121 results about "Cpu load" patented technology

The invention discloses an industrial personal computer and multi-graphics card collaborative parallel computation acceleration system, which relates to the technical field of industrial resource allocation and parallel computation, and comprises a resource monitoring and predicting module, a resource management module and a prediction type resource preparation module, the task splitting and collaborative execution module comprises a task splitting module, a cross-node collaborative module and a collaborative operation engine; the intelligent scheduling and dynamic resource allocation module comprises an intelligent scheduler, a dynamic resource allocation module and a conflict avoidance module. According to the method, the GPU video memory utilization rate, the core utilization rate, the temperature, the video memory fragment rate, the available video memory total amount, the CPU core total utilization rate, the load condition and the idle core number index are collected in real time through a resource monitoring module, and the video memory capacity, the GPU core occupancy rate and the CPU load requirement are predicted in advance before a task is submitted in combination with a gradient boosting decision tree and a neural network prediction model; and resources are reserved, so that the scheduling delay is remarkably reduced, and the scheduling hit rate is improved.

The invention discloses a sintering process real-time optimization method based on edge calculation, and the method comprises the steps: obtaining original multi-index time series data collected at the edge side of a sintering process in real time, and recording sampling data and corresponding process labels for sensing points at different workshop section positions in a sintering production line; based on the real-time prediction task load, the edge node computing power and the network bandwidth monitoring result, whether the current node reaches an elastic migration triggering condition or not is judged, and if yes, subsequent task migration operation is executed; the calculation and storage quotas of edge nodes are dynamically adjusted based on the GPU / CPU load, the storage space and the prediction task fluctuation of each node; according to the invention, by capturing the deep relationship of time-varying coupling between process indexes, an elastic and dynamic resource scheduling and model adaptive mechanism is constructed, the edge-side dynamic graph modeling capability is enhanced, and a model self-learning optimization mechanism is introduced; the accuracy, the real-time performance and the edge system adaptability of multi-index collaborative prediction in the complex industrial process are effectively improved.

The invention discloses an AI-driven RDMA network card hardware unloading acceleration method and system based on an RISC-V architecture, the system comprises an RISC-V processing unit, the RISC-V processing unit comprises a control module and a driving module, and the driving module stores RDMA exclusive acceleration instruction set extension; the AI reasoning unit is connected with the driving module and is used for expanding and running a pre-trained network optimization model based on an RDMA exclusive acceleration instruction set to analyze a network state and an application requirement, generating a hardware unloading strategy optimization instruction and sending the hardware unloading strategy optimization instruction to the control module; the control module generates a corresponding management command according to the hardware unloading strategy optimization instruction; and the hardware unloading unit comprises a management module and an RDMA component module, and the management module receives the management command and performs parameter configuration on the RDMA component module according to the management command. According to the invention, the burden of the CPU is reduced, and the throughput is improved.

The invention discloses a large model deployment method based on heterogeneous resource scheduling and multi-model collaborative reasoning, and relates to the field of artificial intelligence and server system deployment. Comprising the following steps of: 1, calculating a deployment score of a target model by utilizing a deployment scoring function according to a GPU video memory residual rate, a CPU load, network delay and a current task quantity of each node through a heterogeneous resource sensing scheduler, and selecting a node with the highest score of the target model to load or reuse the target model; a cache index key is generated, and whether key value KV pairs supporting multiplexing of the target model exist or not is inquired; step 3, if the cache is hit, skipping a forward propagation stage, and directly decoding based on a cache result, otherwise, executing complete forward reasoning and writing the result into the cache; 4, based on a plurality of user requests, performing fusion according to user priorities, model weights, reasoning costs and cue word lengths to generate execution sorting weights, and scheduling execution models in batches according to the weights; and 5, returning the model response to the user, and releasing or updating node state information.

