443 results about "Load balancing (computing)" patented technology

An AI orchestration system dynamically manages multiple artificial intelligence (AI) agents within a cloud computing environment to efficiently process user requests. A model orchestration subsystem determines whether a request is handled locally using a domain-specific database or by invoking one or more AI agents. The system maintains AI agents in active and inactive states, provisioning computing resources for inactive agents as needed. Real-time model metrics guide the selection of target AI agents, and if a degrading performance trend is detected, the system preemptively spins up additional AI instances. The system provisions processor cycles, memory, and network bandwidth through a cloud-based resource manager, instantiates containerized execution environments or virtual machines, and performs automated load balancing among AI instances.

The invention discloses an energy-efficient task scheduling method for an edge computing system, which comprises the following steps of: acquiring a task state, a computing node resource state, a link state and an energy consumption state according to a unified time slot under a computing power network control domain, and constructing a system state vector; performing priority evaluation on the to-be-scheduled task based on the residual delay budget, the candidate node energy efficiency coefficient and the queue position to obtain a target task set; inputting a system state vector and a target task set into an energy efficiency perception deep reinforcement learning scheduling model, outputting a task-node allocation decision under the constraint of computing node resources and task time delay, and introducing a system-level energy consumption ratio, self-adaptive energy consumption penalty and exploration bias facing high-energy-efficiency nodes into rewards; and scheduling tasks according to the allocation decision, recording state transition and instant rewards, updating a double-commentator and actor network, and performing iterative execution in continuous time slots. According to the method, task success rate, time delay, load balancing and energy-saving performance are considered, and system energy consumption is reduced.

The present application relates to the technical field of large-scale data processing, and particularly relates to a parallel computing method and system suitable for large-scale data processing. The system comprises a task management unit, a distributed load balancing module, an elastic expansion architecture, an intelligent communication optimization module, and a resource monitoring unit, wherein the task management unit divides large-scale data into a plurality of sub-tasks by means of a task decomposer, and distributes the sub-tasks to computing nodes by means of a task scheduler and a priority distributor; the distributed load balancing module achieves global load balancing by means of a load sensing unit, a dynamic adjustment unit and a balance optimization unit; the elastic expansion architecture dynamically adjusts system resources by means of a node manager, a resource pool controller and an expansion decision-making device; the intelligent communication optimization module optimizes inter-node communication by means of a communication path planning unit, a bandwidth distribution unit and a delay compensation unit; and the resource monitoring unit monitors the system performance in real time by means of a performance collector, a state analyzer and an anomaly detector.

The invention relates to the technical field of load balancing optimization, in particular to an intelligent computing center load balancing and energy-saving optimization method and system, and the method comprises the following steps: obtaining task execution duration and resource state data, calculating a multi-dimensional index, carrying out the weighted analysis, dynamically outputting a regulation and control scheme, and achieving the load balancing and energy-saving optimization of an intelligent computing center. According to the method, response time comparison based on a service level protocol is introduced, a task execution duration distribution structure is clearly recognized, the abnormal task recognition precision is improved, and the judgment accuracy of an overload trend is enhanced by comparing resource utilization rate changes of adjacent periods in the aspect of node state monitoring; in the task allocation process, the available memory of the node and the resource adaptation degree are combined, real performance load performance is obtained through energy efficiency and throughput data cross analysis, task allocation better conforms to the actual energy consumption level, the resource waste risk is effectively reduced, the response flexibility of a scheduling strategy to system load changes is enhanced, and the task allocation efficiency is improved. And the task execution efficiency and the energy efficiency distribution accuracy are improved.

The invention discloses a storage and calculation integrated server optimization method based on an NPU, relates to the technical field of server architecture, and discloses the storage and calculation integrated server optimization method based on the NPU by dynamically adjusting the configuration of an arithmetic logic unit, a multiplication and accumulation calculation unit and a cache module to generate a reconstruction calculation unit. By combining heterogeneous hardware performance evaluation and a task migration mechanism, dynamic optimal configuration and load balancing of computing resources are realized, the task execution efficiency can be improved, and the heterogeneous hardware collaborative energy efficiency ratio can be optimized.

