target cloud application through the target graphics processor comprise:

[0019] obtaining identity information of a target object triggering the start event;

[0020] performing authentication processing based on the identity information to obtain an authentication result;

[0021] When the authentication result represents that the authentication is passed, the target cloud application corresponding to the identity information of the target object is started on the GPU cloud desktop platform, the started target cloud application is directly connected with the terminal, and the target cloud application is responded to by the target graphics processor in response to a graphics computing power request of the target cloud application.

[0022] In one of the embodiments, the method further comprises:

[0023] When the CPU cloud desktop obtains a data file transmitted by a hardware interface of the terminal, and stores the data file in a cloud storage device corresponding to the identity information, the GPU cloud desktop platform accesses the data file stored in the cloud storage device corresponding to the identity information after receiving an access request for the data file triggered by the terminal with the identity information.

[0024] In a second aspect, the application further provides a cloud application running device based on a GPU cloud desktop platform.

[0025] The device comprises:

[0026] A graphics processor allocation module is configured to, in response to a starting event of a target cloud application requiring a graphics processor on a CPU cloud desktop connected with a terminal, obtain a preset graphics computing power requirement value of the target cloud application, and determine a target graphics processor from a preconfigured graphics processor sharing pool according to the preset graphics computing power requirement value, so that a remaining graphics computing power value of the target graphics processor is greater than or equal to the preset graphics computing power requirement value.

[0027] A graphics computing power request response module is configured to start the target cloud application on the GPU cloud desktop platform, directly connect the target cloud application with the terminal, and respond to a graphics computing power request of the target cloud application by the target graphics processor.

[0028] In a third aspect, the application further provides a computer device. The computer device comprises a memory and a processor, the memory stores a computer program, and the processor implements the following steps when executing the computer program:

[0029] In response to a starting event of a target cloud application requiring a graphics processor on a CPU cloud desktop connected with a terminal, a preset graphics computing power requirement value of the target cloud application is obtained.

[0030] According to the preset graphics computing power requirement value, a target graphics processor is determined from a preconfigured graphics processor sharing pool, so that a remaining graphics computing power value of the target graphics processor is greater than or equal to the preset graphics computing power requirement value.

[0031] start the target cloud application on the GPU cloud desktop platform, directly connect the target cloud application with the terminal, and respond to a graphics computing power request of the target cloud application through the target graphics processor.

[0032] In a fourth aspect, the present application also provides a computer readable storage medium. The computer readable storage medium has a computer program stored thereon, and the computer program, when executed by a processor, implements the following steps:

[0033] In response to a start event of a target cloud application requiring a graphics processor on a CPU cloud desktop connected by a terminal, obtaining a preset graphics computing power requirement value of the target cloud application;

[0034] According to the preset graphics computing power requirement value, determining a target graphics processor from a preconfigured graphics processor sharing pool, so that a remaining graphics computing power value of the target graphics processor is greater than or equal to the preset graphics computing power requirement value;

[0035] starting the target cloud application on the GPU cloud desktop platform, directly connecting the target cloud application with the terminal, and responding to a graphics computing power request of the target cloud application through the target graphics processor.

[0036] In a fifth aspect, the present application also provides a computer program product. The computer program product comprises a computer program, and the computer program, when executed by a processor, implements the following steps:

[0037] In response to a start event of a target cloud application requiring a graphics processor on a CPU cloud desktop connected by a terminal, obtaining a preset graphics computing power requirement value of the target cloud application;

[0038] According to the preset graphics computing power requirement value, determining a target graphics processor from a preconfigured graphics processor sharing pool, so that a remaining graphics computing power value of the target graphics processor is greater than or equal to the preset graphics computing power requirement value;

[0039] starting the target cloud application on the GPU cloud desktop platform, directly connecting the target cloud application with the terminal, and responding to a graphics computing power request of the target cloud application through the target graphics processor.

[0040] The cloud application running method, device, equipment, medium and computer program product based on the GPU cloud desktop platform can trigger a starting event of a target cloud application through a CPU cloud desktop, the GPU cloud desktop platform starts the target cloud application, and the target cloud application can be run when the CPU cloud desktop is used by responding to a graphics computing power request of the target cloud application through a target graphics processor, which creates conditions for better graphics presentation and rendering performance when the target cloud application is run. Moreover, the target graphics processor does not need to be bound to the CPU cloud desktop, but is determined from a graphics processor shared pool shared by multiple CPU cloud desktops according to a preset graphics computing power requirement value of the target cloud application. The remaining graphics computing power value of the target graphics processor can match the preset graphics computing power requirement value of the target cloud application, which can support running the target cloud application, and realizes flexible scheduling of multiple graphics processors with high resource utilization. BRIEF DESCRIPTION OF DRAWINGS

[0041] Figure 1 An application environment diagram of the cloud application running method based on the GPU cloud desktop platform in one embodiment;

[0042] Figure 2 A flowchart of the cloud application running method based on the GPU cloud desktop platform in one embodiment;

[0043] Figure 3 A flowchart of the step of determining the target graphics processor according to the preset graphics computing power requirement value from the preconfigured graphics processor shared pool in one embodiment;

[0044] Figure 4 A specific application scenario diagram of the cloud application running method based on the GPU cloud desktop platform in one embodiment;

[0045] Figure 5 A structural block diagram of the cloud application running device based on the GPU cloud desktop platform in one embodiment;

[0046] Figure 6 An internal structure diagram of the computer equipment in one embodiment. DETAILED DESCRIPTION

[0047] In order to make the purpose, technical scheme and advantages of the present application clearer, further detailed description of the present application will be given below in combination with the drawings and embodiments. It should be understood that the specific embodiments described herein are only used to explain the present application and do not limit the present application.

