47 results about "Extreme scale computing" patented technology

The application discloses a network model parameter protection method based on a multi-layer key, relates to the technical field of data protection, and calculates a low-entropy page upper limit at a single moment based on device parameters and a maximum layer scale of a model, and determines a page budget balance in combination with a remaining count of a previous round; in response to a hierarchical decryption request, if the balance is sufficient, a certain amount is deducted and a one-time session key is derived in combination with a device fingerprint; then, the assigned physical page is subjected to no-information-residue verification, after the verification, the parameter ciphertext is decrypted using the session key and written, and the current low-entropy page count is increased in real time; after the page is used up, forced overwriting and clearing verification are performed, after the verification, the page budget balance is backfilled and the low-entropy page count is reduced, and the one-time session key is destroyed; finally, a digital signature is generated and stored as evidence. The application establishes a budget gating and recycling mechanism, realizes dynamic limitation of the amount of plaintext in the runtime memory, and solves the exposure risk problem caused by full-decryption of model parameters of an edge device.

The invention discloses a server heat dissipation device and method, electronic equipment and a storage medium, and relates to the field of financial science and technology or other related technical fields, and the device comprises a cold plate which is internally provided with a micro-channel flow channel, and liquid cooling heat dissipation is carried out through the micro-channel flow channel; the fin assemblies are distributed in the edge area of the cold plate or the non-flow-channel covering area in the cold plate and used for conducting air cooling heat dissipation; the standby fan group is used for performing heat exchange by utilizing the surface of the fin assembly in an air cooling mode; and the intelligent control system comprises a flow sensor and a temperature sensor, and is used for monitoring the inlet flow of the cold plate and the temperature of the heating element to obtain a monitoring result, and performing intelligent heat dissipation adjustment based on the monitoring result in combination with at least one of the cold plate, the fin assembly and the standby fan group. According to the method and the device, the technical problems that server heat dissipation depends on single liquid cooling heat dissipation, the heat dissipation effect is poor and the emergency capacity is poor in a large-scale calculation scene in the prior art are solved.

The invention discloses a calculation method and application of an underwater random bubble curtain pressure value simulation model. The method comprises the following steps: firstly, establishing a bubble generation model based on random characteristics, and realizing random distribution of bubble positions and sizes through an MATLAB program and a randperm function; then, a fluid and bubble dynamics bidirectional coupling model is constructed, and pressure field association is established by keeping grid sizes consistent; generating a grid by adopting an SALE method, and combining the bubble model, the target plate model and the fluid model through open and import commands; and finally, carrying out numerical calculation by adopting an LS-DYNA solver, iteratively adjusting parameters based on an optimization algorithm, and outputting a pressure value simulation result. According to the method, full-process automatic modeling is realized, the efficiency is improved by more than 50%, million unit-level large-scale calculation is supported, the model precision is high, the universality is high, and the method can be widely applied to simulation analysis and device optimization of underwater noise reduction, antifouling and oxygenation scenes.

The utility model discloses a large -scale calculation interactive integrated machine of multimode data, including base, the top front end fixedly connected with display case of base, the rear end fixedly connected with a plurality of groups of wire conduit of display case, the rear end fixedly connected with terminal block of wire conduit, the top fixedly connected with terminal box of terminal block, be equipped with a plurality of groups of wire storage cavity that go up and down and pass through in terminal box, the top still fixedly connected with a plurality of groups of terminal post of terminal block, and terminal post all are arranged in the inside of wire storage cavity, and the outer wall top of terminal post all fixedly connected with spring wire, and the outer wall of spring wire all fixedly connected with protective sheath, and the top of spring wire all fixedly connected with joint, through setting the weight block, when it slides down, can drag spring wire to wire storage cavity inside, not only save staff's time, improve the use efficiency of whole integrated machine simultaneously, through setting the protective sheath, can protect spring wire to prevent spring wire breakage.