The invention relates to the technical field of power system automation, and particularly discloses a cross-domain heterogeneous computing power real-time calling and unified scheduling system for the power industry, which comprises a computing power resource global sensing unit, a business demand dynamic modeling unit, a cross-domain unified scheduling decision unit and a task execution and feedback control unit, according to the method, heterogeneous computing power states are collected in real time, task requirements are dynamically modeled, a scheduling scheme is generated based on multi-objective optimization, and task execution monitoring and load balancing are realized by means of closed-loop feedback, so that the utilization efficiency of global computing power resources is improved, and the real-time performance and reliability of power business are guaranteed.

The invention discloses a numerical control machine tool data acquisition device with distributed edge computing capability and a data processing method. The device comprises a local computing unit and a distributed collaborative management unit, wherein the local computing unit is used for performing data acquisition and processing on a numerical control machine tool and executing a stand-alone analysis task. In the distributed collaborative management unit, a resource discovery and communication module is used for automatically discovering other data acquisition devices of the same type in a local area network and establishing communication connection to form a distributed edge computing cluster; the task collaborative scheduling engine is used for evaluating the real-time calculation load, the resource idle condition and the complexity of a current data processing task of the device, and the cross-equipment analysis module is used for initiating or responding to a cross-machine-tool collaborative analysis task; according to the invention, a distributed edge computing network is constructed, and dynamic distribution, load balancing and cross-equipment intelligent analysis of computing tasks are realized, so that the overall efficiency, the reliability and the intelligent level of a data acquisition system are remarkably improved.

According to the SpMM optimization method and device oriented to the GPU-like architecture, on the premise that the standard CSR storage format is not changed, dual load balancing in the thread bundles and between the thread bundles can be achieved, and computing power waste caused by row length difference of a traditional row-split strategy is eliminated. The method comprises the following steps: (1) inputting a sparse matrix A; (2) calculating the number of non-zero elements in the longest row of the sparse matrix A; (3) judging whether rows with more than 64 non-zero elements exist in the sparse matrix A or not, and if yes, executing the step (4); if not, executing the step (6); (4) binning operation: dividing matrix rows into three categories according to the number of non-zero elements; (5) each category is processed by a corresponding computing kernel, a short traveling box calls a short traveling box kernel to compute, a middle traveling box calls a middle traveling box kernel to compute, and a long traveling box calls a long traveling box kernel to compute; skipping to the step (7); (6) calling kernel calculation in the short-row box; and (7) writing back a result.

A real-time target detection method and system based on RTSP flow and NPU collaborative optimization belong to the technical field of computer vision and artificial intelligence, and are characterized by comprising the following steps: adopting dynamic memory optimization management of a hybrid pipeline architecture, and performing single-time continuous copying through a CPU (Central Processing Unit); through deep integration of innovative technologies such as DMA direct transmission NPU continuous memory pool management, dynamic batch processing scheduling, hybrid assembly line processing, intelligent equipment load balancing, parallel preprocessing optimization and vectorization post-processing, RTSP flow collaborative optimization and the like, the NPU utilization rate is improved to 85% or above, the overall average FPS is improved by 200% or above, the assembly line parallelism degree achieves three times of performance gain, and the production efficiency is greatly improved. The data transmission delay is reduced by 80%, the system stability is remarkably improved, performance degradation is avoided after long-time operation, and the method is suitable for various scenes such as edge calculation and cloud reasoning.

The invention relates to the technical field of calculation, and provides a CFD solution acceleration method and system based on parallel calculation and load dynamic balance. Physical characteristics such as a CFL number and a residual error are quantified into load indexes, so that a load balancing decision and a flow field evolution rule are tightly coupled; by adopting an asynchronous communication overlapping framework deeply coupled with a CFD solver, blocked communication waiting time is converted into effective calculation time, and communication delay is systematically hidden; based on an incremental grid migration strategy of real-time load evaluation, high-load grid blocks in a hotspot process can be intelligently identified and migrated, and differentiation and high-efficiency utilization of computing power are realized.