[0048] The cloud application running method based on the GPU cloud desktop platform provided by the embodiments of the present application can be applied to, for example Figure 1The application environment shown. Among them, the terminal 102 can operate the CPU cloud desktop 104 through the remote network, the GPU cloud desktop platform 106 can communicate with the terminal 102 and the CPU cloud desktop 104 through the network, the GPU cloud desktop platform 106 can run on the platform server, for managing at least one graphics processor configured on the platform server including the target graphics processor, at least one graphics processor constitutes a graphics processor shared pool. Specifically, the GPU cloud desktop platform 106 can respond to the terminal 102 through the remote network operation and trigger the start event of the target cloud application on the CPU cloud desktop 104, determine the target graphics processor from the graphics processor shared pool according to the preset graphics computing power requirement value of the target cloud application, start the target cloud application on the GPU cloud desktop platform 106, connect the target cloud application with the running client on the terminal 102 directly, and respond to the graphics computing power request of the target cloud application through the target graphics processor.

[0049] Among them, the terminal 102 plays the role of thin client in the cloud desktop scenario, which can be various desktop computers, notebook computers, smart phones, tablet computers or special cloud terminals, etc. The CPU cloud desktop 104 can be virtualized by a cloud server, which can be realized by an independent server or a server cluster composed of multiple servers. The GPU cloud desktop platform 106 can be a server running on a platform server, which can be realized by an independent server or a server cluster composed of multiple servers. The data storage system can store the data processed by the cloud application running method based on the GPU cloud desktop platform. The data storage system can be integrated on the platform server where the GPU cloud desktop platform 106 is located, or placed on the cloud or other servers.

[0050] In one embodiment, as Figure 2 shown, a cloud application running method based on a GPU cloud desktop platform is provided. This embodiment takes the GPU cloud desktop platform 106 in the application environment as an example to illustrate the method, which includes the following steps: Figure 1

[0051] Step 202, in response to the start event of the target cloud application needing to use the graphics processor on the CPU cloud desktop connected by the terminal, obtaining the preset graphics computing power requirement value of the target cloud application.

[0052] Among them, the CPU (central processing unit) cloud desktop is a virtual computer virtualized by a server, and the CPU cloud desktop runs cloud applications based on CPU. The graphics processor is a processor specially used for image and graphics related operation work. The graphics processor can be various consumer-level graphics processors, enterprise-level graphics processors, automotive-level graphics processors or others.

[0053] ​The cloud application is software installed and running on the cloud desktop. The cloud application requiring a graphics processor refers to software requiring processing of graphics tasks by a graphics processor, such as design software, cloud game software, ultra-high-definition video software, or the like. The target cloud application is the cloud application requiring a graphics processor to which the method is directed. The preset graphics computing power requirement value is a value of the graphics processor computing power expected to be consumed, which can be determined based on the resource usage rate of the target cloud application expected to occupy the graphics processor.

[0054] The start event of the target cloud application refers to an event of starting the target cloud application to run the target cloud application, which can be a manual trigger operation or an automatic trigger event. The manual trigger operation is, for example, a click on the identification information of the target cloud application. The automatic trigger event can be automatically triggered when the CPU cloud desktop is started, or can be automatically triggered after the target cloud application is abnormally closed and restored to normal.

[0055] In one embodiment, the GPU cloud desktop platform can query, in response to the start event of the target cloud application on the CPU cloud desktop, from a quick query table recording the preset graphics computing power requirement value of the target cloud application, the preset graphics computing power requirement value of the target cloud application when the preset graphics computing power requirement value of the target cloud application is queried.

[0056] In one embodiment, the GPU cloud desktop platform can trigger, in response to a trigger operation on the identification information of the target cloud application in the application panel on the CPU cloud desktop, a start event of the target cloud application installed on the GPU cloud desktop platform, and obtain the preset graphics computing power requirement value of the target cloud application.

[0057] Step 204, according to the preset graphics computing power requirement value, determining the target graphics processor from the pre-configured graphics processor sharing pool, so that the remaining graphics computing power value of the target graphics processor is greater than or equal to the preset graphics computing power requirement value.

[0058] The graphics processor sharing pool is a collection of graphics processors shared by multiple CPU cloud desktops. The graphics processor sharing pool includes at least one graphics processor, and when including one graphics processor, the graphics processor can be shared by multiple cloud applications. The target graphics processor is a graphics processor selected from the graphics processor sharing pool to respond to the graphics computing power request of the target cloud application. The remaining graphics computing power value is the graphics computing power value remaining in the current running state of the corresponding graphics processor. The remaining graphics computing power value can be determined based on the usage rate of the corresponding graphics processor.

[0059] In an embodiment, the GPU cloud desktop platform can determine at least one to-be-assigned graphics processor in the graphics processor sharing pool according to a preset assignment condition, determine a target graphics processor from the at least one to-be-assigned graphics processor, so that a remaining graphics computing power value of the target graphics processor is greater than or equal to a preset graphics computing power requirement value. The preset assignment condition can be to preferentially assign an idle graphics processor as a to-be-assigned graphics processor, or to preferentially assign a graphics processor that has been assigned a cloud application as a to-be-assigned graphics processor. The graphics processor that has been assigned a cloud application is used to respond to a graphics computing power request of the assigned cloud application.

[0060] In an embodiment, the GPU cloud desktop platform can traverse graphics processors in the preconfigured graphics processor sharing pool, compare a remaining graphics computing power value of the traversed graphics processor with a preset graphics computing power requirement value, and in a case where the remaining graphics computing power value of the traversed graphics processor is greater than or equal to the preset graphics computing power requirement value, determine that the traversed graphics processor is a target graphics processor, and stop traversing.

[0061] Step 206: starting a target cloud application on the GPU cloud desktop platform, connecting the target cloud application directly to the terminal, and responding to a graphics computing power request of the target cloud application through the target graphics processor.

[0062] The graphics computing power request is an action of requesting resources from a graphics processor to process a cloud application graphics task on the graphics processor. Direct connection means that the target cloud application is connected to the terminal by skipping the CPU cloud desktop, and the target cloud application running on the server where the GPU cloud desktop platform is located can be directly operated through the terminal.

[0063] In an embodiment, the GPU cloud desktop platform can start the target cloud application, return a target cloud application running interface to the terminal after starting the target cloud application, receive a graphics computing power request triggered by the terminal through the target cloud application running interface, and respond to the graphics computing power request of the target cloud application through the target graphics processor.

[0064] In an embodiment, the target graphics processor can only respond to a graphics computing power request of the target cloud application triggered by a target object that has passed authentication. In this embodiment, the GPU cloud desktop platform can start the target cloud application through the identity of the target object, and respond to a graphics computing power request of the target cloud application triggered by the target object through the target graphics processor.