The invention discloses a hardware-friendly Transform column balance pruning model compression and efficient deployment method. A model compression algorithm, a lightweight parameter storage format, an operation data buffer, a systolic array operation block, a vector operation unit, a nonlinear operator unit, a data flow controller and a DMA unit are included. A model compression algorithm and an efficient deployment architecture are explored according to Transform network software and hardware collaborative reasoning requirements: in a software level, the scale calculation complexity of model parameter quantities is reduced through a fine-grained column balance structured pruning strategy, and parameters are stored in a single-instruction multi-data-stream format and the parameter storage efficiency is optimized through mask code storage; according to the hardware level, an edge computing-oriented Transform special accelerator architecture is designed, so that the architecture can support column balance structured pruning characteristics and a lightweight parameter storage scheme in an original manner. According to the Transform model compression and efficient deployment method, the parameter sparsity after structured pruning is fully utilized, so that the parameter storage pressure of a hardware architecture is reduced, the complex balance of an arithmetic unit is ensured, the operation efficiency of an accelerator is improved, and load balance and efficient reasoning during software and hardware collaborative optimization are realized; the method is widely applicable to efficient deployment scenes of Transform models for edge calculation.

The invention discloses an automatic selection method and system for monitoring indexes of a large-scale computer system, and the automatic selection method for out-of-band monitoring indexes comprises the following steps: aiming at a time sequence data sequence of M in-band monitoring indexes and N out-of-band monitoring indexes from the same computer node, quantifying the dynamic association strength between the out-of-band monitoring indexes and the in-band monitoring indexes, and selecting a plurality of out-of-band monitoring indexes with relatively high dynamic association strength with the in-band monitoring indexes to obtain a key out-of-band monitoring index set; and for the key out-of-band monitoring index set, carrying out space-time correlation analysis of in-band index anomaly and out-of-band hardware alarm by extracting in-band anomaly, and further optimizing the key out-of-band monitoring index to obtain a finally selected out-of-band index. The invention aims to select an out-of-band monitoring index with strong system operation indication in an automatic and unsupervised manner by associating the in-band operation state of the system, thereby reducing the problems of large cardinal number of an acquisition sensor, large data transmission pressure and large processing delay.

A manufacturing and test methodology for Zetta-scale computing engines is disclosed. The system employs a reticle-sized MEMS probe card to sequentially verify the electrical continuity of passive interconnects on a Wafer-Scale Silicon Circuit Board (WSSCB) prior to component assembly. By stepping across the wafer and testing one module at a time, the apparatus generates a defect map of the substrate's high-density wiring without the need for full-wafer probing or active powering. Following verification, Known Good Die (KGD) stacks—comprising Logic and HBM—are attached to the valid sites of the WSSCB via micro-bonding. This "step-and- repeat" verification strategy ensures high manufacturing yield for the complex all-silicon domain assembly.

The application provides a runtime suspension detection and positioning method and device for parallel programs, and applies to the technical field of large-scale computing. The method comprises the following steps: constructing a program counter value queue of each target execution unit at each sampling time; recording a stagnation program counter value of a target execution unit in a stagnation state; clustering the stagnation program counter value of the target execution unit to determine a suspended target execution unit; based on a machine instruction corresponding to the stagnation program counter value of the suspended target execution unit, obtaining source file and line number information corresponding to the stagnation program counter value through debugging information; and when the machine code corresponding to the stagnation program counter value of the suspended target execution unit is a call instruction, obtaining function call information. In this way, non-intrusive and low-load monitoring of parallel programs is realized, normal synchronization waiting and abnormal suspension processes are effectively distinguished, and problems are positioned to source code lines.

This invention relates to the field of electric vehicle charging and discharging technology, and more particularly to a method, system, and computer-readable storage medium for extrapolating a vehicle-to-grid (V2G) interaction supply-demand coordination model. The method first collects data and determines model input parameters based on local policies, electricity pricing mechanisms, and technological development trends. Then, it constructs a user utility function, eliminates differences in indicator dimensions through normalization, and calculates the probability of user selection for each charging mode, thus constructing a demand-side mode selection model. Based on user selection probabilities and facility operation indicators, it establishes a model for calculating facility expansion scale, and uses a dual-objective optimization framework to balance maximizing user satisfaction with minimizing investment costs to determine the annual number of new facilities. The number of new facilities is fed back to the user model to update facility accessibility parameters until the supply and demand status meets preset convergence conditions. This invention significantly improves the scientific nature and prediction accuracy of V2G interaction planning through dynamic closed-loop modeling of supply and demand coordination and a dual-objective optimization strategy.