The application provides a Windows three-dimensional cloud rendering system, method and device, relates to the technical field of cloud computing, and the system comprises a model data acquisition unit which judges whether a corresponding historical rendering model is stored in a cloud server; three-dimensional model data to be rendered and rendering parameters corresponding to the three-dimensional model data are acquired according to a judgment result; a load balancing unit dynamically distributes according to the node state of a service node, and distributes rendering tasks to a plurality of rendering nodes; a task scheduling unit constructs a cloud rendering node cluster according to all the rendering nodes; rendering tools are deployed according to the cloud rendering node cluster, and local resources are called; and a rendering unit utilizes the rendering tools and the local resources to perform rendering according to the rendering parameters. The application performs task acquisition, load balancing, resource scheduling and rendering execution through various units, enhances the reliability and scalability of the system, and improves the cloud rendering efficiency of Windows.

The invention discloses a hybrid expert model training optimization method based on expert calculation load balancing scheduling, and the method comprises the steps: firstly predicting current iteration global expert load data, and obtaining sampling data; and secondly, designing a multi-dimensional performance perception expert placement evaluation method, and realizing quantitative comparison and decision in a stage of searching an expert placement scheme by converting multi-dimensional communication and calculation into a unified time delay index. And then designing an expert placement strategy by adopting an expert grouping search algorithm of dynamic load feature perception and a matching algorithm of expert grouping and resource perception based on the sampling data and an expert placement evaluation method. Finally, on the basis of an expert placement strategy, an asynchronous communication strategy is adopted to implement expert and optimizer parameter scheduling, and training optimization is completed. According to the method, by improving the computing load balance and asynchronous communication concealment between the devices, synchronous waiting is reduced, and the training efficiency and the overall performance of the hybrid expert model are effectively enhanced.

The invention discloses a power equipment anomaly detection method and device, equipment and a storage medium. According to the dynamic graph division method based on computational load awareness, partition layout of load balance is achieved, the problem of load imbalance between processing nodes is remarkably reduced, hotspot bottlenecks caused by high-frequency equipment concentration are avoided, the overall throughput and parallel processing capacity of a system are improved, and the performance requirement for real-time processing of high-frequency data is met. And a constructed partition-unit-device multi-level anomaly detection mechanism triggers the detection task of each level as required, so that the system resource overhead is effectively reduced, and the overall detection efficiency is improved. By analyzing the abnormal propagation path in the electrical topology and combining the spatial position information of the equipment, the abnormal region boundary expression is generated, so that the conversion from point-level identification to chain-level analysis and region-level visualization is realized, the operation and maintenance personnel can quickly position and respond, and the interpretability of the abnormal result and the practical value of scheduling are improved.

The invention provides an intelligent computing center path distribution method, a data transmission method and a network system. According to the technical scheme, based on the time sequence characteristics of the AI training data flow and the characteristic that each set communication library only selects one communication algorithm to drive data flow transmission in the same communication period, the data flow identification capable of identifying the data flow and the communication algorithm corresponding to the data flow identification are obtained, and path planning is carried out based on the communication algorithm. In this way, during actual communication, data streams driven by different communication algorithms on the same path cannot be transmitted at the same time, the link congestion risk is reduced, and idle links are reserved for subsequent data streams of the same communication algorithm; different links are allocated to data flow identifiers with the same source IP address or the same destination IP address corresponding to the same communication algorithm in the same set communication domain, multi-path load sharing is carried out, data flow dispersed transmission based on the same communication algorithm drive is ensured, the congestion risk of a single link is further reduced, and the transmission efficiency is improved. The communication efficiency of an intelligent computing center network is expected to be improved.

Provided in the embodiments of the present application are a data processing method and a related apparatus, which are applied to the field of artificial intelligence and used for partitioning, into a plurality of computing units, a plurality of computing tasks corresponding to a plurality of sub-sequences, so as to achieve load balance of the computing units and implement attention operations. The method comprises: on the basis of the number of computing units, partitioning a sequence into a plurality of sub-sequences; partitioning a plurality of computing tasks of the plurality of sub-sequences into a plurality of computing units, wherein the computing tasks refer to performing attention operations on the basis of the sub-sequences, and the number of computing tasks in each computing unit is the same; and on the basis of a preset communication rule, performing an operation on each computing task in each computing unit, wherein the preset communication rule is used for indicating that a first computing unit among the plurality of computing units communicates with any second computing unit among the plurality of computing units.