[0065] In the cloud application running method based on the GPU cloud desktop platform, the CPU cloud desktop triggers a starting event of the target cloud application, the GPU cloud desktop platform starts the target cloud application, and the target graphics processor responds to a graphics computing power request of the target cloud application, so that the target cloud application requiring the graphics processor can be run when the CPU cloud desktop is used, and conditions for better graphics presentation and rendering performance when the target cloud application is run are created. Moreover, the target graphics processor does not need to be bound to the CPU cloud desktop, but is determined from a graphics processor shared pool shared by multiple CPU cloud desktops according to a preset graphics computing power requirement value of the target cloud application. The remaining graphics computing power value of the target graphics processor can match the preset graphics computing power requirement value of the target cloud application, and the target cloud application can be supported to be run, so that flexible scheduling of multiple graphics processors is implemented, and resource utilization is high.

[0066] In one embodiment, the starting event of the target cloud application is triggered in response to a triggering operation on identification information of the target cloud application in an application panel. The application panel is displayed on the terminal by the CPU cloud desktop, and the application panel displays identification information of at least one cloud application requiring a graphics processor.

[0067] The application panel is software for displaying identification information of other applications. The identification information of the target cloud application is a visual element representing the target cloud application, which can be a shortcut icon or text representing the target cloud application.

[0068] In one embodiment, the CPU cloud desktop can display the application panel on the terminal and display identification information of at least one cloud application requiring a graphics processor in the application panel in response to a starting event of the application panel installed on the CPU cloud desktop. The starting event of the application panel can be a manual triggering operation or an automatic triggering event. The manual triggering operation is, for example, clicking on the application panel identification information, which can be an icon or text representing the application panel. The automatic triggering event can be automatically triggered when the CPU cloud desktop is started, or can be automatically triggered after the application panel is abnormally closed and restored to normal.

[0069] In one embodiment, the CPU cloud desktop can respond to a login event triggered on the application panel, obtain object identity information triggering the login event, determine a cloud application requiring a graphics processor corresponding to the object identity information, obtain identification information of the cloud application, and display the obtained identification information of the cloud application in the application panel.

[0070] In one embodiment, the CPU cloud desktop platform can determine at least one cloud application requiring a graphics processor from the GPU cloud desktop platform after displaying the application panel, generate identification information of the at least one cloud application, and display the generated identification information in the application panel.

[0071] In one embodiment, the GPU cloud desktop platform can determine at least one cloud application requiring a graphics processor after the CPU cloud desktop displays the application panel, generate identification information of the at least one cloud application, and send the generated identification information to the CPU cloud desktop through the network to trigger the display of the identification information of the at least one cloud application in the application panel.

[0072] In this embodiment, the cloud applications capable of running through the GPU cloud desktop platform are presented through the application panel on the CPU cloud desktop, which can efficiently and conveniently use the service of running cloud applications provided by the GPU cloud desktop platform and reduce the resource consumption of the CPU cloud desktop.

[0073] In one embodiment, as shown in Figure 3 According to the preset graphics computing power requirement value, the target graphics processor is determined from the preconfigured graphics processor sharing pool, so that the remaining graphics computing power value of the target graphics processor is greater than or equal to the preset graphics computing power requirement value, including the following steps 302 to 306:

[0074] Step 302, obtain the remaining graphics computing power value of the plurality of graphics processors in the preconfigured graphics processor sharing pool.

[0075] In one embodiment, the GPU cloud desktop platform can determine all graphics processors in the preconfigured graphics processor sharing pool and obtain the remaining graphics computing power value of all graphics processors.

[0076] In one embodiment, the GPU cloud desktop platform can filter out the graphics processors in the preconfigured graphics processor sharing pool in a preset state, obtain the remaining graphics computing power value of the graphics processors in the preset state, and continue to filter the graphics processors based on the obtained remaining graphics computing power value in the subsequent step 304. The graphics processors in the preset state can be idle graphics processors or graphics processors that have been allocated with cloud applications.

[0077] In one embodiment, when the sum of the remaining graphics computing power values of all graphics processors in the preconfigured graphics processor sharing pool is maintained at the preset minimum computing power threshold for a preset length of time, the sharing pool of output graphics processors needs to increase the computing power prompt information.

[0078] Step 304, filter out the graphics processors with the remaining graphics computing power value greater than or equal to the preset graphics computing power requirement value.

[0079] In one embodiment, the GPU cloud desktop platform can filter out all graphics processors with the remaining graphics computing power value greater than or equal to the preset graphics computing power requirement value.

[0080] In an embodiment, the GPU cloud desktop platform can screen a preset number of graphics processors whose remaining graphics processing power values are greater than or equal to the preset graphics processing power requirement value. In this embodiment, the GPU cloud desktop platform can traverse the graphics processors whose remaining graphics processing power values are obtained in step 302, and when a graphics processor whose remaining graphics processing power value is greater than or equal to the preset graphics processing power requirement value is traversed, mark the traversed graphics processor as a candidate graphics processor. When the preset number of candidate graphics processors is reached, stop the traversal, and subsequently select the target graphics processor from the preset number of candidate graphics processors.

[0081] Step 306: Select the target graphics processor from the screened graphics processors according to a preset selection condition.

[0082] In an embodiment, the GPU cloud desktop platform can determine an idle graphics processor in a preconfigured graphics processor sharing pool that does not respond to a graphics processing power request, obtain a remaining graphics processing power value of the idle graphics processor, screen a graphics processor whose remaining graphics processing power value is greater than or equal to the preset graphics processing power requirement value from the idle graphics processors, and randomly select any screened idle graphics processor as the target graphics processor.

[0083] The preset selection condition is a condition that needs to be reached for automatically selecting the target graphics processor. The preset selection condition can be random selection, selection according to the maximum remaining graphics processing power value, or selection according to the minimum remaining graphics processing power value.

[0084] In an embodiment, the random selection can be random selection from idle graphics processors first, and then random selection from graphics processors that have been allocated with cloud applications. This embodiment can quickly select a suitable target graphics processor, and can ensure that the target graphics processor has sufficient processing power to cope with a sudden increase in the processing power requirement of the target cloud application.