This application discloses an anomaly localization method and a cloud management platform based on a cloud management platform, which can improve the efficiency of anomaly localization to a certain extent. The method includes: the cloud management platform can create a large-scale computing node cluster for tenants, which can contain multiple computing nodes. If these multiple computing nodes fail to execute a job task, the cloud management platform can first obtain the logs generated by these multiple computing nodes during the execution of the job task, and generate log space characteristics and log time characteristics of these multiple computing nodes based on the logs. Then, the cloud management platform can evaluate these multiple computing nodes based on these log space characteristics and these log time characteristics to obtain evaluation values ​​for these multiple computing nodes. Subsequently, the cloud management platform can identify computing nodes whose evaluation values ​​are greater than or equal to a preset threshold as computing nodes that have experienced an anomaly.

The invention discloses a storage particle, a storage controller, a storage chip, a storage device and equipment, relates to the technical field of computers, and is used for reducing the requirement of a model parameter matrix for memory capacity in a large-scale calculation process and solving the problem that the large-scale calculation performance is influenced by the storage medium interface bus rate. The storage device comprises a storage controller used for sending a plurality of model parameter sub-matrixes included in a model parameter matrix to a plurality of storage particles so as to store the plurality of model parameter sub-matrixes in the plurality of storage particles; the plurality of storage particles are used for concurrently calculating intermediate calculation results of the input data matrix and at least one model parameter sub-matrix stored in the storage particles, and sending the intermediate calculation results obtained by the storage particles to the storage controller; and the storage controller is also used for calculating the calculation results of the input data matrix and the model parameter matrix according to the plurality of intermediate calculation results correspondingly obtained by the plurality of storage particles.

The invention discloses a scheduling method for cloud edge cooperative computing tasks of a smart steel mill. The scheduling method comprises the steps of collecting computing tasks of a whole production link of the smart steel mill; obtaining the total number of CPU cores and the total number of memories in the cloud side server cluster; a K-Means algorithm is adopted to divide calculation tasks into cloud-assisted edge calculation and edge-assisted cloud calculation; constructing a constraint model, and constraining the communication time delay and the communication cost of the calculation task by taking minimization of the maximum communication time delay and the total communication cost of the task as a target; establishing two types of computing task agents, training the two types of computing task agents, and optimizing communication time delay and communication cost; and determining a distribution scheme of the calculation tasks according to the trained two types of calculation task intelligent agents. By adopting the method, dynamic scheduling optimization of large-scale computing tasks in the cloud edge collaborative manufacturing system is realized, the computing efficiency of the cloud edge server and the flexibility of the manufacturing process are effectively improved, and the communication burden of the cloud edge end is reduced.

The invention provides a method and device for creating an intelligent decision data set based on log data, and relates to the technical field of cloud computing, cluster management, big data processing and artificial intelligence, and the method comprises the steps: collecting original log data from a server and a cluster, and achieving the collection and preprocessing of the log data; key information is extracted from the preprocessed log data, and log event analysis and structured processing are realized based on a self-attention mechanism; high-quality features are extracted and constructed from structured log data, and feature engineering and data enhancement are achieved; and combining the enhanced feature set with a predefined decision target to generate an intelligent decision data set for model training and evaluation, thereby realizing construction and evaluation of the intelligent decision data set. According to the method, high-quality data support is provided for the intelligent decision model, the system resource demand, the task execution time and the failure risk can be effectively predicted, the resource allocation efficiency is greatly optimized, and a solid foundation is provided for large-scale computing system resource management and scheduling.

The invention discloses a hardware fault repairing method and device, a medium and a program product, and relates to the technical field of automatic operation and maintenance of large-scale computing clusters. The hardware fault repairing method comprises the steps of obtaining a preliminary repairing result of a target state machine; when the preliminary restoration result is that restoration cannot be carried out, acquiring preliminary restoration alarm data; and collecting alarm context information according to the preliminary repair alarm data, and generating repair plan data according to the alarm context information and a large language model. The technical scheme provided by the embodiment of the invention can be suitable for operation and maintenance requirements of complex hardware faults of a large-scale cluster, and the dependence on manpower is broken.