The invention discloses a method for dynamically setting dog feeding time of a watchdog circuit, which comprises the following steps of: collecting and monitoring power consumption of a CPU (Central Processing Unit) and a corresponding load, and fitting the load and power of the CPU according to monitoring data so as to reflect the running state condition of a system; power consumption of a monitoring CPU is collected in real time, the real-time CPU load is calculated according to the fitting relation between the CPU load and power, and the watchdog feeding time of a watchdog circuit is determined according to the CPU load; a square wave period is set, square waves are collected to set timeout time, and the timeout time is kept consistent with dog feeding time, determined according to CPU loads, of a watchdog circuit. According to the method, the CPU power consumption and the load are fitted, the CPU load is monitored in real time, the square wave period is adjusted, the dog feeding time is dynamically adjusted according to the CPU load, resource waste is avoided, and the method is suitable for the fields of embedded systems, industrial control and the like.

The invention discloses a resource allocation method and device based on bit map pre-calculation, and the method comprises the steps: generating a global resource state bitmap according to the CORESET configuration in a PDCCH, and carrying out the pre-calculation of each possible candidate PDCCH to form a candidate pre-allocation bit map set; extracting a corresponding pre-allocation bitmap from the candidate pre-allocation bitmap set as a candidate set bitmap; and comparing the candidate set bitmap with the global resource state bitmap, judging whether the resources are conflicted or not, and if the resources are not conflicted, determining that the resources are available and further occupying the resources. Complex function calculation at the scheduling moment is transferred to pre-calculation at the initialization stage, only simple bitmap retrieval and bit operation are needed during scheduling, the PDCCH resource allocation time consumption is greatly shortened, real-time intensive calculation is replaced with pre-stored bitmaps, the CPU load is remarkably reduced, a processor focuses on a core scheduling algorithm, resource redundancy consumption is reduced, and the scheduling efficiency is improved. And the overall capacity and the energy utilization efficiency of the communication system are effectively improved.

Systems and methods are disclosed for optimizing network routing and data processing through dynamic route optimization. The invention features a dynamic route optimization engine that uses real-time performance metrics such as latency, bandwidth availability, memory utilization, CPU load, and throughput capacity. It allows for rule-based configurations, enabling custom parameters for transaction processing, data retrieval, and user authorization. Integration with network monitoring tools provides continuous visibility into network performance. The system supports multi-cloud and hybrid cloud environments, dynamically selecting optimal paths based on real-time conditions and regulatory compliance. Predictive analytics anticipate network congestion, allowing proactive routing adjustments. Adaptive algorithms learn from past conditions to optimize performance continuously. Scalability is achieved through automated updates to routing tables and security groups. The invention reduces latency and ensuring efficient data transfer, providing a cost-effective and reliable solution for modern network management, addressing a long-standing need for adaptability and efficiency in network routing.

The invention relates to a cloud computing resource load prediction method based on TCN-PatchTST (Trusted Cryptographic Network-PatchTST). Comprising the following steps: collecting CPU load time sequence data to form a training sample set; carrying out preprocessing and feature construction on the training sample, and selecting a memory utilization rate load value which has obvious influence on a CPU load value as an input feature to predict the CPU load value; according to the method, a TCN-PatchTST neural network is constructed, a two-channel feature extraction architecture is designed, local time sequence features are extracted through causal convolution and expansion convolution of TCN, a long-term dependency relationship is captured by using partitioning processing and a channel independence mechanism of PatchTST, and a feature fusion layer is designed to integrate multi-scale information; the TCN-PatchTST neural network is trained; and inputting a test sample into the trained TCN-PatchTST neural network to obtain a CPU load value prediction result of the test sample. According to the method, load prediction is carried out through the historical load time sequence data, the prediction precision of the CPU load value of the cloud computing resources is effectively improved, and technical support is provided for intelligent scheduling of the cloud resources.

The invention discloses an intelligent task scheduling and collaborative execution method and a heterogeneous computing system, and the method comprises the steps: receiving and analyzing a computing task, and extracting the computing, data and constraint features of the computing task; decomposing the task into sub-tasks with a dependency relationship; acquiring load, computing power and power consumption states of the NPU, the GPU and the CPU in real time; on the basis of task characteristics and real-time states, multiple allocation schemes are evaluated through a pre-constructed cost model, and an optimal scheduling decision for allocating the sub-tasks among the heterogeneous computing units is dynamically generated; a flow line collaborative execution plan across at least two computing units is arranged according to a decision, sub-task execution is scheduled, meanwhile, cross-unit data flow and synchronization are managed, and a corresponding system comprises an NPU, a GPU, a CPU, a system memory and an intelligent heterogeneous resource scheduler for executing the method. According to the method, fine scheduling, self-adaptive load balancing and multi-objective optimization of heterogeneous computing resources are realized, and the comprehensive efficiency and energy efficiency of the system are remarkably improved.

