A computing power resource access method of an intelligent computing center

Abstract

The application discloses a kind of intelligent computing center's computing power resource access method, it is related to computing power resource access technical field, including the identity of user, the computing power demand of user this time computing power resource access task;Real-time redundant computing power is calculated, and the maximum computing power access amount corresponding to the unit of the computing power resource access authority level of user this time computing power resource access task is obtained, and auxiliary access analysis early warning is sent out outward;After receiving auxiliary access analysis early warning, real-time computing power anomaly coefficient is calculated, and computing power correction instruction is sent outward;After receiving computing power correction instruction, real-time redundant computing power and real-time computing power anomaly coefficient are obtained, correction redundant computing power is calculated, in combination with the computing power demand of user this time computing power resource access task, the permitted computing power of user this time computing power resource access is output and permitted computing power access duration is accessed and executed, in computing power demand peak period, system dynamically allocates correction redundant computing power, ensure that training task is completed on time.

Description

Technical Field

[0001] The present invention relates to the technical field of computing power resource access, and in particular to a computing power resource access method for an intelligent computing center. Background Art

[0002] With the rapid development of technologies such as big data, artificial intelligence, and the Internet of Things, computing demand is exploding. Traditional computing centers, faced with massive computing tasks, are increasingly facing problems such as uneven resource allocation, high access latency, and insufficient security. To address these challenges, intelligent computing centers have emerged. These are comprehensive computing platforms that integrate advanced computing technologies, intelligent management systems, and efficient network infrastructure. By integrating multiple computing models, including cloud computing, edge computing, and distributed computing, they provide users with flexible and scalable computing resources. The core goal of intelligent computing centers is to achieve intelligent scheduling, efficient utilization, and convenient access to computing resources to meet the needs of a wide range of complex computing tasks. In intelligent computing centers, access to computing resources is a critical link between users and computing resources. However, existing methods for accessing computing resources still have many limitations.

[0003] In the Chinese invention application with application publication number CN118842657A, a method and device for accessing computing power resources of an intelligent computing center are disclosed, including step S1: receiving an access instruction from a user, the access instruction is used to indicate the target computing power resources that the user wants to access, the access instruction carries the user's access token, and the access token carries the user's identity information; step S2: performing a hash operation on the access token to obtain a hash value to be verified corresponding to the access token; step S3: judging whether a first condition is met, if so, executing step S4; if not, executing step S5; the first condition includes: the hash value to be verified is consistent with the reference hash value; step S4: verifying whether the user has the authority to access the target computing power resources based on the user's identity information; step S5: denying the user access to the target computing power resources.

[0004] The above invention application combines hashing and access tokens to ensure authentication security and data immutability. By verifying user permissions, only authorized users can access specific computing resources, thereby reducing the risk of unauthorized users accessing computing resources. However, access to computing resources by authorized users is not restricted, and users may use computing resources without restraint, resulting in an overall imbalance in the computing center's load. Some users may occupy a large amount of resources, affecting the computing tasks of other users, causing system performance degradation and even triggering resource contention and conflicts.

[0005] To this end, the present invention provides a method for accessing computing resources of an intelligent computing center. Summary of the Invention

[0006] Technical problems solved

[0007] In view of the deficiencies of the prior art, the application provides a computing power resource access method of an intelligent computing center, and the application calculates the corrected redundant computing power by taking the real-time redundant computing power and the real-time computing power abnormality coefficient Ycx, combines the computing power demand Sx of the user's current computing power resource access task, and outputs the permitted computing power Zx and the permitted computing power access duration Sc of the user's current computing power resource access to perform access, so that the corrected redundant computing power is dynamically allocated by the system during the computing power demand peak period, and the training task is ensured to be completed on time, thereby solving the technical problems recorded in the background art.

[0008] Technical solutions

[0009] To achieve the above object, the application is implemented by the following technical solutions: a computing power resource access method of an intelligent computing center, comprising the following steps:

[0010] The identity of the user is identified, the access permission corresponding to the identity of the user is queried according to the identity of the user, the single task computing amount Dj and the concurrent task number Bf of the user's current computing power resource access task are obtained, and the computing power demand Sx of the user's current computing power resource access task is calculated.

[0011] The real-time redundant computing power Ry is calculated i , the unit maximum computing power access amount Fw corresponding to the access permission level of the user's current computing power resource access task is obtained, and an auxiliary access analysis warning is sent out.

