18 results about "Concurrent computing" patented technology

The invention provides an acceleration processing method and device for sparse matrix vector multiplication. The method comprises the following steps: acquiring a sparse matrix; dividing the sparse matrix into segments, and distributing threads for the segments; a matrix multiplication-accumulation instruction in a preset instruction set architecture is called, a tensor calculation core Tensor Core is used for carrying out matrix multiplication calculation of small blocks on the fragments, result data are obtained, and the result data are used for representing vector data objects of the linear equation set solution vectors. According to the method, the obtained sparse matrix is divided into the fragments and the threads are distributed, so that Tensor Core concurrent calculation is realized, the calculation efficiency is greatly improved, and the calculation time is remarkably shortened especially for solving a large-scale linear equation set.

The present invention provides a risk-weighted asset measurement method, which relates to the field of risk-weighted asset calculation. The method includes information acquisition, data preprocessing and blocking, and calling multiple computing units to form distributed computing. The method converts the entire measurement process into a data simplification and blocking processing and multi-unit concurrent computing mode, wherein the data simplification and blocking preprocessing step divides the original information into simplified versions of multiple block information, and then the multiple block information is respectively handed over to multiple computing units for processing, thereby realizing multi-threaded synchronous processing. The method has strong scalability and can use multiple computing units according to actual conditions, and has strong flexibility.

This application provides a method, device, apparatus, and storage medium for selecting computing power nodes. The method includes: acquiring multiple undetermined service requests received through the same entry computing power routing node; and determining, based on whether the senders of the multiple undetermined service requests are the same user, selecting target computing power nodes that meet the requirements of each undetermined service request according to a first rule or a second rule. This application addresses specific application scenarios by prioritizing the processing of multiple service requests from the same user using the same or similar computing power nodes, thereby achieving parallel data forwarding, reusing transmission resources, and improving resource utilization. Furthermore, by using different computing power nodes belonging to different computing power routing nodes to process service requests of the same type from multiple users as much as possible, it avoids service paralysis across the entire region due to faults, thus accurately and specifically selecting the optimal computing power node for multiple concurrent computing power services in specific application scenarios.

The application provides a data processing method, device and equipment and a storage medium. The method comprises the following steps: obtaining to-be-processed data and a pre-designed calculation graph corresponding to the to-be-processed data; the pre-designed calculation graph represents the relationship between a plurality of operators, the plurality of operators comprise a connection operator and a task operator determined by a data processing task of the to-be-processed data, the task operator comprises an upstream task operator and at least two concurrent task operators; based on the upstream task operator, the to-be-processed data is processed to obtain an upstream data processing result; based on the at least two concurrent task operators, the upstream data processing result is processed concurrently to obtain a concurrent data processing result corresponding to each concurrent task operator; based on the connection operator, the concurrent data processing results corresponding to each concurrent task operator are combined to obtain a data combination result. The application can realize concurrent calculation of a single message among a plurality of operators, shorten the message processing time and improve the system response speed.

This application provides a dynamic empirical route generation method, relating to the field of intelligent scheduling and path planning technology for new energy commercial vehicles. The method includes: collecting multi-dimensional state data across the entire process; filtering valid data from the multi-dimensional state data to generate first state data; constructing a hierarchical energy consumption prediction model, which includes a lightweight model and a high-precision model; employing the NSGA-II algorithm and particle swarm optimization algorithm, with the objective functions of minimizing overall energy cost, minimizing time consumption, and minimizing refueling waiting time, and combining this with the hierarchical energy consumption prediction model to predict energy consumption, thereby generating a candidate route set; determining the total empirical matching score for each candidate route in the candidate route set, and determining the optimal route based on the total empirical matching score. This solution ensures the reliability of range prediction under extreme operating conditions while significantly reducing the load pressure of large-scale concurrent computing in the cloud.

A method, apparatus, medium, and product for generating distributed tile data are disclosed. The method includes, based on the concurrent computing power of a computing cluster, assigning multiple location point information of location data to corresponding computing nodes of the computing cluster to concurrently determine multiple first tile identifiers corresponding to at least some of the multiple location point information, and multiple pixels for the multiple location point information, wherein at least two location point information corresponding to the same first tile identifier are assigned to different computing nodes of the computing cluster. The method also includes, based on the multiple first tile identifiers, assigning multiple pixels to corresponding computing nodes of the computing cluster to concurrently generate multiple first tiles with multiple first tile identifiers, each first tile being uniquely identified by its corresponding first tile identifier, wherein at least two pixels corresponding to the same first tile identifier are assigned to the same computing node of the computing cluster.