The invention discloses a power plant node impedance matrix collaborative correction method based on distributed parallel computing, relates to the technical field of power plant node correction, and aims to solve the problem of poor correction effect caused by inaccurate node analysis in a power plant. Regions are divided in a hierarchical clustering mode based on topological relevance and other characteristics, dynamic optimization is carried out in combination with load balancing and other constraints, node resource accurate allocation tasks are matched, regional topological integrity is guaranteed, calculation loads are balanced, resource waste is reduced, key parameters are synchronized in time through a standardized boundary information exchange mechanism, and the reliability of the system is improved. Coupling errors caused by region segmentation are eliminated, local autonomy and boundary collaboration are achieved, local calculation pertinence is reserved, consistency of electrical characteristics among regions is ensured, global deviation is corrected by combining layered optimization, it is ensured that a matrix conforms to an electrical law, multi-reference verification and four-dimensional index evaluation are conducted, and it is ensured that results are accurate, stable and practical.

The invention relates to the technical field of cross-platform collaboration of an ERP (Enterprise Resource Planning) system, in particular to an ERP operation method for realizing cross-platform collaboration, which comprises a data conversion module, a protocol adaptation module, a task scheduling module and a resource allocation module. The data conversion module unifies heterogeneous data formats, the protocol adaptation module analyzes communication protocols, the task scheduling module dynamically adjusts task priorities, and the resource allocation module coordinates calculation and storage resources. According to the method, through improvements of a weight distribution algorithm, dynamic mapping updating, security verification, load balancing and the like, data integration and task collaboration among multiple platforms are optimized. The flexibility and adaptability of the system can be improved, the technical threshold is reduced, and the efficient and intelligent operation requirement of enterprises is met.

The invention relates to the technical field of meteorological data visualization, in particular to a meteorological data visualization method and device, electronic equipment and a storage medium. According to the meteorological data visualization method and device, firstly, a lightweight multi-resolution data organization unit is established based on a spatial index, a server-side pre-rendering slice is replaced, the storage pressure is reduced, and the data updating flexibility is improved; secondly, a dynamic thread pool and a load balancing mechanism are adopted, data processing tasks are intelligently decomposed and executed in parallel, client computing resources are fully utilized, the processing efficiency is remarkably improved, then coordinate conversion, interpolation and graph drawing are completed in a working thread through an off-screen rendering engine, calculation and rendering decoupling is achieved, and the system performance is improved. And finally, an incremental updating mechanism and a dirty marking technology are introduced, only local redrawing is carried out on a change area, and a cache multiplexing mechanism is combined, so that global rendering overhead is effectively avoided, and real-time, smooth and high-interactivity visualization of mass meteorological data is realized at a Web end.

The invention provides a method and system for efficiently converting an FBX format file into an STEP format file, and relates to the technical field of data format conversion. The method comprises the following steps: dynamically fragmenting an input FBX file, generating fragmented metadata in combination with geometric complexity, a hierarchical structure and a resource load, and distributing the fragmented metadata to a heterogeneous computing thread through load balancing; performing non-manifold edge repair, hole repair, self-intersection surface elimination, material mapping and hierarchical relationship reconstruction on the fragmented data; the preprocessed data are converted into STEP files conforming to the ISO 10303 standard in parallel, geometric consistency and topological closure verification is carried out, and an XML quality report is output; scheduling is carried out based on resource demand labels (CPU intensive type, GPU acceleration type and memory sensitive type), efficient and complete geometric repair and conversion are achieved, and industrial manufacturing and engineering design requirements are met.

The invention discloses a cloud computing task scheduling method, device and system based on deep reinforcement learning, and the method comprises the steps: carrying out the preprocessing of a task, and obtaining the priority of the task; the task with the highest priority is selected according to the priorities of the tasks, partition selection is converted into a Markov decision process, the Markov decision process is executed through a first D3QN agent, and the tasks are distributed to corresponding virtual machine partitions; and converting task scheduling into a Markov decision process, and executing the Markov decision process through the second D3QN intelligent agent to distribute the tasks to the corresponding virtual machines so as to realize task scheduling. A task scheduling model is designed, task scheduling is regarded as a packing problem by the model, and resource requirements of tasks and execution capability of heterogeneous virtual machines are considered at the same time. Meanwhile, reduction of task completion time and maintenance of load balance of a cluster environment are taken as optimization objectives, so that multi-objective optimization of task scheduling is realized.