[0012] After receiving the auxiliary access analysis warning, the real-time computing power excess coefficient δ, the memory occupation amount Nz and the bandwidth occupation amount Dz of the real-time unit computing power are calculated, the real-time computing power abnormality coefficient Ycx is calculated, and a computing power correction instruction is sent out.

[0013] After receiving the computing power correction instruction, the real-time redundant computing power Ry i and the real-time computing power abnormality coefficient Ycx are obtained, the corrected redundant computing power Jry i is calculated, the computing power demand Sx of the user's current computing power resource access task is combined, and the permitted computing power Zx and the permitted computing power access duration Sc of the user's current computing power resource access are output to perform access.

[0014] Further, the identity of the user is identified through an identity verification mechanism (such as a username and password, a digital certificate, etc.), the access permission corresponding to the identity of the user is queried according to the identity of the user, if the computing power resource access permission does not pass, the computing power resource access program is exited, and if the computing power resource access permission passes, the user's current computing power resource access task and the computing power resource access permission level are obtained.

[0015] The query of the corresponding access rights can be achieved through access control lists, role allocation tables or attribute rule bases.

[0016] Furthermore, the single-task computing amount Dj of the user's current computing resource access task is obtained based on the PyTorch tool library, and the number of concurrent tasks Bf of the user's current computing resource access task is obtained based on the system monitoring tool.

[0017] thop or ptflops in the PyTorch tool library: automatically counts FLOPs based on the input model and example data. It is applicable to standard modules (such as nn.Conv2d and nn.Linear). torch_flops supports capturing all operators in the model (including non-modular operations such as torch.add), avoiding the problem of traditional tools ignoring some operations.

[0018] The calculation method of single-task computation amount Dj varies depending on the task type:

[0019] Algorithmic tasks (such as matrix multiplication) Dj = 2 × matrix dimension 3 (For example, the multiplication of two 1000×1000 matrices requires 2×10 9 FLOPs)

[0020] Deep learning tasks: Dj = model forward FLOPs + reverse FLOPs (usually the reverse calculation amount is 2-3 times that of the forward)

[0021] The number of concurrent tasks Bf is calculated differently depending on the task type:

[0022] Real-time tasks: Number of concurrent tasks Bf = number of requests per second (QPS) × single task processing time (seconds)

[0023] Offline tasks: number of concurrent tasks Bf = total number of samples ÷ batch size

[0024] Furthermore, the single-task computing amount Dj and the number of concurrent tasks Bf of the user's current computing resource access task are obtained, and the computing power requirement Sx of the user's current computing resource access task is calculated:

[0025] Sx=Dj*Bf*α

[0026] Here, α is the safety redundancy factor, which is used to reserve resources to cope with sudden loads and is usually 1.2 to 2.0.

[0027] Furthermore, the continuous computing power Cs and real-time computing power Ys of the computing resources of the intelligent computing center are obtained based on the operation log. i , calculate the real-time redundant computing power Ry i :

[0028] Ryi =Cs-Ys i

[0029] Here, i represents the time of the data.

[0030] Sustained computing power refers to the average computing power that a computing center can continuously output when operating at full or near full capacity for extended periods of time (e.g., hours to days). It takes into account the actual operational limitations of the hardware (such as heat dissipation, power supply, and software scheduling), rather than relying solely on theoretical maximums.

[0031] Furthermore, the unit maximum computing power access Fw corresponding to the computing power resource access permission level of the user's current computing power resource access task is obtained. If the unit maximum computing power access Fw is less than the real-time redundant computing power Ry i , then the user's current computing power resource access task is accessed and executed according to the unit maximum computing power access. If the unit maximum computing power access Fw is not less than the real-time redundant computing power Ry i , an auxiliary access analysis warning is issued.

[0032] Furthermore, after receiving the auxiliary access analysis warning, the computing power demand Sx of each computing power resource access task in the access execution state is obtained. a , computing power access per unit time Fw a , computing power access time Fs a And computing power access task completion progress Jd a , calculate the real-time computing power excess coefficient δ:

[0033]

[0034] Among them, a represents the sequence number of each computing resource access task in the access execution state, a=1, 2, ..., m, m is the total number of each computing resource access task in the access execution state, and the computing resource access task status includes access request, access execution and access completion.