The high-performance computing method of CALPUFF is realized based on workflow, which relates to the fields of atmospheric environment and computer processing technology. It solves the problems that the existing meteorological models have poor processing capabilities for abnormal situations, cannot support concurrency in calculations, have low computing efficiency, and are insufficient in analyzing and processing calculation results; and cannot support calculations for complex situations of multiple models, and can only solve single problems with a single script. The present invention is implemented based on an atmospheric model computing service system. According to the workflow model, a computing task is divided into multiple task nodes, which are completed one by one. This method integrates complex data, data analysis and processing, and realizes cloud service sharing. The atmospheric model computing service system of the present invention adopts a numerical simulation method to analyze the causes of atmospheric pollution under different climatic conditions and at different times within the range of the numerical model, and to identify the contribution rates of various pollution sources, various types of pollution sources, and each grid pollution source that affect atmospheric pollution, providing a decision-making support system for atmospheric pollution prevention and control.

Methods and systems for implementing division of process resources of running processes into individually locked partitions, and indirect mapping of keys to process resources to locks of each partition, are provided. In computing systems implementing concurrent processing, applications may generate and destroy concurrently running processes with high frequency. Real-time security monitoring may cause the computing system to run monitoring processes collecting large volumes of data regarding system events occurring in context of various other processes, causing threads of processes of the security monitoring application to make frequent write access and read access to resources of those processes in memory. Indirect mapping of lock acquisition across locks provides scalable alleviation of lock contention and thread blocking that result from computational concurrency, while handling read and write requests which arise at unpredictable times from kernel-space monitoring processes, and which request unpredictable resources of monitored user-space processes.

The invention discloses a disease transmission simulation architecture based on a large language model, and the architecture deeply integrates a mathematical model, agent modeling and the large language model, and achieves the efficient and accurate simulation of disease transmission. The architecture comprises a bottom-layer technical support module and an upper-layer disease transmission dynamics logic module, the bottom-layer module adopts a distributed computing framework, an asycio asynchronous programming library and an aiohttp network communication technology, and high-concurrency computing of a million-level agent is ensured; the upper layer module supports user-defined disease characteristic parameters and propagation modes, and defines agent state transition rules based on a classical mathematical model. A policy issuing-execution-adjustment closed loop module is designed, and a scientific basis is provided for a policy maker. According to the method, the accuracy and practicability of epidemic situation propagation simulation are remarkably improved, and powerful technical support is provided for precise epidemic situation prevention and control and scientific decision making.

The invention relates to the technical field of data analysis, in particular to a data path tracking and visualization method and system. The data path tracking and visualization method comprises the following steps: querying a database, obtaining an original interaction record set, extracting a timestamp, a numerical value and an identity identification field of each record, and constructing a cache structure; setting a time range and a numerical threshold value in a self-defined manner, and screening target records meeting conditions; performing bidirectional link tracking on each screened target record, constructing a visual map structure, and binding the visual map structure to an original record; and carrying out deduplication processing, and inserting into a database in batches after deduplication. According to the data path tracking and visualization method and system, time and numerical value double constraints are combined, the path recognition accuracy is improved, input and output bidirectional tracing is supported, and complex service requirements can be met; meanwhile, the performance is improved in combination with asynchronous concurrent calculation, the expandability is high, and the method is suitable for various business scenes such as finance, audit and supervision.

The invention discloses a relational data restoration value evaluation and cleaning decision method and system, and aims to solve the problems that when a multi-field combination in relational data is restored, a CPU (Central Processing Unit) cannot complete utility evaluation within a limited time delay in a large-scale concurrent computing scene; the problem that cleaning resource allocation is low in efficiency due to the fact that the repair value cannot be estimated in real time and an executable priority ranking basis cannot be provided for computer hardware due to the fact that candidate distribution, model parameters and a verification set cannot be maintained in a limited memory for simultaneous residence operation is solved. Obtaining a request of a to-be-repaired field according to an abnormal result of the relational data and the problem label set, calculating and decomposing joint conditional distribution of the relational data corresponding to the request to obtain a candidate repair field set, solving an expected utility increment and a mean value of the relational data of the candidate repair field, obtaining a cleaning value of each relational data, and obtaining a to-be-repaired field; and traversing all the relational data in a descending order according to the calculated unit budget score, and obtaining a repair list in combination with a preset cost budget.

The invention discloses a meta-universe architecture system and a meta-universe data processing method, and relates to the field of meta-universe SIMD processor design, the system comprises a meta-universe system control terminal, the meta-universe system control terminal comprises a meta-universe system control processor and a special processor for meta-universe performance calculation, the meta-universe system control processor and the meta-universe performance calculation special processor are connected with each other, the meta-universe performance calculation special processor is a plurality of concurrent calculation cores for executing calculation of each layer of the meta-universe, and the meta-universe system control processor is used for receiving a multi-layer calculation task. The multi-level calculation task is sent to a meta-universe performance calculation special processor, and the meta-universe performance calculation special processor is used for decomposing and concurrently executing the multi-level calculation task through a performance concurrence calculation core and a processing mechanism of a meta-universe processor to obtain a calculation / data processing result required by a meta-universe user. And feeding back a calculation / data processing result to the user terminal of each level.