The invention provides an online interactive programming education system based on a micro-service architecture, and the system comprises an infrastructure layer which serves as a bottom layer computing power and provides an elastic computing power through a cross-cloud deployment technology and a dynamic resource scheduling technology; the platform core layer constructs a stateless micro-service cluster through a service registration discovery module, a load balancing module and a fusing mechanism module depending on elastic resources of the infrastructure layer, and the infrastructure layer provides physical resource guarantee for operation of the micro-service cluster of the platform core layer; the platform support layer is used for integrating an authority management component, a workflow engine component and a general service component based on a stateless micro-service cluster constructed on the platform core layer, and providing resource isolation and high-concurrency task scheduling support in a multi-tenant environment for the programming education system; the platform core layer is combined with distributed transaction management to provide stable service scheduling and communication capability for an authority management component, a workflow engine component and a general service component of the platform support layer; and the platform application layer combines service scheduling and communication capabilities, realizes teaching functions such as multi-language development, code debugging visualization, AI personalized suggestion and collaborative learning, and directly serves programming learning of students and teaching management requirements of teachers.

The invention relates to the field of cloud computing big data secure storage, in particular to a cloud computing-based big data secure storage method and system. The method comprises the following steps: performing encryption and fragmentation processing on an original data stream to generate a fragmentation sequence with an index; generating a dynamic key sequence and a verification key group based on the fragment information; constructing a secure transmission channel through a quantum key distribution mechanism in combination with a geographic position weight and a load balancing strategy, and realizing encapsulation and transmission of a secure fragment packet; node state monitoring and redundancy management are implemented, and an update mapping table is generated; and finally, a complete data stream is recovered through fragment positioning retrieval, data recombination and dynamic decryption. According to the invention, the security and reliability of data storage and the overall load balancing capability of the system are improved.

The invention discloses a distributed document analysis method and device, equipment and a medium. The method comprises the steps that a target to-be-analyzed document is acquired; extracting content feature data of the target to-be-analyzed document; the content feature data is used for reflecting the content complexity of the target to-be-analyzed document; according to the content feature data, determining estimated time consumption for analyzing the target to-be-analyzed document; based on load balancing, according to the estimated time consumption, distributing a target parsing task of the target to-be-parsed document to a target computing power node; and the target computing power node executes the target analysis task to obtain an analysis result. Therefore, the analysis duration is pre-estimated based on the content feature data of the document instead of determining the pre-estimated time consumption singly depending on the size of the document, the method is adaptive to the time consumption fluctuation characteristics of different documents, the precision of the pre-estimated time consumption is improved, calculation power stage distribution is performed based on load balance according to the pre-estimated time consumption, and the calculation efficiency is improved. The resource utilization rate and the load balancing effect of the system are improved.

The invention discloses an artificial intelligence platform computing power resource scheduling management method and system, and relates to the technical field of artificial intelligence computing power scheduling, and the method comprises the specific steps: hardware topology collection traverses heterogeneous resources, and generates and updates a graph; the collaborative strategy decision is based on the atlas, and optimal precision and parallel combination are screened; load balancing deploying split tasks, matching hardware capability and planning a transmission path; executing and monitoring multi-dimensional real-time tracking; the quantization deviation is dynamically adapted and adjusted, and the strategy is timely optimized; according to the invention, through hardware topology acquisition and collaborative strategy decision, a computing power resource optimal configuration scheme is formed, and the training task starting efficiency and stability are improved; load balancing deployment and dynamic adaptation adjustment are combined, resource balancing allocation and whole-process dynamic optimization are achieved, the problems of resource imbalance and response lag are effectively solved, the computing power utilization efficiency is improved to the maximum extent, and efficient and stable training is guaranteed.