[0035] Furthermore, the cumulative computing power access volume Fw of each computing power resource access task in the access execution state is obtained based on the operation log. a , cumulative memory usage Nc a and cumulative bandwidth usage Dk a , calculate the memory usage Nz and bandwidth usage Dz of real-time unit computing power:

[0036]

[0037] Furthermore, the real-time computing power excess coefficient δ, the memory usage Nz and bandwidth usage Dz of the real-time unit computing power are obtained, and the real-time computing power abnormality coefficient Ycx is calculated:

[0038]

[0039] Among them, Δδ represents the excess coefficient of standard computing power, ΔNz represents the memory usage of standard unit computing power, and ΔDz represents the bandwidth usage of standard unit computing power.

[0040] Furthermore, the real-time computing power abnormality coefficient Ycx is obtained. When the real-time computing power abnormality coefficient Ycx exceeds 0.2, a computing power correction instruction is sent out.

[0041] When the real-time computing power abnormality coefficient Ycx does not exceed 0.2, the real-time redundant computing power Ry i Access and execute the user's computing resource access task.

[0042] Furthermore, after receiving the computing power correction instruction, the real-time redundant computing power Ry is obtained. i And the real-time computing power abnormality coefficient Ycx, calculate the corrected redundant computing power Jry i :

[0043]

[0044] Furthermore, the computing power requirement Sx and the corrected redundant computing power Jry of the user's computing power resource access task are obtained. i , calculate the user's current computing power resource access permitted computing power Zx and permitted computing power access duration Sc:

[0045]

[0046] The user's current computing power resource access permitted computing power Zx and permitted computing power access duration Sc are output for access execution.

[0047] (3) Beneficial effects

[0048] The present invention provides a method for accessing computing resources of an intelligent computing center, which has the following beneficial effects:

[0049] 1. Identify the user's identity, query the corresponding access rights based on the user's identity, obtain the single task computing volume Dj and the number of concurrent tasks Bf of the user's current computing resource access task, and calculate the computing power demand Sx of the user's current computing resource access task. This can accurately allocate computing resources, improve efficiency, reduce costs, and provide data support for strategic decision-making.

[0050] 2. Calculate real-time redundant computing power Ry i , and obtain the unit maximum computing power access Fw corresponding to the computing power resource access permission level of the user's current computing power resource access task, and issue auxiliary access analysis warnings to the outside. The system warning prevents resource exhaustion and adjusts task allocation according to redundant computing power to avoid node overload.

[0051] 3. After receiving the auxiliary access analysis warning, calculate the real-time computing power excess coefficient δ, the real-time unit computing power memory usage Nz and bandwidth usage Dz, and calculate the real-time computing power abnormality coefficient Ycx. Send computing power correction instructions to the outside, monitor computing power usage in real time, dynamically schedule tasks, and avoid node overload by adjusting task allocation, thereby improving overall computing efficiency.

[0052] 4. After receiving the computing power correction instruction, obtain the real-time redundant computing power Ry i And the real-time computing power abnormality coefficient Ycx, calculate the corrected redundant computing power Jry i , combined with the computing power demand Sx of the user's computing power resource access task, output the user's computing power resource access allowed computing power Zx and allowed computing power access time Sc for access execution. During the peak period of computing power demand, the system dynamically allocates and corrects redundant computing power to ensure that the training task is completed on time. BRIEF DESCRIPTION OF THE DRAWINGS

[0053] Figure 1 The figure is a flow chart of a method for accessing computing resources of an intelligent computing center according to the present invention. DETAILED DESCRIPTION

[0054] The following will clearly and completely describe the technical solutions in the embodiments of the present invention in conjunction with the accompanying drawings. Obviously, the described embodiments are only part of the embodiments of the present invention, not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without making creative efforts are within the scope of protection of the present invention.

[0055] See also Figure 1 The present invention provides a method for accessing computing resources of an intelligent computing center, comprising the following steps:

[0056] Step 1: Identify the user's identity, query the corresponding access rights based on the user's identity, obtain the single task computing amount Dj and the number of concurrent tasks Bf of the user's current computing resource access task, and calculate the computing power requirement Sx of the user's current computing resource access task.

[0057] The step 1 includes the following:

[0058] Step 101: Identify the user's identity through an authentication mechanism (such as user name and password, digital certificate, etc.), and query the corresponding access rights based on the user's identity. If the computing power resource access rights are not passed, exit the computing power resource access program. If the computing power resource access rights are passed, obtain the user's current computing power resource access task and computing power resource access rights level.