Embodiments of the present disclosure disclose a data processing method based on reinforcement learning, an electronic device and a readable medium. The method is applied to an execution node, a policy node and a training node included in a reinforcement learning system, and a specific implementation of the method comprises: performing environment simulation by the execution node to obtain an observation result; performing policy updating by the training node according to a training sample and a model to be trained, wherein the training sample is transmitted from the execution node to the training node through a sample transmission stream; the policy node performs policy inference according to the observation result and updated policy information, wherein in a non-in-line inference mode, the execution node and the policy node perform data transmission on the observation result and an action to be executed through an inference transmission stream; and the execution node executes the action to be executed to generate an updated observation result. The implementation realizes effective scheduling of large-scale computer hardware resources, meets the training requirements for complex reinforcement learning tasks, and improves the training speed.

The invention relates to the technical field of electric energy substitution optimization, and discloses an electric energy substitution optimization method and system. The method comprises the following steps: processing an earlier-stage multi-view through a first feature selection model to obtain an initial low-dimensional feature and an incidence matrix so as to initialize feature selection model parameters corresponding to a later-stage multi-view, and constructing a third feature selection model in combination with a multi-constraint condition; target low-dimensional features are obtained through a minimum total loss solving model, and finally an electric energy substitution optimization strategy is determined by combining with a historical load data input prediction model. According to the method, the data dimension is remarkably reduced through feature selection and dimension reduction processing, and computing resource consumption is reduced; the calculation efficiency is improved through staged model parameter initialization and constraint condition optimization. The technical problems that in the prior art, when large-scale power grid data is processed, calculation resource consumption is too large, calculation efficiency is low, and particularly in the face of complex power grid data and large-scale calculation, the real-time requirement is difficult to meet in the prior art are solved.

The application discloses a vehicle computing task scheduling method based on vehicle-edge-cloud cooperation and belongs to the technical field of intelligent transportation systems and Internet of Vehicles. The method is aimed at the problems of limited computing resources of intelligent networked vehicles, high task delay and large energy consumption. An improved greedy algorithm is used to optimize the task execution order, and local greedy selection is performed according to the data volume, computation volume and maximum tolerable delay of the task to output an optimal platform allocation label. Then, a feedforward neural network is trained using the labels. The trained neural network can quickly predict an optimal platform vehicle-edge-cloud cooperative allocation scheme according to the input task characteristics in a real-time scenario. Through the fusion of the algorithm and the neural network, the application realizes efficient online real-time scheduling, significantly improves the task scheduling efficiency and system reliability, has excellent real-time performance and scalability, effectively reduces the task execution delay and energy consumption, and is suitable for efficient scheduling and resource allocation of large-scale computing tasks in the Internet of Vehicles.

The embodiment of the invention provides a hierarchical configuration management and automatic synchronization method and device for efficient molecular dynamics calculation, and the method comprises the steps: obtaining configuration information from a superior node of a current node when the current node is started for each node in a molecular dynamics calculation cluster three-level architecture, establishing a version number, a change log and an event subscription relationship in the configuration information; when the current node configuration information is changed, generating a change event and a version number, recording a change log according to the change event, and spreading the change event through an event bus; regularly determining whether the configuration version numbers of the current node and the superior node are consistent or not; if not, sending a synchronization request to the superior node; according to the configuration version number of the current node, incremental data is extracted from the superior node, so that the current node applies incremental configuration to change the configuration information of the current node, and the three-level layered architecture can realize configuration management of a large-scale computing cluster and support differentiated configuration requirements.

This invention provides a model deployment method and system for resource-constrained environments. By jointly introducing a tensor library structure and a matrix pairing mechanism, the parameter size is significantly reduced while maintaining the model's expressive power. The constructed compressed model only reconstructs and replaces the fully connected layers on the original network structure, without changing the overall forward propagation logic. The training process is simple and stable, and requires no additional large-scale computing resources; model training can be completed on ordinary GPUs or high-performance personal computers. The construction of the tensor library and the matrix pairing operation have low computational overhead, involving only standard tensor shrinkage operations, and the overall inference speed remains basically unchanged. The method of this invention maintains high prediction accuracy under different network sizes and compression ratios, and has good versatility and scalability, providing an efficient compression solution for deploying traditional neural networks and large models in resource-constrained environments.