The invention provides a software operation synchronization method and system based on a dual-chip architecture, and abandons a traditional mode depending on a UART / SPI / CAN communication protocol or an additional synchronization circuit. Firstly, a unique PWM duty ratio mapping relation with any difference not smaller than 5% is preset for the initial state, the system self-checking state, the OTA upgrading state, the running state, the fault state, the dormancy preparation state and other running states of the double-chip system, and then a chip A and a chip B continuously output PWM signals matched with the current states of the chip A and the chip B through common GPIO pins in a push-pull output mode. Meanwhile, PWM signals of GPIO pins of the opposite side are collected in real time in a bidirectional mode through respective hardware capture units, the duty ratio is analyzed to obtain the state of the opposite side, response operation is executed according to preset logic, synchronization failures caused by communication delay and data frame error codes of a traditional software protocol are effectively avoided, the system stability is guaranteed, meanwhile, a complex protocol stack does not need to be designed, and the cost is reduced. The CPU load rate is reduced, an additional synchronous circuit is omitted, the chip area and power consumption are reduced, and the hardware cost is reduced.

The embodiment of the invention provides an intelligent equipment temperature control method and device, equipment, a medium and a program product, and relates to the technical field of intelligent temperature control. Performing feature engineering processing on the training data to obtain feature engineering data; wherein the characteristic engineering data comprises a temperature change rate, a CPU load and a historical temperature control mode; combining weighted splicing and attention mechanism fusion feature engineering data to obtain fusion features; based on the fusion features, training an LSTM long short-term memory network model; using an LSTM model to predict a temperature rise trend in a first set time period; adjusting the rotating speed of the cooling fan based on the predicted temperature rise trend; the intelligent dynamic control of the fan is realized, the response and control precision is improved, and the energy consumption is reduced.

A method for quality of service (QoS) flow control dynamically optimizes network resource allocation based on terminal power consumption status. An access network function receives terminal status information from an application function (AF) and determines a QoS profile for associated QoS flows from alternative profiles. The selected QoS profile is transmitted to a core network function or terminal for QoS updates. The terminal status information, which may include battery level, CPU load, or overheating status, informs the selection process to balance performance and efficiency. Core network functions manage QoS updates by interacting with the AF, access network, and terminals. The method, apparatus and computer readable medium ensure adaptive and energy-efficient QoS management for improved user experience and resource utilization in diverse network environments.

The invention provides a system for secure encryption of CPU boot firmware and a working method thereof. The system comprises a CPU unit, an MCU unit, a FLASH unit and an electronic switch unit. By introducing an MCU unit, conventional CPU boot firmware is firstly subjected to data encryption processing and then stored in a first FLASH chip to complete initial configuration, in the system starting process, the MCU firstly reads, decrypts and restores the securely encrypted boot firmware in the first FLASH chip into conventional boot firmware, and writes the conventional boot firmware into a second FLASH chip, so that the boot firmware in the second FLASH chip is subjected to data encryption processing; and then the CPU loads a conventional boot firmware starting system from the second FLASH chip, and after the CPU is started, the MCU erases the boot firmware in the second FLASH chip. Only the first FLASH chip stores the securely encrypted boot firmware, and the second FLASH chip only temporarily stores the boot firmware in the starting stage, so that the risk of physical leakage and cracking of the boot firmware is effectively prevented.

The invention discloses a software and hardware collaborative RDMA atomic operation acceleration method, which comprises the following steps: dynamically collecting and reporting atomic operation target address popularity through a client, and dynamically formulating a table locking slot scheduling strategy by a server in combination with a CPU load according to the popularity so as to realize a fine-grained unloading decision; the client carries out real-time scheduling according to a strategy; a low-heat slot is executed by a network card hardware atomic instruction, and a high-heat slot is redirected to an RPC to be processed by a server CPU; the switching process depends on a client side consensus and a global consensus; and the RPC thread of the server side verifies the strategy identifier and the consensus state before execution, and rejects an inconsistent request. According to the method, performance optimization of RDMA atomic operation is realized through heat perception, dynamic scheduling, network card and CPU collaborative execution and a two-stage consensus mechanism.