[0085] In an embodiment, the random selection can also be random selection from graphics processors that have been allocated with cloud applications first, and then random selection from idle graphics processors. This embodiment can quickly select a suitable target graphics processor, and can improve the utilization rate of the graphics processors in the graphics processor sharing pool to a certain extent to cope with the situation where a large number of cloud applications need to use the GPU cloud desktop platform.

[0086] In an embodiment, the selection according to the maximum remaining graphics processing power value can ensure that the target graphics processor has sufficient processing power to cope with a sudden increase in the processing power requirement of the target cloud application to a certain extent.

[0087] In one embodiment, the minimum residual graphics processing power value selection is performed so that the residual graphics processing power value is closest to the preset graphics processing power requirement value, and the utilization of the graphics processors in the graphics processor shared pool is improved to cope with the situation that a large number of cloud applications need to use the GPU cloud desktop platform.

[0088] In one embodiment, the GPU cloud desktop platform can adopt an adaptive preset selection condition at different times. Specifically, the utilization of the platform server supporting the GPU cloud desktop platform as a whole can be used as an index to determine the applicable preset selection condition. For example, when the index does not reach a threshold value, random selection or maximum residual graphics processing power value selection can be adopted, and when the index reaches the threshold value, minimum residual graphics processing power value selection can be adopted.

[0089] In this embodiment, by comparing the residual graphics processing power of the graphics processors in the graphics processor shared pool with the preset graphics processing power requirement value, the graphics processors with sufficient graphics processing power to respond to the graphics processing power request of the target cloud application are screened out, and then the target graphics processor is uniquely determined according to the preset selection condition, so that the target graphics processor is sufficient to respond to the graphics processing power request of the target cloud application, thereby creating conditions for the normal operation of the target cloud application.

[0090] In one embodiment, when the full-load state of the target graphics processor is maintained for a preset time length, the target graphics processor is re-determined from the graphics processor shared pool, so that the residual graphics processing power value of the re-determined target graphics processor is greater than the residual graphics processing power value of the previously determined target graphics processor; and the graphics processing power request of the target cloud application is responded to by the re-determined target graphics processor, and the response of the previously determined target graphics processor to the graphics processing power request of the target cloud application is stopped.

[0091] In this embodiment, during the operation of the target cloud application, when the full-load state of the target graphics processor is maintained for a preset time length, it indicates that the current target graphics processor is insufficient to cope with the graphics processing power request of the target cloud application, and the graphics processing power request is migrated to a target graphics processor with a larger residual graphics processing power value to provide better operation conditions for the target cloud application.

[0092] In one embodiment, step 306 includes: sorting the screened graphics processors according to a preset sorting condition based on the residual graphics processing power values of the screened graphics processors; and determining the graphics processor ranked first in the sorting result as the target graphics processor.

[0093] The preset sorting condition is a rule for sequentially arranging the screened graphics processors. The preset sorting condition can be random sorting of the screened graphics processors, sorting according to the residual graphics processing power values from large to small, or sorting according to the residual graphics processing power values from small to large.

[0094] In an embodiment, the GPU cloud desktop platform can determine a preset priority of the target cloud application, determine a preset sorting condition corresponding to the preset priority of the target cloud application, sort the filtered graphics processors according to the preset sorting condition, and determine the graphics processor ranked first in the sorting result as the target graphics processor. For example, the preset priority of the cloud application can be high priority, medium priority or low priority. In the case of a high priority target cloud application, the preset sorting condition can be sorting according to the remaining graphics computing power value from large to small, and the target graphics processor with the largest remaining graphics computing power value is selected as the target graphics processor.

[0095] In this embodiment, the graphics processors are sorted according to the preset sorting condition based on the remaining graphics computing power value, which can quickly determine the target graphics processor, so as to quickly start the target cloud application subsequently, so that the target graphics processor responds to the graphics computing power request of the target cloud application.

[0096] In an embodiment, the GPU cloud desktop platform starts the target cloud application, directly connects the target cloud application with the terminal, and responds to the graphics computing power request of the target cloud application through the target graphics processor, including: obtaining the identity information of the target object triggering the starting event; performing authentication based on the identity information to obtain an authentication result; when the authentication result indicates that the authentication is passed, starting the target cloud application corresponding to the identity information of the target object on the GPU cloud desktop platform, directly connecting the started target cloud application with the terminal, and responding to the graphics computing power request of the target cloud application through the target graphics processor.

[0097] The target object is an object with an independent identity in the computer system, which can be embodied in the form of an account, such as an account for logging into the target cloud application. The target object can be a personal account of a personal user or an enterprise management account. The identity information of the target object is information recorded in the computer system, which represents the permission of the target object to use the computer system, such as an account and an account password. The authentication process is a process of verifying whether the target object has the right to use the target graphics processor. The authentication result is the result of the authentication process, which can represent one of the cases of passing the authentication or failing the authentication.

[0098] In an embodiment, the GPU cloud desktop platform can obtain the identity information input by the target object through the login interface of the target cloud application on the CPU cloud desktop, query from the recorded target cloud application identity information table, and output an authentication result indicating that the authentication is successful when the identity information of the target object is queried.

[0099] In one embodiment, the GPU cloud desktop platform can obtain identity information of the target object when the target object logs in an application panel including target cloud application identification information through the CPU cloud desktop, perform authentication processing based on the identity information, and obtain an authentication result.

[0100] In one embodiment, the GPU cloud desktop platform can read an executable file of the target cloud application and an archive corresponding to the identity information of the target object and the target cloud application from the cloud storage device, start the target cloud application based on the executable file and the archive, and send an interface of the started target cloud application to the terminal.

[0101] In one embodiment, the GPU cloud desktop platform can respond to a graphics computing power request sent by the target object by operating the target cloud application on the terminal, instruct the target graphics processor to respond to the graphics computing power request, and send a result processed by the target graphics processor in response to the graphics computing power request to the terminal.

[0102] In one embodiment, the GPU cloud desktop platform can respond to a close event of the target cloud application triggered by the target object on the terminal, disconnect the target cloud application from the terminal, and stop responding to the graphics computing power request corresponding to the target object and the target cloud application by the target graphics processor.