This invention relates to the field of high-precision industrial printing technology and discloses a multi-printerhead synchronous printing system based on a distributed architecture, including a workstation, an FPGA processing board, and print head modules. This system separates non-real-time large-scale computing from real-time high-precision control through a distributed architecture. When increasing the number of print heads, only additional computing tasks need to be added to the workstation, and logic units and I / O interfaces are expanded on the FPGA. The workstation parses the image data of each print job, completes dynamic load segmentation, establishes a segmentation index table, and then packages the data and embeds timestamps, ensuring that the FPGA processing board has sufficient time to complete decompression and caching, and also ensuring precise start of the printing operation at the specified time point. The distributed parsing has strong scalability, and the FPGA processing board constructs a global synchronization mechanism to ensure precise alignment of data from different channels in the time dimension. Waveform queries are performed on data packets to generate personalized drive waveforms adapted to each piezoelectric print head, resulting in excellent multi-printer collaboration.

The invention discloses a configuration optimization method in multi-agent path planning, which is used for an algorithm in path finding of agent robots such as a logistics robot and an unmanned aerial vehicle. On the basis of an original algorithm, a multi-node generation strategy is adopted to avoid falling into a local optimal solution, and the overall problem solving capacity is improved. Comprising the steps of creating an initial node Ninit, obtaining a node to be processed, comparing node configuration with target configuration, pruning invalid nodes, generating new configuration, processing the configuration, and finally outputting a found feasible solution: for a problem instance capable of being solved, returning a solution within finite time; and if the problem instance has no solution, reporting that no feasible solution exists. Aiming at the problems of poor initial solution quality, low convergence speed and the like of the existing graph path search algorithm, the method is more suitable for industrial automation application scenes with strict requirements on real-time performance and large-scale calculation.

This invention provides an accelerated method, system, product, medium, and device for LDU sparse matrix computation, relating to the field of high-performance numerical computation technology. It is applied to the Sunway many-core processor, which includes a main core and multiple slave cores. The method includes: discretizing the simulation region of the target computation example into an unstructured mesh to generate an lduMatrix format sparse matrix; dividing the cells and surfaces of the unstructured mesh into blocks and allocating them to each slave core, generating a communication graph containing a computation mapping table and a communication mapping table; based on the communication graph, each slave core performs two-stage parallel computation in parallel, and finally writes the complete computation results back to main memory in batches. This invention effectively overcomes the computational bottleneck of unstructured meshes on the Sunway many-core processor, significantly reduces the drag on overall performance from main memory access latency, greatly leverages the computing power of the Sunway many-core processor, and can meet the stability requirements of industrial-grade CFD simulation for long-term, large-scale computation.

The present application relates to the technical field of optical imaging, and particularly relates to a kind of aerodynamic optical aberration calculation method, first for the aircraft model of optical imaging system is equipped with external flow field, set its speed variable condition, multiple flight height sampling points and corresponding state boundary condition;According to the Reynolds number and viscous interference parameter under each state boundary condition, flight height sampling point is divided into two kinds of calculation aberration and predicted aberration;For predicted aberration sampling point, first randomly select one as reference point and calculate its aerodynamic optical aberration, then the aerodynamic optical aberration of reference point is used to predict the aerodynamic optical aberration of other predicted sampling points;For the calculation aberration sampling point, directly through the density distribution of its external flow field, conversion refractive index distribution is solved to obtain the corresponding aerodynamic optical aberration.The present application expands to different flight heights through once aerodynamic optical transmission aberration calculation, significantly reduces the demand for large-scale computing resources and the time required for evaluation.

The invention discloses a data batch processing scheduling method and device, computer equipment and a storage medium, and the method comprises the steps: responding to a received data batch processing task request, and the data batch processing task request comprises a business dimension field; based on the business dimension field, performing scale evaluation on the target data, splitting the target data into at least one task subset according to a preset splitting strategy, and classifying the task subsets; adjusting the type of the task subset based on the scale of the task subset, the calculation complexity and the data distribution characteristics, and generating a corresponding type identifier; according to the type identifier, executing a corresponding task processing strategy, processing the task subset, obtaining a processing result and storing the processing result; and in response to completion of processing of each task subset, obtaining a processing result corresponding to each task subset, performing aggregation calculation, and outputting a final result, so that the batch processing efficiency and the resource utilization rate can be improved.