The application discloses a Kalman filtering state decoupling and parallel computing method for multi-target tracking, and comprises the following steps: S1, generating several two-dimensional Kalman subsystems according to the original high-dimensional state space and column vectors of Kalman filtering hardware; S2, constructing two-dimensional Kalman filtering processing units for each two-dimensional Kalman subsystem, and outputting a high-dimensional state vector; and S3, generating a high-dimensional target state estimation result according to the high-dimensional state vector. When the multi-target tracking algorithm is applied, the parallel two-dimensional Kalman filtering processing array is simultaneously executed in the FPGA, and the processing delay of single filtering is greatly shortened. The soft and hard collaborative architecture completely sinks the heavy matrix operation to the FPGA, significantly reduces the CPU load, enables the CPU to focus on upper logic control, and thus realizes real-time multi-target tracking with high frame rate and low delay on the end-side heterogeneous platform.

The application provides a CPU load balancing-based inter-server communication scheduling method and system, and relates to the technical field of CPU load balancing, wherein the method comprises: performing quality evaluation on network state data to generate strategy data, and meanwhile converting electromagnetic intensity signals and temperature distribution signals in a computer room into corresponding digital signals to form an environment data set; fusing the strategy data and the environment state data set through a dynamic weight adjustment algorithm to generate a frequency modulation scheme and a channel switching instruction; screening a set of available CPU cores according to load data; and finally distributing data packets to target CPU cores for processing and transmission according to the frequency modulation scheme and the channel switching instruction, establishing multiple concurrent connection channels, and thereby realizing load balancing and high stability transmission between servers. The application improves the communication efficiency and transmission reliability of a server cluster.

The invention discloses a one-master multi-slave time division multiplexing communication method, and relates to the technical field of multi-node real-time communication.The method comprises the steps of star topology construction and configuration, time division multiplexing time sequence establishment, data transmission optimization and transmission fault detection.The method comprises the steps that firstly, a star topology structure is constructed, a multi-core heterogeneous SoC master control core and a full duplex differential bus are configured, and the full duplex differential bus is configured; hardware interconnection verification and core parameter initialization are completed; secondly, establishing a time division multiplexing time sequence, distributing exclusive time slices according to a polling period and calibrating the time sequence to realize ordered periodic data interaction; data transmission is optimized through mechanisms such as a shared cache read-write lock and overflow protection; the method has the advantages of being determined in communication delay, high in data throughput, excellent in resource utilization rate and wide in adaptive scene, the real-time performance is improved, the CPU load is reduced, and the flexibility is further improved.

This invention relates to the field of instrument display technology, and in particular to a method and instrument device suitable for multilingual display of instruments with limited resources. The method includes: an external device assigning globally unique macro definition IDs to translation entries and extracting actually used non-repeating characters to generate a trimmed dot matrix font library; constructing a binary resource package with a direct addressing structure, which includes a file header, an index addressing area, and a data entity area; and a microcontroller directly locating and quickly rendering the target entry using the ID as an offset through non-retrieval addressing. This invention provides a method and instrument device suitable for multilingual display of instruments with limited resources, solving the bottleneck of multilingual display in embedded devices with extremely low resources. It significantly reduces storage consumption and CPU load without requiring runtime parsing and string comparison, and enables independent updating and maintenance of language resources.

The invention provides a zero-trust dynamic encryption routing method, device and equipment based on traffic awareness, and the method comprises the steps: determining an identity feature and a behavior feature corresponding to a first to-be-processed message when the data plane receives the first to-be-processed message through a zero-trust security tunnel; performing decryption operation on the first to-be-processed message through target key information corresponding to the zero-trust security tunnel to obtain a decrypted second to-be-processed message; if the data plane has the first forwarding strategy matched with the identity feature, determining a routing action of the second to-be-processed message based on the first forwarding strategy; operating the second to-be-processed message based on the routing action; and when the identity feature and the behavior feature are received through the control plane, determining a second forwarding strategy matched with the identity feature based on the behavior feature, and updating the second forwarding strategy to the data plane. Through the scheme of the invention, on the premise of keeping zero trust, the data volume mirrored to the control plane and the CPU load are reduced.