[0103] In this embodiment, the GPU cloud desktop platform needs to authenticate the target object before starting the target cloud application, so as to avoid attacks on the GPU cloud desktop platform by unknown identities and cause insecurity; after the target object is successfully authenticated, the corresponding target cloud application is started with the identity information of the target object, so that the use object of the resource is clear, resource management is facilitated, and resource waste is avoided.

[0104] In one embodiment, the cloud application running method based on the GPU cloud desktop platform further includes a data file access step, which specifically includes: after the CPU cloud desktop obtains a data file transmitted by a hardware interface of the terminal and stores the data file in the cloud storage device corresponding to the identity information, the GPU cloud desktop platform accesses the data file stored in the cloud storage device corresponding to the identity information after receiving an access request for the data file triggered by the terminal with the identity information.

[0105] The hardware interface is a physical interface on the terminal for connecting external devices to transmit data between the terminal and the connected external devices. The hardware interface can be a USB (Universal Serial Bus) interface or a serial port. The cloud storage device is a remote storage device relative to the local terminal, which can be a remote server or a virtual server.

[0106] A data file is an electronic record of information that can be recognized by a computer. The data file can be an electronic document, can be a multimedia file such as a picture, a video, etc., can be other electronic records stored in a cloud storage device. An access request is an action of proposing an operation on a data file; the operation on the data file can be a read operation, a write operation, a copy operation, a download operation, an upload operation, or other.

[0107] In one embodiment, the GPU cloud desktop platform can trigger an access request for a data file by identity information in response to a terminal, verify whether the identity information has the corresponding permission of the access request, and access the data file on the cloud storage device with the identity information when it is verified that the identity information has the corresponding permission of the access request.

[0108] In this embodiment, when the GPU cloud desktop platform accesses the data file on the cloud storage device, the corresponding identity information is required for access, which can ensure that the data file on the cloud storage device cannot be randomly read or tampered with, and the privacy and security of the data file are ensured.

[0109] In one embodiment, as shown in Figure 3 a specific application scenario of a GPU cloud desktop platform-based cloud application running method is provided, and the terminal can be a thin client. The application scenario specifically includes the following steps:

[0110] The CPU cloud desktop can respond to a login operation on the CPU cloud desktop triggered by the thin client through a remote network, and log in to the CPU cloud desktop. In response to a start event of an application panel on the CPU cloud desktop, the application panel is displayed, and identification information of at least one cloud application requiring a graphics processing unit is displayed in the application panel.

[0111] The CPU cloud desktop can respond to a start trigger operation of the thin client on the identification information of the target cloud application in the identification information of the at least one cloud application, display a login interface of the target cloud application, receive identity information input by a target object in the login interface, and trigger a start event of the target cloud application.

[0112] The GPU cloud desktop platform responds to the start event of the target cloud application, obtains a remaining graphics computing power value of all graphics processing units in a preconfigured graphics processing unit sharing pool, filters out graphics processing units with a remaining graphics computing power value greater than or equal to the preset graphics computing power requirement value, and selects any idle graphics processing unit in the filtered graphics processing units as a target graphics processing unit.

[0113] After identifying the target graphics processor, the GPU cloud desktop platform can obtain the identity information of the target object that triggered the launch event. Based on this identity information, it performs authentication processing to obtain the authentication result. When the authentication result indicates that the authentication is successful, it loads thin client data, including data files, corresponding to the identity information and the target cloud application, from the cloud storage device. Based on the thin client data, it launches the target cloud application corresponding to the identity information of the target object on the GPU cloud desktop platform, directly connects the launched target cloud application to the thin client, and responds to the graphics computing power request of the target cloud application through the target graphics processor.

[0114] When the CPU cloud desktop receives the data file from the terminal's hardware interface and uploads the data file to the cloud storage device for storage based on the corresponding identity information, the GPU cloud desktop platform, upon receiving an access request for the data file triggered by the terminal's identity information, accesses the data file stored on the cloud storage device based on the corresponding identity information.

[0115] It should be understood that although the steps in the flowcharts of the embodiments described above are shown sequentially according to the arrows, these steps are not necessarily executed in the order indicated by the arrows. Unless explicitly stated herein, there is no strict order restriction on the execution of these steps, and they can be executed in other orders. Moreover, at least some steps in the flowcharts of the embodiments described above may include multiple steps or multiple stages. These steps or stages are not necessarily completed at the same time, but can be executed at different times. The execution order of these steps or stages is not necessarily sequential, but can be performed alternately or in turn with other steps or at least some of the steps or stages of other steps.

[0116] Based on the same inventive concept, this application also provides a cloud application running device based on a GPU cloud desktop platform for implementing the aforementioned cloud application running method based on a GPU cloud desktop platform. The solution provided by this device is similar to the solution described in the above method. Therefore, the specific limitations in one or more embodiments of the cloud application running device based on a GPU cloud desktop platform provided below can be found in the limitations of the cloud application running method based on a GPU cloud desktop platform described above, and will not be repeated here.

[0117] In one embodiment, such as Figure 5 As shown, a cloud application running device 500 based on a GPU cloud desktop platform is provided, including: a graphics processor allocation module 510 and a graphics computing power request response module 520, wherein:

[0118] The graphics processor allocation module 510 is configured to, in response to a starting event of a target cloud application requiring use of a graphics processor on a CPU cloud desktop connected by the terminal, obtain a preset graphics computing power requirement value of the target cloud application; and determine a target graphics processor from a preconfigured graphics processor sharing pool according to the preset graphics computing power requirement value, so that a remaining graphics computing power value of the target graphics processor is greater than or equal to the preset graphics computing power requirement value.

[0119] The graphics computing power request response module 520 is configured to start the target cloud application on the GPU cloud desktop platform, directly connect the target cloud application with the terminal, and respond to a graphics computing power request of the target cloud application by using the target graphics processor.

[0120] In an embodiment, the starting event of the target cloud application is triggered in response to a triggering operation on identification information of the target cloud application in an application panel, and the application panel is provided by the CPU cloud desktop to the terminal for display, and the application panel displays identification information of at least one cloud application requiring use of a graphics processor.