An adaptive signaling architecture compensates for electrical and thermal variation among multiple power domains of a wafer- or panel-scale computing substrate. Each domain receives power from an independent power-supply board, and inter-domain signal channels include adaptive control circuitry that dynamically adjusts voltage swing, bias, pre-emphasis, and termination in response to local voltage and temperature sensors. The adaptive network maintains uniform impedance and timing across domain boundaries without retimers, reducing interconnect power and improving reliability. The system sustains per-channel data rates above 32 Gb / s and rack-scale throughput around 1 zettaFLOPS (FP4 sparse AI inference) within practical power limits.

The application discloses a kind of complex special-shaped building segmentation methods based on voxel model, step 1, the point cloud data of special-shaped building is obtained;Step 2, according to the definition of voxel unit length of special-shaped building point cloud model complexity;Step 3, construct model library, based on point cloud model through matrix calculation to voxelization reconstruction is carried out;Step 4, based on voxel model segmentation entity building, realize the building scale calculation of complex special-shaped building;Step 5, the output of calculation result.Through point cloud model, the voxel model of building is constructed, the voxel model is segmented, whereby the building scale of special-shaped building is measured and calculated, the measuring and calculating efficiency is improved, the measuring and calculating cost is reduced, and it is suitable for large-area popularization.

The application discloses a non-uniform and conformal grid-oriented electromagnetic particle adaptive simulation method and system, and relates to the technical field of electromagnetic particle simulation. The method comprises the following steps: performing non-uniform right-angle grid division based on the average free path of charged particles, and identifying conformal grid units; adopting macro particles for initialization; performing electromagnetic field propulsion, field interpolation, particle propulsion, charge and current backfilling and collision simulation in each time step; and realizing accelerated calculation through a CPU and GPU hybrid parallel architecture. Through the application, the numerical simulation precision can be improved, the adaptability of complex geometry can be enhanced, the stability of electromagnetic field and particle coupling can be ensured, the information interaction between the field and the particles can be improved, and the large-scale calculation efficiency can be improved.

The present disclosure provides techniques and solutions for reducing coupling between an application management control plane and a data plane in a very large scale computing environment. A super-large scale computing environment includes a first data plane super-large scale cluster having a plurality of physical or virtual servers hosting a first software application and a first communication controller. The very large scale computing environment also includes an application management control plane cluster having a plurality of physical or virtual servers hosting a second software application supporting the first software application, and a second communication controller. The second communication controller routes operations from the second software application to the first communication controller, which then routes them to the first software application. In some cases, management operations for the data plane are performed only when originating from the application management control plane. In further cases, the management operation is sent only under requests of the data plane.

This invention discloses an intelligent cooperative spatial modulation method and system based on GNN-MARL, mainly addressing the antenna resource competition and energy efficiency issues in multi-user spatial modulation systems. The method includes: first, establishing a multi-user interference graph model to abstract the interference relationships between users into a graph structure; second, using a graph attention network for feature extraction and message passing to update the environmental awareness embedding; then, employing multi-agent reinforcement learning, where each user agent independently selects distributed antennas based on local embedding; subsequently, combining Polar code channel coding and CCDM probabilistic shaping techniques to spatially modulate and map information bits for transmission; and finally, performing joint detection and decoding of the signal at the receiving end. This invention not only effectively reduces the large-scale computational complexity of the system but also significantly reduces the antenna collision rate and average transmit power, greatly improving system throughput. For example, in an 8-user scenario, it controls the collision rate to 19.6% and achieves a power saving of 32.8%, providing an efficient and practical innovative solution for intelligent network distributed resource management.

The present disclosure provides techniques and solutions for reducing coupling between an application management control plane and a data plane in a hyperscale computing environment. The hyperscale computing environment includes a first data plane hyperscale cluster with a plurality of physical or virtual servers hosting a first software application, and a first communication controller. The hyperscale computing environment also includes an application management control plane cluster with a plurality of physical or virtual servers hosting a second software application that supports the first software application, and a second communication controller. The second communication controller routes operations from the second software application to the first communication controller, which then routes them to the first software application. In some cases, management operations for the data plane are only executed when originating from the application management control plane. In further cases, management operations are only sent at the request of the data plane.