[0059] The query of the corresponding access rights can be achieved through access control lists, role allocation tables or attribute rule bases.

[0060] Step 102: Obtain the single-task computing amount Dj of the user's current computing resource access task based on the PyTorch tool library, and obtain the number of concurrent tasks Bf of the user's current computing resource access task based on the system monitoring tool.

[0061] thop or ptflops in the PyTorch tool library: automatically counts FLOPs based on the input model and example data. It is applicable to standard modules (such as nn.Conv2d and nn.Linear). torch_flops supports capturing all operators in the model (including non-modular operations such as torch.add), avoiding the problem of traditional tools ignoring some operations.

[0062] The calculation method of single-task computation amount Dj varies depending on the task type:

[0063] Algorithmic tasks (such as matrix multiplication) Dj = 2 × matrix dimension 3 (For example, the multiplication of two 1000×1000 matrices requires 2×10 9 FLOPs)

[0064] Deep learning tasks: Dj = model forward FLOPs + reverse FLOPs (usually the reverse calculation amount is 2-3 times that of the forward)

[0065] The number of concurrent tasks Bf is calculated differently depending on the task type:

[0066] Real-time tasks: Number of concurrent tasks Bf = number of requests per second (QPS) × single task processing time (seconds)

[0067] Offline tasks: number of concurrent tasks Bf = total number of samples ÷ batch size

[0068] Step 103: Obtain the single-task computing amount Dj and the number of concurrent tasks Bf of the user's current computing resource access task, and calculate the computing power requirement Sx of the user's current computing resource access task:

[0069] Sx=Dj*Bf*α

[0070] Here, α is the safety redundancy factor, which is used to reserve resources to cope with sudden loads and is usually 1.2 to 2.0.

[0071] When using, combine the contents in steps 101 to 103:

[0072] Identify the user's identity, query the corresponding access rights based on the user's identity, obtain the single task computing volume Dj and the number of concurrent tasks Bf of the user's current computing resource access task, calculate the computing power demand Sx of the user's current computing resource access task, accurately allocate computing power resources, improve efficiency, reduce costs, and provide data support for strategic decision-making.

[0073] Step 2: Calculate the real-time redundant computing power Ry i , and obtain the unit maximum computing power access Fw corresponding to the computing power resource access permission level of the user's current computing power resource access task, and issue an auxiliary access analysis warning.

[0074] The second step includes the following:

[0075] Step 201: Obtain the continuous computing power Cs and real-time computing power Ys of the computing resources of the intelligent computing center based on the operation log. i , calculate the real-time redundant computing power Ry i :

[0076] Ry i =Cs-Ys i

[0077] Here, i represents the time of the data.

[0078] Sustained computing power refers to the average computing power that a computing center can continuously output when operating at full or near full capacity for extended periods of time (e.g., hours to days). It takes into account the actual operational limitations of the hardware (such as heat dissipation, power supply, and software scheduling), rather than relying solely on theoretical maximums.

[0079] Step 202: Obtain the unit maximum computing power access Fw corresponding to the computing power resource access permission level of the user's current computing power resource access task. If the unit maximum computing power access Fw is less than the real-time redundant computing power Ry i , then the user's current computing power resource access task is accessed and executed according to the unit maximum computing power access. If the unit maximum computing power access Fw is not less than the real-time redundant computing power Ry i , an auxiliary access analysis warning is issued.

[0080] When using, combine the contents in steps 201 and 202:

[0081] Calculate real-time redundant computing power Ry i , and obtain the unit maximum computing power access Fw corresponding to the computing power resource access permission level of the user's current computing power resource access task, and issue auxiliary access analysis warnings to the outside. The system warning prevents resource exhaustion and adjusts task allocation according to redundant computing power to avoid node overload.

[0082] Step 3: After receiving the auxiliary access analysis warning, calculate the real-time computing power excess coefficient δ, the real-time unit computing power memory usage Nz and bandwidth usage Dz, and calculate the real-time computing power abnormality coefficient Ycx, and send a computing power correction instruction.