[0121] In an embodiment, the graphics processor allocation module 510 is further configured to obtain remaining graphics computing power values of a plurality of graphics processors in the preconfigured graphics processor sharing pool, filter out graphics processors with a remaining graphics computing power value greater than or equal to the preset graphics computing power requirement value, and select the target graphics processor from the filtered graphics processors according to a preset selection condition.

[0122] In an embodiment, the graphics processor allocation module 510 is further configured to sort the filtered graphics processors according to a preset sorting condition based on the remaining graphics computing power values of the filtered graphics processors, and determine the graphics processor with a first position in the sorting result as the target graphics processor.

[0123] In an embodiment, the graphics computing power request response module 520 is further configured to obtain identity information of a target object triggering the starting event, perform authentication based on the identity information to obtain an authentication result, and when the authentication result indicates that the authentication is passed, start a target cloud application corresponding to the identity information of the target object on the GPU cloud desktop platform, directly connect the started target cloud application with the terminal, and respond to a graphics computing power request of the target cloud application by using the target graphics processor.

[0124] In an embodiment, the cloud application running device 500 based on the GPU cloud desktop platform further includes a data file access module, which is configured to, when the CPU cloud desktop obtains a data file transmitted by a hardware interface of the terminal and stores the data file in a cloud storage device corresponding to identity information, access the data file stored in the cloud storage device corresponding to the identity information after the GPU cloud desktop platform receives an access request for the data file triggered by the terminal with the identity information.

[0125] The cloud application running device based on the GPU cloud desktop platform triggers a starting event of the target cloud application through the CPU cloud desktop, the GPU cloud desktop platform starts the target cloud application, and the target graphics processor responds to the graphics computing power request of the target cloud application, so that the target cloud application requiring the graphics processor can be run when the CPU cloud desktop is used, and conditions for obtaining better graphics presentation and rendering performance when running the target cloud application are created. Moreover, the target graphics processor does not need to be bound to the CPU cloud desktop, but is determined from a graphics processor shared pool shared by multiple CPU cloud desktops according to a preset graphics computing power requirement value of the target cloud application. The remaining graphics computing power value of the target graphics processor can match the preset graphics computing power requirement value of the target cloud application, and can support running the target cloud application, thereby realizing flexible scheduling of multiple graphics processors and high resource utilization.

[0126] The various modules in the cloud application running device based on the GPU cloud desktop platform can be realized by software, hardware and combinations thereof in whole or in part. The various modules can be embedded in or independent of the processor in the computer device in hardware form, or can be stored in the memory in the computer device in software form, so as to call and execute the operations corresponding to the various modules by the processor.

[0127] In one embodiment, a computer device, which can be a server, has an internal structure diagram as shown in Figure 6 The computer device includes a processor, a memory, an input / output interface (I / O) and a communication interface. The processor, the memory and the input / output interface are connected through a system bus, and the communication interface is connected to the system bus through the input / output interface. The processor of the computer device is used to provide computing and control capabilities. The memory of the computer device includes a non-volatile storage medium and an internal memory. The non-volatile storage medium stores an operating system, a computer program and a database. The internal memory provides an environment for the operating system and the computer program in the non-volatile storage medium to run. The database of the computer device is used to store data required to be stored when executing the cloud application running method based on the GPU cloud desktop platform. The input / output interface of the computer device is used to exchange information between the processor and external devices. The communication interface of the computer device is used to communicate with external terminals through network connection. The computer program is executed by the processor to implement a cloud application running method based on a GPU cloud desktop platform.

[0128] Those skilled in the art can understand that Figure 6The structure shown in the figure is only a block diagram of part of the structure related to the scheme of the present application, and does not constitute a limitation on the computer device to which the scheme of the present application is applied. The specific computer device can include more or fewer components than those shown in the figure, or combine certain components, or have a different arrangement of components.

[0129] In one embodiment, a computer device is provided, comprising a memory and a processor, the memory storing a computer program, and the processor implementing the following steps when executing the computer program:

[0130] In response to a start event of a target cloud application requiring a graphics processor on a CPU cloud desktop to which the terminal is connected, a preset graphics computing power requirement value of the target cloud application is obtained;

[0131] According to the preset graphics computing power requirement value, a target graphics processor is determined from a preconfigured graphics processor sharing pool, so that the remaining graphics computing power value of the target graphics processor is greater than or equal to the preset graphics computing power requirement value;

[0132] Starting the target cloud application on the GPU cloud desktop platform, connecting the target cloud application directly to the terminal, and responding to the graphics computing power request of the target cloud application through the target graphics processor.

[0133] In one embodiment, the processor further implements the following when executing the computer program: the start event of the target cloud application is triggered in response to a trigger operation on identification information identifying the target cloud application in an application panel, the application panel being provided by the CPU cloud desktop to the terminal for display, and the application panel displaying identification information of at least one cloud application requiring a graphics processor.

[0134] In one embodiment, the processor further implements the following steps when executing the computer program: obtaining the remaining graphics computing power values of a plurality of graphics processors in the preconfigured graphics processor sharing pool; screening out graphics processors with a remaining graphics computing power value greater than or equal to the preset graphics computing power requirement value; and selecting the target graphics processor from the screened graphics processors according to a preset selection condition.

[0135] In one embodiment, the processor further implements the following steps when executing the computer program: sorting the screened graphics processors according to a preset sorting condition based on the remaining graphics computing power values of the screened graphics processors; and determining the graphics processor ranked first in the sorting result as the target graphics processor.

[0136] In one embodiment, the processor further implements the following steps when executing the computer program: obtaining identity information of a target object triggering the starting event; performing authentication based on the identity information to obtain an authentication result; when the authentication result indicates that the authentication is passed, starting a target cloud application corresponding to the identity information of the target object on the GPU cloud desktop platform, directly connecting the started target cloud application with the terminal, and responding to a graphics computing power request of the target cloud application through the target graphics processor.

[0137] In one embodiment, the processor further implements the following steps when executing the computer program: when the CPU cloud desktop obtains a data file transmitted by the hardware interface of the terminal and stores the data file in the cloud storage device corresponding to the identity information, the GPU cloud desktop platform, after receiving an access request for the data file triggered by the terminal with the identity information, accesses the data file stored in the cloud storage device corresponding to the identity information.