The embodiment of the invention provides a message sending method and device, electronic equipment and a storage medium, the method is applied to a current node, the current node is any node in a live broadcast P2P system, and the method comprises the following steps: determining a sending interval of a block bitmap message according to the number of connected nodes, the sending interval being represented by a target number of logic fragments, the target number is smaller than or equal to the connected node number; according to the sending interval, when data downloading of one logic fragment in the live broadcast slice is completed, a target node which is currently polled is determined from the nodes connected with the current node, a block bitmap message is sent to the target node, and the block bitmap message is sent to each connected node once in one sending interval. According to the embodiment of the invention, the sending frequency of the block bitmap message is reduced, the updating message of the block bitmap is reduced, and the P2P thread focuses on processing the data downloading and uploading message, so that the pressure of a CPU (Central Processing Unit) is reduced, and the slice live broadcast lagging is reduced.

The present application relates to the technical field of data processing, and relates to a data shunting collection and adaptive storage method and system based on a heterogeneous SoC, to solve the problems of excessively high CPU load and large storage delay when multiple sensors are accessed. The present application accesses multiple sensors at the PL end of a heterogeneous SoC and caches data, sets a shunting threshold to divide the data stream into a high-speed data stream and a low-speed control stream channel, and uses corresponding interfaces for processing to reduce bus load; a logical ring buffer is constructed in the PS end DDR memory, the number of DMA transmission blocks is adjusted according to the occupancy, and after adaptive storage of data is completed in three modes of low delay, standard batch, and extremely fast continuity, the disk read pointer is updated, the buffer area of completed disk writing is released, and the PL end is provided with a new data cyclic writing function. The present application realizes efficient and low-loss storage of a large data stream of multiple sensors through the synergistic effect of a heterogeneous SoC, improves data collection continuity and storage reliability, and prolongs the service life of the storage device.

The invention discloses an NVMeoF far-end storage system and method based on GPU direct command control, an I / O processing task in an NVMeoF environment is clearly divided into two modules, namely a GPU processing NVMe command module and a CPU processing network protocol module, cooperation of the two modules is realized through a doorbell notification mechanism, a CPU is enabled to concentrate on the skilled network protocol processing, and the NVMeoF remote storage system and method based on GPU direct command control are realized. The GPU is enabled to concentrate on the skilled NVMe command generation and data processing, so that the load of the CPU is remarkably reduced, an end-to-end zero-copy data path from remote NVMe storage to a GPU video memory is constructed, unnecessary intermediate data copy is eliminated, and the data throughput is maximized. According to the invention, in an NVMeoF environment, approximate localization low-delay and high-bandwidth access of the GPU to the remote NVMe storage device can be realized.

The application discloses a low-delay media processing method under an Android system with a domestic ARM chip. By constructing a direct data path of "Java application layer->JNI bridging layer->native C / C++ layer->kernel layer hardware driver", the high-delay system services such as Android MediaServer and / or AudioFlinger are bypassed. Specifically, the method is realized through three steps: integrating a bottom media library and writing a packaging module, establishing a JNI bridging layer to expose native layer functions, and creating an adapter class compatible with the Android standard MediaCodec API for application calling. The scheme reduces the streaming media output delay to below 20 milliseconds, significantly reduces CPU load and memory bandwidth occupation, and adapts to the upper application without substantial modification, breaks the real-time performance bottleneck of the Android system, expands the professional application scenarios of the domestic ARM chip Android device, and improves the system running efficiency and market competitiveness.

A PLC high-frequency data management system based on a built-in database comprises an embedded database, a database operation engine, a data recording function block, a database operation function block, an annular memory cache region, a nonvolatile storage medium and a communication service module during PLC operation. A lightweight database and an intelligent cache mechanism are deeply integrated in the PLC, and high-frequency real-time small data packets are written and converted into low-frequency batch large transaction operation, so that high-frequency, stable and structured recording and convenient query of process data are realized on the premise of ensuring control instantaneity and low CPU load.