[0083] The step three includes the following:

[0084] Step 301: After receiving the auxiliary access analysis warning, obtain the computing power requirements Sx of each computing power resource access task in the access execution state. a , computing power access per unit time Fw a , computing power access time Fs a And computing power access task completion progress Jd a , calculate the real-time computing power excess coefficient δ:

[0085]

[0086] Among them, a represents the sequence number of each computing resource access task in the access execution state, a=1, 2, ..., m, m is the total number of each computing resource access task in the access execution state, and the computing resource access task status includes access request, access execution and access completion.

[0087] Step 302: Obtain the cumulative computing power access amount Fw of each computing power resource access task in the access execution state according to the operation log a , cumulative memory usage Nc a and cumulative bandwidth usage Dk a , calculate the memory usage Nz and bandwidth usage Dz of real-time unit computing power:

[0088]

[0089] Step 303: Obtain the real-time computing power excess coefficient δ, the memory usage Nz and bandwidth usage Dz of the real-time unit computing power, and calculate the real-time computing power abnormality coefficient Ycx:

[0090]

[0091] Among them, Δδ represents the excess coefficient of standard computing power, ΔNz represents the memory usage of standard unit computing power, and ΔDz represents the bandwidth usage of standard unit computing power.

[0092] Step 304: Obtain the real-time computing power abnormality coefficient Ycx. When the real-time computing power abnormality coefficient Ycx exceeds 0.2, send a computing power correction instruction.

[0093] When the real-time computing power abnormality coefficient Ycx does not exceed 0.2, the real-time redundant computing power Ry iAccess and execute the user's computing resource access task.

[0094] When using, combine the contents in steps 301 to 304:

[0095] After receiving the auxiliary access analysis warning, it calculates the real-time computing power excess coefficient δ, the real-time unit computing power memory usage Nz and bandwidth usage Dz, and the real-time computing power abnormality coefficient Ycx, and sends out computing power correction instructions, monitors computing power usage in real time, and dynamically schedules tasks. By adjusting task allocation, it avoids node overload and improves overall computing efficiency.

[0096] Step 4: After receiving the computing power correction instruction, obtain the real-time redundant computing power Ry i And the real-time computing power abnormality coefficient Ycx, calculate the corrected redundant computing power Jry i , combined with the computing power requirement Sx of the user's computing power resource access task, output the user's computing power resource access permitted computing power Zx and permitted computing power access duration Sc for access execution.

[0097] The step 4 includes the following contents:

[0098] Step 401: After receiving the computing power correction instruction, obtain the real-time redundant computing power Ry i And the real-time computing power abnormality coefficient Ycx, calculate the corrected redundant computing power Jry i :

[0099]

[0100] Step 402: Obtain the computing power requirement Sx and the corrected redundant computing power Jry for the current computing power resource access task of the user. i , calculate the user's current computing power resource access permitted computing power Zx and permitted computing power access duration Sc:

[0101]

[0102] The user's current computing power resource access permitted computing power Zx and permitted computing power access duration Sc are output for access execution.

[0103] When using, combine the contents in steps 401 and 402:

[0104] After receiving the computing power correction instruction, obtain the real-time redundant computing power Ry i And the real-time computing power abnormality coefficient Ycx, calculate the corrected redundant computing power Jry i , combined with the computing power demand Sx of the user's computing power resource access task, output the user's computing power resource access allowed computing power Zx and allowed computing power access time Sc for access execution. During the peak period of computing power demand, the system dynamically allocates and corrects redundant computing power to ensure that the training task is completed on time.

[0105] The present application provides another specific embodiment:

[0106] Video rendering task computing power estimation: render a 10-minute 4K video (60fps), each frame needs ray tracing calculation, single frame calculation amount: 1TFLOPs, redundancy coefficient: 1.5

[0107] Total calculation amount:

[0108] 10 minutes x 60 seconds / minute x 60fps x 1TFLOPs / frame*1.5 = 216,000*1.5TFLOPS

[0109] The above embodiments can be realized wholly or partially by software, hardware, firmware or any other combination. When realized by software, the above embodiments can be realized wholly or partially in the form of a computer program product. Those skilled in the art can realize that the units and algorithm steps of the examples described in conjunction with the embodiments disclosed herein can be realized by electronic hardware or a combination of computer software and electronic hardware. Whether the functions are realized in hardware or software depends on the specific application and design constraints of the technical solutions.

[0110] The units described as separate components can or can not be physically separated, and the components shown as units can or can not be physical units, i.e., they can be located in one place or distributed on multiple network units. Part or all of the units can be selected to achieve the purpose of the embodiments according to actual needs.