[0138] In one embodiment, a computer readable storage medium is provided, and the computer readable storage medium stores a computer program. The computer program is executed by a processor to implement the following steps:

[0139] In response to a starting event of a target cloud application requiring a graphics processor on a CPU cloud desktop connected by a terminal, obtaining a preset graphics computing power requirement value of the target cloud application;

[0140] According to the preset graphics computing power requirement value, determining a target graphics processor from a preconfigured graphics processor sharing pool, so that the remaining graphics computing power value of the target graphics processor is greater than or equal to the preset graphics computing power requirement value;

[0141] Starting the target cloud application on the GPU cloud desktop platform, directly connecting the target cloud application with the terminal, and responding to a graphics computing power request of the target cloud application through the target graphics processor.

[0142] In one embodiment, the computer program is further executed by the processor to implement: the starting event of the target cloud application is triggered in response to a triggering operation on identification information of the target cloud application in an application panel, the application panel is provided by the CPU cloud desktop to the terminal for display, and the application panel displays identification information of at least one cloud application requiring a graphics processor.

[0143] In one embodiment, the computer program is further executed by the processor to implement the following steps: obtaining the remaining graphics computing power values of a plurality of graphics processors in the preconfigured graphics processor sharing pool; screening graphics processors with a remaining graphics computing power value greater than or equal to the preset graphics computing power requirement value; and selecting the target graphics processor from the screened graphics processors according to a preset selection condition.

[0144] In one embodiment, the computer program, when executed by the processor, further implements the following steps: sorting the filtered graphics processors according to a preset sorting condition based on the remaining graphics processing power values of the filtered graphics processors; and determining the graphics processor with the first median in the sorting result as the target graphics processor.

[0145] In one embodiment, the computer program, when executed by the processor, further implements the following steps: obtaining identity information of a target object triggering the starting event; performing authentication processing based on the identity information to obtain an authentication result; when the authentication result indicates that the authentication is passed, starting a target cloud application corresponding to the identity information of the target object on the GPU cloud desktop platform, directly connecting the started target cloud application with the terminal, and responding to a graphics processing power request of the target cloud application through the target graphics processor.

[0146] In one embodiment, the computer program, when executed by the processor, further implements the following steps: when the CPU cloud desktop obtains a data file transmitted by the hardware interface of the terminal and stores the data file in the cloud storage device corresponding to the identity information, the GPU cloud desktop platform receives an access request for the data file triggered by the terminal with the identity information, and accesses the data file stored in the cloud storage device corresponding to the identity information.

[0147] In one embodiment, a computer program product is provided, comprising a computer program which, when executed by a processor, implements the following steps:

[0148] In response to a starting event of a target cloud application requiring a graphics processor on a CPU cloud desktop connected by a terminal, obtaining a preset graphics processing power requirement value of the target cloud application;

[0149] According to the preset graphics processing power requirement value, determining a target graphics processor from a preconfigured graphics processor sharing pool, so that the remaining graphics processing power value of the target graphics processor is greater than or equal to the preset graphics processing power requirement value;

[0150] Starting the target cloud application on the GPU cloud desktop platform, directly connecting the target cloud application with the terminal, and responding to a graphics processing power request of the target cloud application through the target graphics processor.

[0151] In one embodiment, the computer program, when executed by the processor, further implements: the starting event of the target cloud application is triggered in response to a triggering operation on identification information of the target cloud application in an application panel, the application panel is provided by the CPU cloud desktop to the terminal for display, and the application panel displays identification information of at least one cloud application requiring a graphics processor.

[0152] In an embodiment, the computer program, when executed by the processor, further implements the following steps: obtaining a remaining graphics computing power value of each graphics processor in the pre-configured graphics processor shared pool; screening out graphics processors whose remaining graphics computing power values are greater than or equal to the preset graphics computing power requirement value; and selecting a target graphics processor from the screened graphics processors according to a preset selection condition.

[0153] In an embodiment, the computer program, when executed by the processor, further implements the following steps: sorting the screened graphics processors according to a preset sorting condition based on the remaining graphics computing power values of the screened graphics processors; and determining the graphics processor ranked first in the sorting result as the target graphics processor.

[0154] In an embodiment, the computer program, when executed by the processor, further implements the following steps: obtaining identity information of a target object triggering the start event; performing authentication based on the identity information to obtain an authentication result; when the authentication result indicates that the authentication is passed, starting a target cloud application corresponding to the identity information of the target object on the GPU cloud desktop platform, directly connecting the started target cloud application with the terminal, and responding to a graphics computing power request of the target cloud application through the target graphics processor.

[0155] In an embodiment, the computer program, when executed by the processor, further implements the following steps: when the CPU cloud desktop obtains a data file transmitted by the hardware interface of the terminal and stores the data file in the cloud storage device corresponding to the identity information, the GPU cloud desktop platform, after receiving an access request for the data file triggered by the terminal with the identity information, accesses the data file stored in the cloud storage device corresponding to the identity information.

[0156] It should be noted that the user information (including but not limited to user equipment information, user personal information, etc.) and data (including but not limited to data for analysis, stored data, displayed data, etc.) involved in the present application are all information and data authorized by the user or authorized by all parties, and the collection, use and processing of the related data need to comply with the relevant laws, regulations and standards of the relevant countries and regions.

[0157] Those skilled in the art can understand that all or part of the processes in the above-mentioned embodiment methods can be completed by instructing the relevant hardware through a computer program. The computer program can be stored in a non-volatile computer readable storage medium, and when the computer program is executed, the processes of the above-mentioned embodiments of the methods can be included. Any reference to memory, database or other medium used in the embodiments provided in the present application can include at least one of non-volatile and volatile memory. Non-volatile memory can include read-only memory (Read-Only Memory, ROM), magnetic tape, floppy disk, flash memory, optical storage, high-density embedded non-volatile memory, resistive memory (ReRAM), magnetoresistive random access memory (Magnetoresistive Random Access Memory, MRAM), ferroelectric memory (Ferroelectric Random Access Memory, FRAM), phase change memory (Phase Change Memory, PCM), graphene memory, etc. Volatile memory can include random access memory (Random Access Memory, RAM) or external cache memory, etc. As an illustration but not limitation, RAM can be in various forms, such as static random access memory (Static Random Access Memory, SRAM) or dynamic random access memory (Dynamic Random Access Memory, DRAM), etc. The database involved in the embodiments provided in the present application can include at least one of a relational database and a non-relational database. The non-relational database can include a distributed database based on a block chain, etc., without being limited thereto. The processor involved in the embodiments provided in the present application can be a general-purpose processor, a central processing unit, a graphics processing unit, a digital signal processor, a programmable logic device, a data processing logic device based on quantum computing, etc., without being limited thereto.