The invention discloses a laboratory management and control scheduling system and method, and belongs to the field of laboratory management, and the system comprises an Internet of Things equipment monitoring module, a cloud computing resource scheduling module, an experiment task distribution and progress monitoring module, a data analysis and report generation module, an early warning and exception handling module, and a data security and privacy protection module. Through linkage of the cloud computing resource scheduling module and the experiment task allocation module, accurate matching of equipment, personnel and tasks is realized through an intelligent algorithm, the tasks are automatically allocated based on equipment loads and personnel skills, a scheduling plan can be dynamically adjusted according to real-time progress, the problems of resource idleness and task backlog are solved from the source, and the task scheduling efficiency is improved. The cooperative effects of efficient equipment operation, personnel load balance and ordered task promotion are achieved; equipment and environment data are collected in real time through the Internet of Things equipment monitoring module, and the early warning and exception handling module synchronously monitors and responds to exceptions.

The invention relates to a multi-agent reinforcement learning training method, system and device driven by an unreal engine and a medium. The method comprises the following steps: acquiring core attribute data of multiple agents, including real-time communication bandwidth, initial communication topology configuration, training stage progress, single agent computing power and cluster overall computing load state; analyzing the real-time communication bandwidth and the initial communication topology configuration to form an optimized topology configuration scheme; in combination with an optimized topology configuration scheme and a training stage progress, generating a cooperative training progress state by quantifying progress differences among intelligent agents; extracting load balance data according to the single agent computing capability and the cluster load state; integrating load balance data, a collaborative training progress state and an optimized topology configuration scheme, and constructing a multi-layer collaborative training framework; and integrated optimization is carried out to generate a cooperative training scheme for guiding multi-agent communication, training and resource allocation, so that the training efficiency is effectively improved, the resource load is balanced, and the cooperative synchronism is enhanced.

The invention discloses a GPU cluster resource allocation method and device, equipment and a storage medium, and relates to the technical field of GPU cluster resource allocation. According to the method, a hardware performance description vector is generated by fusing static hardware parameters and dynamic micro-benchmark test data, the limitation that the GPU performance is represented only by depending on the static parameters is broken through, and the actual operation capacity of different architecture GPUs in a heterogeneous cluster is accurately matched; task feature vectors are generated by extracting task calculation features and resource demands, and quantitative description of reasoning task resource consumption features is achieved; information of hardware, tasks and load dimensions is integrated through a machine learning model, performance degradation characteristics during multi-task parallel are effectively captured, and the accuracy of execution time prediction is improved; scheduling decision operation including candidate node screening, comprehensive cost evaluation and resource reservation backfilling is executed in combination with the predicted execution time, and the task execution efficiency and cluster load balancing are both considered.

The invention discloses a hotspot sensing hierarchical storage and computing resource scheduling method oriented to multi-modal data, and belongs to the technical field of data center resource management. The core of the method is as follows: firstly, monitoring characteristics, access modes and computing node loads of multi-modal data in real time; secondly, a comprehensive popularity evaluation model is used for dynamically calculating the comprehensive popularity value of the data object, and the model comprehensively considers the data type, the access frequency, the data size and the task priority; then, intelligently scheduling the data to different performance levels of the hierarchical storage system according to the popularity value; and finally, when a calculation task is scheduled, synchronously sensing the heat and the position of required data, and carrying out collaborative scheduling by taking optimal data locality and load balance as targets. According to the method, storage and computing resources are managed cooperatively through the unified heat index, the problem of resource scheduling disjunction in a multi-modal data environment is effectively solved, and the system throughput rate, the response speed and the resource utilization efficiency are remarkably improved.

Embodiments of the invention provide a model input request scheduling method and device, a storage medium and a program product. The method comprises the steps of receiving an input request of a preset language model; determining prediction time consumption of the input request in a pre-filling stage in a preset language model processing process; determining the latest execution time of the pre-filling stage of the input request according to the receiving time of the input request, the predicted consumed time of the pre-filling stage and the target first lexical element delay; and according to the latest execution time of the pre-filling stage, sending the input request to a service node of a preset language model for processing. According to the embodiment of the invention, the latest execution time of the pre-filling stage is determined based on the predicted time consumption of the pre-filling stage of the input request to schedule the input request, and the system throughput and the calculation load balance are optimized while the service level target is met, so that the resource utilization efficiency and the service quality guarantee capability of the service node of the preset language model can be improved.