[0111] The above is only a specific implementation of the present application, but the protection scope of the present application is not limited thereto, and any skilled person in the art can easily think of changes or replacements within the technical scope disclosed in the present application, which should be covered within the protection scope of the present application.

Claims

1. A method for accessing computing resources in an intelligent computing center, characterized by: The steps include: Identify the user's identity, query the corresponding access rights based on the user's identity, obtain the single task computing amount Dj and the number of concurrent tasks Bf of the user's current computing resource access task, and calculate the computing power requirement Sx of the user's current computing resource access task; Calculate real-time redundant computing power Ry i , and obtain the unit maximum computing power access Fw corresponding to the computing power resource access permission level of the user's current computing power resource access task, and issue an auxiliary access analysis warning; Obtain the continuous computing power Cs and real-time computing power Ys of the computing resources of the intelligent computing center based on the operation log i , calculate the real-time redundant computing power Ry i : Ry i =Cs-Ys i Where i represents the time of the data; After receiving the auxiliary access analysis warning, it calculates the real-time computing power excess coefficient δ, the memory usage Nz and bandwidth usage Dz of the real-time unit computing power, and calculates the real-time computing power abnormality coefficient Ycx, and sends a computing power correction instruction. After receiving the auxiliary access analysis warning, obtain the computing power requirements Sx of each computing power resource access task in the access execution state a , computing power access per unit time Fw a , computing power access time Fs a And computing power access task completion progress Jd a , calculate the real-time computing power excess coefficient δ: Where a represents the sequence number of each computing resource access task in the access execution state, a=1, 2, ..., m, m is the total number of computing resource access tasks in the access execution state, and the computing resource access task status includes access request, access execution, and access completion; Obtain the cumulative computing power access volume Fw of each computing power resource access task in the access execution state based on the operation log a , cumulative memory usage Nc a and cumulative bandwidth usage Dk a , calculate the memory usage Nz and bandwidth usage Dz of real-time unit computing power: Get the real-time computing power excess coefficient δ, the real-time unit computing power memory usage Nz and bandwidth usage Dz, and calculate the real-time computing power abnormality coefficient Ycx: Among them, Δδ represents the standard computing power excess coefficient, ΔNz represents the memory usage of the standard unit computing power, and ΔDz represents the bandwidth usage of the standard unit computing power; Obtain the real-time computing power abnormality coefficient Ycx. When the real-time computing power abnormality coefficient Ycx exceeds 0.2, send a computing power correction instruction. When the real-time computing power abnormality coefficient Ycx does not exceed 0.2, the real-time redundant computing power Ry i Access and execute the user's current computing resource access task; After receiving the computing power correction instruction, obtain the real-time redundant computing power Ry i And the real-time computing power abnormality coefficient Ycx, calculate the corrected redundant computing power Jry i , combined with the computing power requirement Sx of the user's computing power resource access task, output the user's current computing power resource access permitted computing power Zx and permitted computing power access duration Sc for access execution; After receiving the computing power correction instruction, obtain the real-time redundant computing power Ry i And the real-time computing power abnormality coefficient Ycx, calculate the corrected redundant computing power Jry i :

2. The method for accessing computing resources of an intelligent computing center according to claim 1, characterized in that: Obtain the single-task computing amount Dj and the number of concurrent tasks Bf for the user's current computing resource access task, and calculate the computing power requirement Sx for the user's current computing resource access task: Sx=Dj*Bf*α Among them, α is the safety redundancy coefficient, which is used to reserve resources to cope with sudden loads.

3. The method for accessing computing resources of an intelligent computing center according to claim 1, characterized in that: Get the unit maximum computing power access Fw corresponding to the computing power resource access permission level of the user's current computing power resource access task. If the unit maximum computing power access Fw is less than the real-time redundant computing power Ry i , then the user's current computing power resource access task is accessed and executed according to the unit maximum computing power access. If the unit maximum computing power access Fw is not less than the real-time redundant computing power Ry i , an auxiliary access analysis warning is issued.

4. The method for accessing computing resources of an intelligent computing center according to claim 1, characterized in that: Obtain the computing power requirement Sx and corrected redundant computing power Jry for the user's current computing power resource access task i , calculate the user's current computing power resource access permitted computing power Zx and permitted computing power access duration Sc: The user's current computing power resource access permitted computing power Zx and permitted computing power access duration Sc are output for access execution.

Images