[0158] Any combination of the technical features of the above embodiments can be made. In order to make the description simple, all possible combinations of the technical features in the above embodiments are not described, however, as long as the combination of the technical features does not exist contradictory, it should be considered as the scope of the present application.

[0159] The above embodiments only express several implementation manners of the present application, and the description is more specific and detailed, but it should not be understood as a limitation on the scope of the patent of the present application. It should be pointed out that for ordinary skilled in the art, without departing from the concept of the present application, a number of modifications and improvements can be made, which are within the scope of protection of the present application. Therefore, the protection scope of the present application should be subject to the appended claims.

Claims

1. A cloud application running method based on a GPU cloud desktop platform, characterized in that, The method comprises: in response to a starting event of a target cloud application requiring a graphics processor on a CPU cloud desktop connected by a terminal, obtaining a preset graphics computing power requirement value of the target cloud application; the preset graphics computing power requirement value is determined based on a resource usage rate of the target cloud application expected to occupy a graphics processor; determining a target graphics processor from a preconfigured graphics processor sharing pool according to the preset graphics computing power requirement value, so that a remaining graphics computing power value of the target graphics processor is greater than or equal to the preset graphics computing power requirement value; the graphics processor sharing pool is a collection of graphics processors shared by a plurality of CPU cloud desktops; when the sum of the remaining graphics computing power values of all graphics processors in the graphics processor sharing pool is maintained at a preset minimum computing power threshold for a preset length of time, outputting prompt information that the graphics processor sharing pool needs to increase computing power; starting the target cloud application on the GPU cloud desktop platform, directly connecting the target cloud application with the terminal, and responding to a graphics computing power request of the target cloud application through the target graphics processor; wherein directly connecting the target cloud application with the terminal means skipping the CPU cloud desktop, connecting the target cloud application with the terminal, and directly operating the target cloud application running on the server where the GPU cloud desktop platform is located through the terminal.

2. The method of claim 1, wherein, The starting event of the target cloud application is triggered in response to a triggering operation on identification information of the target cloud application in an application panel, and the application panel is displayed on the terminal by the CPU cloud desktop, and the application panel displays identification information of at least one cloud application requiring a graphics processor.

3. The method of claim 1, wherein, The method comprises: obtaining the remaining graphics computing power values of a plurality of graphics processors in the preconfigured graphics processor sharing pool; screening graphics processors with a remaining graphics computing power value greater than or equal to the preset graphics computing power requirement value; selecting a target graphics processor from the screened graphics processors according to a preset selection condition.

4. The method of claim 3, wherein, The method comprises: sorting the screened graphics processors according to a preset sorting condition based on the remaining graphics computing power values of the screened graphics processors; determining the graphics processor ranked first in the sorting result as the target graphics processor.

5. The method of claim 1, wherein, The method comprises: obtaining identity information of a target object triggering the starting event; performing authentication processing based on the identity information to obtain an authentication result; When the authentication result represents that the authentication is passed, the target cloud application corresponding to the identity information of the target object is started on the GPU cloud desktop platform, the started target cloud application is directly connected with the terminal, and the target cloud application is responded to a graphics algorithm request of the target cloud application through the target graphics processor.

6. The method of claim 5, wherein, The method further comprises: When the CPU cloud desktop obtains a data file transmitted by a hardware interface of the terminal, and stores the data file in a cloud storage device corresponding to the identity information, the GPU cloud desktop platform, after receiving an access request for the data file triggered by the terminal with the identity information, accesses the data file stored in the cloud storage device corresponding to the identity information. 7.A cloud application running apparatus based on a GPU cloud desktop platform, characterized by comprising: The apparatus comprises: A graphics processor allocation module is configured to, in response to a start event of a target cloud application requiring a graphics processor on a CPU cloud desktop connected with a terminal, obtain a preset graphics algorithm requirement value of the target cloud application; the preset graphics algorithm requirement value is determined based on a resource usage rate of the graphics processor predicted to be occupied by the target cloud application; according to the preset graphics algorithm requirement value, a target graphics processor is determined from a preconfigured graphics processor shared pool, so that a residual graphics algorithm value of the target graphics processor is greater than or equal to the preset graphics algorithm requirement value; the graphics processor shared pool is a collection of graphics processors shared by a plurality of CPU cloud desktops; when a sum of residual graphics algorithm values of all graphics processors in the graphics processor shared pool is maintained at a preset minimum algorithm threshold for a preset time length, a prompt information that the graphics processor shared pool needs to increase the algorithm is outputted; A graphics algorithm request response module is configured to start the target cloud application on the GPU cloud desktop platform, directly connect the target cloud application with the terminal, and respond to a graphics algorithm request of the target cloud application through the target graphics processor; wherein, directly connecting the target cloud application with the terminal refers to connecting the target cloud application with the terminal by skipping the CPU cloud desktop, and directly operating the target cloud application running on a server where the GPU cloud desktop platform is located through the terminal. 8.A computer device, comprising a memory and a processor, wherein the memory stores a computer program, and the computer device is configured to perform the method according to any one of claims 1-7. The processor executes the computer program to implement the steps of the method of any one of claims 1 to 6.

9. A computer-readable storage medium having stored thereon a computer program, characterized in that, The computer program is executed by the processor to implement the steps of the method of any one of claims 1 to 6.

10. A computer program product comprising a computer program, characterized in that, The computer program is executed by the processor to implement the steps of the method of any one of claims 1 to 6. The computer program is executed by the processor to implement the steps of the method of any one of claims 1 to 6.

Images