62 results about "Networking hardware" patented technology

The invention provides a system and a method for generating CAN (Controller Area Network) mailbox configuration and data routing configuration based on a communication matrix table. The method comprises the following steps of: reading a predefined finished automobile CAN communication network signal routing information table; analyzing the whole vehicle CAN communication network signal routing information table to obtain CAN communication configuration parameters of each CAN channel; the number of receiving mailboxes and the number of sending mailboxes of each CAN channel are distributed according to the maximum number of receiving mailboxes and the maximum number of sending mailboxes supported by the whole vehicle CAN communication network hardware for each CAN channel and CAN communication configuration parameters of each CAN channel; and when the number of the messages to be processed of a certain CAN channel exceeds the allocated mailbox number, classifying the messages, and multiplexing the same mailbox by the messages belonging to the same class. According to the method, CAN mailbox resources are used to the greatest extent, and the reliability of data routing is ensured; when the whole vehicle communication network signal routing table is frequently changed, the reliability of the routing function is ensured, and meanwhile, the development efficiency is improved.

The invention provides a large-scale constellation network semi-physical simulation test system and method, the system comprises a constellation simulator, a satellite-borne router, ground detection equipment and a protocol conversion network bridge, a Mininet simulation platform is deployed, and dynamic topology adjustment, an OSPF dynamic routing protocol and a multi-scene test function are supported; a user can observe network topology change, route updating time and key performance indexes in real time. A semi-physical simulation architecture is realized through unified IP subnet interconnection and the same routing protocol, a space-time synchronization engine ensures that virtual and real node routing tables take effect at an appointed moment, and in combination with a protocol consistency sandbox, hidden faults caused by inconsistent behaviors of hardware and software protocol stacks are solved, and the service life of the virtual and real node routing tables is prolonged. A plurality of real nodes are accessed in a large-scale network environment, and end-to-end data transmission test, security policy verification and protocol compatibility analysis are carried out; the invention provides an efficient, flexible and reliable solution for the design, verification and optimization of the satellite network, and has a wide application prospect.

One or more computing devices, systems, and / or methods for visual troubleshooting a network device setup. Images of the network device setup are provided to the system. A GenAI component processes the images to generate one or more device identifying features. The features are further processed to identify the device. The system utilizes hardware-specific information to prompt the GenAI component to answer troubleshooting-related questions concerning the device setup. The images may be pre-processed to include one or more visual guides to assist the GenAI component.

Systems and methods are provided for generating libraries for hardware realization of neural networks. The method includes obtaining a plurality of neural network topologies. Each neural network topology corresponds to a respective neural network. The method also includes transforming each neural network topology to a respective equivalent analog network of analog components. The method also includes generating a plurality of lithographic masks for fabricating a plurality of circuits. Each circuit implements a respective equivalent analog network of analog components.

A machine learning-based proxy model may generate measurements of quality of user experience in a vehicle-based communication network in the absence of direct feedback from the users regarding the user experience. Once generated, the proxy model may be applied to observed operational parameters of the on-board network to quantify the user experience for any user in any given instance. User experience measurements (e.g., trends identified therein) may be utilized to identify and implement adjustments to hardware, firmware, software, and / or service procedures associated with implementation of the on-board network. These adjustments may be implemented between transits of the vehicle, or in some cases, during transit of the vehicle to improve the user experience over the duration of use of the vehicle-based communication network.

Systems and methods herein are for at least one circuit that can determine that a network reference associated with a first network hardware is subject to a network performance degradation in a network, can cause suspension of traffic flow associated with the network reference, can save configuration for at least the network reference at a node associated with the first network hardware, and can cause the configuration to be deployed in a second network hardware so that the network reference that was previously in the first network hardware is provided from the second network hardware to resume the traffic flow.

System and method that enable a userspace compartmented IP Stack and network interfaces between a user device and communications devices associated with the user device, including mobile devices. The system and method include userspace bridging of data to / from external communications devices and applications onboard the user device. The system and method include userspace control of the network interfaces, routing between disparate IP networks, and routing of external IP network interfaces into user device interfaces. The system and method includes incorporating userspace device drivers required for external network hardware interface.