The invention relates to the technical field of facility intelligent management, and discloses a sports facility intelligent management system and method based on cloud computing. User behavior characteristics are extracted based on user interaction data, site facility state data, regional energy consumption data and environment data; the behavior prediction model is trained by using high-concurrency computing and a distributed architecture of cloud computing, and a site demand prediction result is generated through the behavior prediction model; and on the basis, a site recommendation list is further generated and adjusted in combination with user interaction data and regional energy consumption data, and a sports facility management decision is generated through an energy consumption optimization model, so that the generation and execution efficiency of the sports facility management decision is fully improved, the minimum distribution of energy consumption is realized while the user demand is met, and the user experience is improved. Therefore, the use efficiency and the energy-saving effect of sports facilities are improved.

The application discloses a kind of engineering budget processing methods based on budget mechanical demand.The method is applied to server, and the method comprises the following steps: constructing mechanical configuration algorithm model;Task execution unit real-time monitoring task scheduling queue, in response to the mechanical resource optimization configuration task initiated by client, extract mechanical resource optimization configuration task from task scheduling queue;Task execution unit extracts known parameter from mechanical resource optimization configuration task;Task execution unit solves mechanical configuration algorithm model according to known parameter, obtains the optimal solution of mechanical configuration algorithm model, the optimal solution is the construction machinery configuration scheme that can meet the progress plan requirement and minimizes the total cost of construction project, and the total cost of construction project corresponding to the optimal solution is calculated.The application realizes the optimal solution of construction machinery configuration automatic calculation without relying on artificial, improves the economy of construction machinery configuration result, and multi-task concurrent calculation can also maximize the use of system resources.

The application discloses a software integration degree monitoring system based on a modular architecture, and particularly relates to the field of software integration degree situation awareness, and comprises a data input module, a task distribution module, a path tracking module, a voting module and an evaluation module; the data input module inputs data contained in software as input variables through a data access pipeline, converts the data into standardized input blocks, simultaneously establishes a redundant data mapping, and makes the same data reused in multiple calculation paths; through the redundant data mapping and a task hash mechanism, the same data is reused in multiple calculation paths, and hash consistency verification is performed when the calculation paths are executed; compared with a traditional external checking mode, the present scheme utilizes redundant information generated in a calculation process to perform real-time error detection, thereby effectively reducing calculation deviation and data pollution risk in a high-concurrency calculation and distributed storage environment, and improving the stability of the overall system.

The application provides a kind of intelligent chip memory access control method and system based on high concurrency computing, applied to high-frequency random memory access scene, method includes: in response to the multiple memory access requests generated by multiple concurrent computing tasks, each memory access request contains the memory address to be accessed;Multiple memory access requests are temporarily stored in request buffer data structure;Detect whether multiple memory access requests temporarily stored in request buffer data structure point to the same memory page;In response to detecting that at least two memory access requests point to the same memory page, at least two memory access requests are merged to generate a merged memory access instruction, and the merged memory access instruction is sent to the memory controller to perform a one-time memory read operation.The application improves the probability of accessing the same page for high concurrency characteristics, reduces the demand for hardware resources, and significantly improves the memory bandwidth utilization and overall system performance.

The invention provides a special function calculation method and a special function unit calculation core of a multi-core processor, the method comprises that the multi-core processor comprises a plurality of cores and the special function unit calculation core in communication connection with the plurality of cores, and the special function unit calculation core comprises a calculation scheduling module and a plurality of calculation units. And the calculation scheduling module receives a task request of a source core in the plurality of cores, matches a target calculation unit in an idle state for the task request in the plurality of calculation units, and allocates the task request to the target calculation unit for calculation, so that the target calculation unit returns an obtained calculation result to the source core. According to the method, the task request is intelligently distributed through the calculation scheduling module, and the idle target calculation unit is dynamically matched in the multiple calculation units to process the task request, so that the calculation efficiency of the SFU in a high-concurrency calculation scene is effectively improved, the calculation delay is reduced, and the overall calculation performance of the multi-core processor is guaranteed.

The present disclosure belongs to the technical field of control systems. Provided are a correlation integration method and apparatus for a cascade system. The correlation integration method for a cascade system is used for obtaining a gradient between an optimization variable and an objective function of a cascade system, wherein the cascade system comprises a plurality of cascade modules sequentially connected in series. The method comprises: concurrently acquiring a static gain matrix of each cascade module; and on the basis of the static gain matrices of all the cascade modules, acquiring a gradient between an optimization variable and an objective function of a cascade system. In the method, a static gain matrix of each cascade module is calculated concurrently, thereby shortening the calculation time of a gradient between an optimization variable and an objective function, and effectively improving the optimization efficiency of a cascade system.