Motor closed-loop detection circuit and method are provided. The circuit includes: motor open-loop path for providing open-loop angle increment and calculating open-loop angle; motor prediction path for estimating motor's angle increment to obtain estimated angle increment and calculating estimated angle; and closed-loop detection module for outputting first closed-loop flag signal when the open-loop angle increment equals the estimated angle increment n consecutive times within an electrical cycle feeding back the first closed-loop flag signal to the motor open-loop path to adjust the open-loop angle increment, and outputting second closed-loop flag signal when the updated open-loop angle equals the estimated angle. The hardware circuit and existing motor paths reuse simplify the design, reduce costs, and minimize CPU load and power consumption. Shared components across modules from a computing library save costs. The method is simple, fast, and easy to implement. A same circuit achieves dual-level closed-loop functionality with minimal area consumption.

The invention relates to a secure atomic energy elastic capacity expansion and contraction and flow distribution method, device, equipment and medium, and is applied to the technical field of resource management, and the method comprises the steps: carrying out the category division of instances according to historical load conditions and business types, and determining a plurality of instance clusters and the instance category of each instance cluster; constructing a traffic prediction model of each instance category; predicting a predicted request number and a predicted CPU load rate of the target instance cluster based on the traffic prediction model; determining a capacity expansion and contraction threshold value of the target instance cluster based on the target instance category; based on the predicted CPU load rate of the target instance cluster, the predicted request number and the capacity expansion and contraction threshold value, whether capacity expansion and contraction need to be carried out or not is judged; and if capacity expansion and shrinkage need to be carried out, determining a capacity expansion and shrinkage strategy of the target instance cluster based on the predicted request number and the target instance category, and determining a traffic distribution strategy of the target instance cluster based on a preset distribution rule. The method has the effect of improving the reliability of resource management.

The application discloses a multi-task execution method, device, equipment and medium, and belongs to the technical field of power electronics. The method comprises the following steps: setting independent threads for each operation task in an IEC61131-3 multi-operation task in advance; setting a task priority list for each operation task in the IEC61131-3 multi-operation task, and setting a target reference list according to the data calling relationship of each operation task; and sending the task priority list and the target reference list to a programmable logic controller, so that the programmable logic controller controls each operation task in the IEC61131-3 multi-operation task according to the task priority list and the target reference list. By using the method, the programmable logic controller can execute each operation task in the IEC61131-3 multi-operation task without being affected by other operation tasks, and the CPU load is more stable.

The invention relates to a multi-core CPU load balancing scheduling method based on dynamic queue migration and electronic equipment, and the method mainly comprises the following steps: firstly, initializing a physical core set, establishing a corresponding virtual machine identification task queue for each physical core, and sorting according to a core number; when a vCPU allocation request of a virtual machine is received, the queue length of each physical core is calculated, and a virtual machine identifier is inserted into a corresponding physical core queue according to a shortest queue priority principle; counting the queue lengths of all physical cores, and calculating the longest queue length and the shortest queue length; judging whether load unevenness (no idle physical core and inconsistent queue length) exists or not, if so, executing differentiated queue rebalance according to length difference, and migrating tasks in a long queue to a short queue; and finally, updating the binding relationship between the vCPU and the physical core to ensure that the load balancing takes effect. Compared with the prior art, the method has the advantages that the computing resources of the virtual machine are distributed fairly, and the physical core utilization rate and the overall system throughput are improved.

The invention provides a CPU load regulation and control method and device, electronic equipment and a storage medium. The CPU load regulation and control method comprises the steps that the number of MAC cores occupied by a current cell is acquired, and the CPU load of the current MAC cores is acquired; determining a current CPU load regulation and control strategy based on the MAC core number and the CPU load; the regulation and control strategy comprises an MAC core adjustment strategy and a single time slot scheduling user number upper limit adjustment strategy; and regulating and controlling the CPU load according to the CPU load regulation and control strategy. According to the CPU load regulation and control method and device, the electronic equipment and the storage medium provided by the invention, the upper limit of the number of scheduled users in a single time slot or the number of MAC cores occupied by a cell is adjusted by executing the adjustment strategy determined in real time, so that accurate CPU load regulation and control in different scenes are realized, the possibility that base station service processing is abnormal is reduced, the stability of MAC message sending is ensured, and the user experience is improved. And user experience is improved.