The application discloses a transconductance variable field effect transistor array suitable for a dendritic network hardware and application, and belongs to the technical field of semiconductor integrated circuits.The application realizes three-element multiplication of a storage variable and two input variables based on a single transconductance variable field effect transistor, and realizes mapping of a dendritic network core algorithm based on a complementary device array.Compared with a traditional neural network hardware which realizes nonlinear transformation by using a neuron activation circuit, the application realizes nonlinear transformation by using intrinsic nonlinearity of a device, effectively reduces design complexity, optimizes area and power consumption of a system peripheral circuit, and has important significance for design of a high-performance artificial intelligence computing system.

A black screen control system for nitric acid production, comprising network hardware nodes deployed in a local area network and control units running on the network hardware nodes, the control units are independently deployed at each operation station and adopt a multi-thread isolation architecture. The invention allocates time-consuming bottom layer data collection, processing and front-end UI interaction to independent threads for asynchronous execution, ensuring the ultimate smoothness and stability of the system under high concurrency; the state control and abnormal response are separated, the system enters the energy-saving black screen state under normal conditions to reduce power consumption, and when an abnormal signal occurs, the system is instantly awakened across threads and a safety alarm is triggered, realizing true safe unattended operation; the control units are independently deployed and run at each operation station, and are uniformly scheduled by a master control module, avoiding the whole line paralysis caused by single point failure or network fluctuation, and ensuring the absolute independence and reliability of a single device.

The invention provides an optimal phase shift parameter design method based on a switch multiplexing type multi-resonant switched capacitor, and the method comprises the steps: determining a control parameter combination according to the value interval of a preset phase shift ratio and a preset frequency ratio; determining resonance network hardware parameters according to the preset target equalization power and the preset resonance frequency; according to a preset soft switching constraint condition, verifying a parameter set composed of all control parameter combinations and resonance network hardware parameters; if the parameter set meets the preset soft switching constraint condition, taking the parameter set as an element of a feasible solution set; screening a parameter set with the minimum root mean square value of the resonance current according to the feasible solution set; and according to the parameter set with the minimum root-mean-square value of the resonance current, screening the parameter set with the minimum peak value of the resonance current as an optimal working point. According to the invention, through collaborative optimization of multiple performance indexes, on the premise of following the soft switching constraint, the switching loss and the operation loss of the switching tube are reduced, and the overall efficiency and the reliability of the equalizer are improved.

Systems and methods for Kubernetes-cluster-driven modification of network hardware configuration. A Kubernetes event is detected in received network traffic and the corresponding packet is examined to identify a pod of a Kubernetes cluster, then the packet is forwarded to the pod for processing. The pod determines a configuration modification of a network hardware element (e.g., a load balancing switch) based on the processing of the event. The pod sends the modification to a controller, which then generates a command (e.g., an API call or CLI command) and sends this command to the network hardware element to implement the desired modification. The configuration of the network element is then updated according to the command received from the controller.

The application discloses a 2-group signed tensor calculation circuit structure based on a 6-bit approximate full adder, relates to the field of neural network hardware acceleration, and comprises a 6-bit approximate full adder module, a signed 8*8 approximate multiplier circuit structure based on the 6-bit approximate full adder and the 2-group signed tensor calculation circuit structure based on the 6-bit approximate full adder. Since the neural network accelerator can sacrifice part of the accuracy of data in exchange for the optimization of delay, area and power consumption of a circuit structure, the 6-bit approximate full adder module ignores part of the carry, thereby reducing the circuit area and lowering the circuit power consumption, the signed 8*8 multiplier calculation process and the 2-group signed tensor calculation process are optimized by using the 6-bit approximate full adder module, approximate calculation is introduced at some positions, and in exchange for the improvement of area and power consumption of the circuit structure, part of the accuracy is lost.

A computing device includes a memory and a processing device. The memory enters an uninitialized state in response to power being applied to the memory. The processing device is coupled to the memory, and is configured to select a portion of the memory in the uninitialized state to seed a random number generator process. The processing device may alternatively select an uninitialized state of the network hardware that is coupled to the processing device to seed a random number generator process. In one embodiment, the computing device is a compute node in a multi-node processing system, and the memory is a High-Bandwidth Memory.

This application discloses a method, apparatus, device, storage medium, and program for adjusting the number of virtual machine connections, relating to the field of cloud platform technology. The method includes: determining the maximum preset number of tracking connections supported by the host machine in a cloud environment based on the network hardware information corresponding to the host machine and the expected load of establishing the virtual machine; sending status detection commands to each of the multiple target virtual machines when multiple target virtual machines already exist in the cloud environment; determining the actual tracking connection data corresponding to multiple connection tracking domains associated with the multiple target virtual machines based on the feedback results of the status detection commands; and adjusting the upper limit of the number of tracking connections supported by each target virtual machine based on the maximum preset number of tracking connections and the actual tracking connection data. This solves the problem in related technologies where the dynamic control of the number of virtual machine tracking connections in a cloud environment is insufficient, thus affecting the communication performance of the host machine.

Various aspects of the subject technology relate to systems, methods, and machine-readable media for cross-platform programmable network communication. The method includes receiving, via a conversational user interface (UI), a request from a user for a radio access network (RAN), the request including a description of a set of requirements for the RAN made in a conversation format. The method also includes generating a set of constraints for network hardware of the RAN based on the request. The method also includes providing, via the conversational UI, a description of the set of constraints to the user. The method also includes generating, based on an approval of the set of constraints from the user, a solution to a deployment for the network hardware of the RAN according to the set of constraints. The method also includes outputting a description of the solution including attributes of the solution.

The invention discloses a data processing method, device and equipment for a long short-term memory network and a medium, relates to the technical field of computers, and is applied to a preset long short-term memory network hardware implementation circuit, and the circuit comprises a plurality of static random access memories used for storing different types of network input data. The method comprises the steps of determining an initialization result based on a data processing task; based on a preset slicing strategy and a preset three-dimensional cyclic data processing strategy, splitting and loading various data in the initialization result, and triggering a storage operation of a corresponding static random access memory; after all storage operations are completed, data processing corresponding to four gates in the long and short-term memory network is completed in sequence based on the data in each static random access memory; and when it is judged that the cyclic data processing of the three dimensions is completed currently, obtaining a task result based on a corresponding data processing result. Through efficient and reasonable hardware circuit design, the hardware acceleration efficiency during data processing can be improved.

This invention relates to the field of network hardware security technology, specifically disclosing a device authentication and key derivation method and electronic device based on FPGA physical binding. The method includes: in response to device power-on, blocking the protected data path; calling the underlying hardware primitives of the FPGA to read the chip's unique physical identifier sequence; comparing the unique physical identifier sequence with preset authentication information to generate an authentication result; and generating a working key based on the unique physical identifier sequence and an external input key when authentication is successful. This invention restricts the protected data path at the initial power-on stage of the device and combines the chip-bound physical identifier sequence with the external input key to dynamically generate the working key, reducing reliance on static key storage mechanisms and mitigating the risk of configuration data being copied to unauthorized chips and continuing to operate.

The invention provides a partial product decoupling-based dynamic precision approximate multiplication implementation method and a multiplier, and the multiplier comprises a significance evaluation unit which splits a first operand and a second operand into high digits and low digits, and correspondingly generates a first amplitude and a second amplitude; different precision instructions are generated based on the amplitudes and the threshold values; the dynamic precision unit performs zero setting processing on the second low digit based on the precision instruction to form an approximate number, and calculates a partial product of the first high digit and the approximate number; the accurate calculation unit accurately calculates products of the other three parts; and the final addition unit sums the products of the four parts according to the weight. The calculation unit can be used for constructing or optimizing a multiplication calculation array in a neural network hardware accelerator, is particularly suitable for scenes with dense fixed-point number operation such as convolution, and performs efficient approximate operation on the input fixed-point number through a data driving mechanism by utilizing the common sparsity and amplitude difference characteristics of neural network data, so that the calculation efficiency is improved. And self-adaptive management of power consumption is realized.

The application provides a dynamic adjustment method, a management device and a system of a vehicle network. The dynamic adjustment method comprises: acquiring data transmission requests of each first function module and each second function module respectively; acquiring network state data of a vehicle-mounted network and external networks corresponding to each first function module respectively; determining network working modes of each first function module according to the network state data and the data transmission requests; determining network use permissions of each first function module and each second function module according to function attributes of each first function module and each second function module; and controlling each first function module to work in a corresponding network working mode, and controlling each first function module and each second function module to transmit data based on the vehicle-mounted network and / or each external network according to the corresponding network use permissions. According to the application, the defects of network hardware can be solved through software, and the real-time performance of the function of the vehicle network is improved.

The application provides in-memory index-based multilayer perceptron high-energy-efficiency in-memory calculation, and relates to the technical field of neural network hardware acceleration. The method comprises the following steps: extracting weight parameters of each layer in a multilayer perceptron model, generating a lookup table, and storing the lookup table in on-chip memory; according to the network layer type, dynamically loading a weight sub-table from the lookup table into a multiple address concurrent cache to form a local lookup table for single-cycle multi-address parallel access; according to the network layer type and a block strategy, batch reading of input feature vectors is performed and the input feature vectors are divided into multiple data blocks; using an activation function, feature values in each data block are transformed into index addresses, the local lookup table is accessed in parallel, and multiple weight data are obtained as lookup table results in a single cycle; through a hierarchical reduction circuit, the lookup table results are accumulated and fused to generate an activation output value; the activation output value is format-converted and written back to a feature data storage as input features of the next layer network or a model inference result. In this way, the requirements of low power consumption and high throughput are effectively balanced.

The invention relates to the technical field of defect detection, in particular to a hardware fitting internal defect detection method and system based on ultrasonic waves. The method comprises the following steps: acquiring ultrasonic waveform data of a to-be-analyzed hardware fitting; after an initial interface return band in the ultrasonic waveform data is extracted, a contact state feature set of the to-be-analyzed hardware fitting is determined; based on the contact state feature set and historical detection data, determining dynamic health reference waveform data matched with the features of the to-be-analyzed hardware fitting; performing data processing on the dynamic health reference waveform data and the ultrasonic waveform data to determine residual signal data; and based on the residual signal data and a preset defect judgment rule, identifying a detection result containing the type and severity of the real defect. The objective of the invention is to solve the problem in the prior art that the identification accuracy of the real defects of the hardware fitting is low, so that the false alarm rate of the defects of the hardware fitting is increased, and the judgment of the health condition of the power transmission network hardware fitting is affected.

The invention belongs to the technical field of neural network hardware accelerator and integrated circuit design. The invention provides a pulse convolutional neural network accelerator-oriented data storage and access method and system. According to the embodiment of the invention, a data storage and access mode matched with the systolic array calculation data flow is designed according to the binary property and the multi-time-step reasoning characteristic of the pulse data in the pulse convolutional neural network. A storage sequence which is mainly based on the input pulse data row direction, is grouped according to input channels and is continuously organized in the time step dimension is adopted in the off-chip memory, so that interaction with the off-chip memory can be carried out in a continuous access mode in the data loading and calculation result write-back process. The method comprises the following steps of: aggregating pulse data of a plurality of input channels in a storage word on an on-chip memory level; and meanwhile, in combination with an input cache design of a multi-BANK structure, multi-channel input data of adjacent rows can be read in parallel in a convolution window expansion process.

The application discloses a data online compression method and device of a neural network hardware accelerator. The method comprises converting a first activation value output by a neural network to obtain a first activation mask; dividing the first activation mask into at least two groups of activation sub-masks, and sequentially performing accumulation processing on each group of activation sub-masks in a preset order to obtain an activation position mask; calculating an activation selection mask based on the first activation mask, the activation position mask and a weight value output by the neural network; performing screening processing on the first activation value according to the activation selection mask to obtain a target activation value, and generating a second activation mask based on the target activation value. Through the online mask setting of the activation value and the offline compression of the weight value, the adaptability to different neural network compressions is high, the data moving efficiency can be improved, the power consumption is reduced, and the throughput is ensured.

Telecommunications network hardware unit comprising a processor connected to a computer-readable memory unit, the memory unit containing instructions which, when executed by the processor, implement a method for dynamically filtering telecommunications network agents, the method comprising: Retrieval, by the processor executing software code with respect to a virtual network function (VNF) and a physical network function (PNF) of the telecommunications network hardware unit, of peripheral data associated with a user device (UE) that is enabled with respect to a telecommunications network associated with the telecommunications network hardware unit; Assigning peripheral data to a predefined filter selection model by the processor; Selecting, by the processor executing the predefined filter selection model, and in response to the results of mapping, a filtering agent model from a pool of model resources; Selecting, by the processor in response to the execution of the filtering agent model, filtering agents from the VNF and in connection with noise and distortion reduction, which is related to the UE in relation to the telecommunications network; Retrieval, by the processor using a plurality of sensors of a global positioning system (GPS sensors), of environmental properties related to the telecommunications network; Identify, by the processor based on the environmental characteristics, a specified location of the UE that is enabled in relation to the telecommunications network; Sending the filtering agents to the UE via push, through the processor, with the filtering agents being stored within a temporary memory area of ​​an operating system of the UE; Generating, by the processor, network instructions related to the execution of the filtering agents; Execution, by the processor in response to the activation of the network instructions, of the filtering agents with respect to the UE, wherein the execution activates the filtering agents with respect to the UE for a specified time frame; and Reduce, by the processor in response to the results of execution, noise and distortion that occur during the operation of the UE in relation to the telecommunications network.

The present disclosure provides a simulation data transmission method adapted to transmit data of a three-dimensional simulation space over a network for rendering presentation on at least one graphical user interface, the method comprising: performing a first calculation comprising processing particles of the three-dimensional simulation space according to a smoothed particle hydrodynamic algorithm, generating first particle data; executing first coding according to the first particle data to generate second particle data; executing compression according to the second particle data to generate third particle data; and transmitting the third particle data through the network. According to the method, the bit mark, the space division, the data compression and the sliding window are combined, the technical scheme for real-time data transmission between the SPH particle system and the graphic rendering system is formed, the method does not depend on network hardware or protocols, and the system complexity and the deployment cost are reduced.

The machine learning-based proxy model may generate a measure of user quality of experience in a vehicle-based communication network in the absence of direct feedback from a user regarding user experience. Once generated, the proxy model may be applied to observed operational parameters of the onboard network to quantify the user experience of any user in any given instance. User experience measurements (e.g., trends in which are identified) may be used to identify and implement adjustments to hardware, firmware, software, and / or service processes associated with implementation of the onboard network. These adjustments may be implemented between traffic of vehicles, or in some cases during traffic of vehicles, to improve user experience during use of the vehicle-based communication network.

This application relates to a hardware acceleration method and apparatus for device fault diagnosis and monitoring based on FPGA. The method loads the parameters of a trained fault diagnosis model into a high-level synthesis tool to construct the basic hardware architecture; it optimizes data flow using row caching and window caching techniques, and achieves parallel data extraction through shift registers, significantly reducing memory accesses and pipeline startup intervals; it achieves computational parallelization through loop structure reconstruction and dependency optimization, eliminates data dependencies using loop swapping techniques, and designs a parallel computing unit array to improve processing efficiency; and it employs double buffering technology to achieve parallel execution of data transmission and computation, further improving system throughput. This method, through a four-step progressive optimization strategy, solves key technical problems of FPGA neural network accelerators, achieves significant performance improvements, provides a complete solution for the hardware deployment of fault diagnosis neural networks, and has significant engineering application value.

The application discloses a kind of circuit, method and neural network hardware system for realizing non-monotonic activation function, belong to the technical field of neural network hardware circuit design, the circuit includes: first inverter, second inverter and logic processing unit;The input end of first inverter and second inverter is connected, as analog accumulation signal input end;First inverter and second inverter are respectively set first logic flip threshold voltage and second logic flip threshold voltage of unequal value;The two input ends of logic processing unit are respectively connected to the output end of first inverter and the output end of second inverter.The application realizes non-monotonic activation function under extremely simple circuit architecture by the combination of double threshold decision and logic processing, without analog-digital conversion module, with the advantages of less transistor number, low power consumption, easy to integrate.

The invention relates to the technical field of digital management, and discloses a cloud-side collaborative digital platform based on regional production and operation integrated management and control, which comprises an infrastructure layer used for constructing a unified hardware resource pool and an automatic operation and maintenance base for supporting a cloud-side collaborative architecture, global unified management and automatic operation and maintenance of a bottom server, storage and network hardware equipment are realized; the platform layer is used for converging global data and integrating an application development module and an AI module to form a common tool chain and data sharing service covering low-code development, micro-service and model full-life-cycle management; and the application layer is used for calling the common technical capability and the data sharing service provided by the platform layer, and constructing and deploying a business application system serving for integrated management and control of regional production and operation. According to the invention, a single-point operation and maintenance mode is converted into a new integrated intelligent management and control mode with cross-domain cooperation, active early warning and accurate decision making, and the safety, operation efficiency and risk resistance of regional production and operation are improved.