55 results about "Runtime system" patented technology

The invention discloses a lightweight cross-platform PLC (Programmable Logic Controller) multi-language code generation method, which relates to the field of industrial process control, and firstly provides a unified intermediate data model, and on the basis, a module for converting each PLC language into the unified intermediate data model is realized; on the basis of a standardized JSON data structure, a modularized unified intermediate language conversion chain is further supported, an analyzer and a generator communicate through a JSON interface, a cross-platform conversion path is supported, and unified intermediate language conversion from PLC to JSON to XML to ST to C is completed; the method comprises the following steps of: integrating an open source compiler and runtime, constructing an open and pluggable automatic execution framework, compiling a generated C code into a platform executable file by using an adaptive GCC tool chain aiming at different target platforms, and deploying a program generated by compiling to a soft PLC runtime system supporting multiple platforms. According to the method, a JSON intermediate data structure which is light in weight, unrelated to languages and uniform in structure and a matched conversion framework of the JSON intermediate data structure are constructed.

The invention discloses a unified programming model compiling and runtime system oriented to a Shenwei supercomputing platform, which comprises a compiling module used for compiling SYCL programming model codes and detecting kernel functions to generate equipment end intermediate representation; processing the equipment end intermediate representation to obtain a kernel function dynamic link library and an executable code of a host end; and the public runtime module is used for calling the kernel function dynamic link library when the executable code at the host side runs. Based on a compiling module and a public runtime module, a vertical optimization system from a compiler to runtime is constructed, efficient operation of SYCL in a domestic supercomputing system is realized, a semantic gap between a domestic many-core architecture and a universal programming model is overcome, and the development of the system is facilitated. The SYCL programming model is processed by a host end and an equipment end respectively after being extracted by a kernel during compiling, and is dynamically called by a common runtime module, so that the cooperative computing capability of the host end and the equipment end is fully exerted.

The present disclosure provides a method for managing allocation of compute resources in a distributed database. Each request on every task is allocated a certain number of tokens when it arrives, such as by using a token bucket algorithm. This initial allocation may be the per-request maximum usage permitted in the system. This request is made to a central quota server in the region to coordinate allocations between tasks. As the task runs, the system evaluates how much of the limit the request is using, and adjusts its limit up or down to keep the quota allocation tight on the actual usage of the request. If there aren't enough tokens to keep the request active, the request can be evicted.

The embodiment of the invention provides a register addressing method, an instruction processing device, a target processor, a system on chip and computing equipment, and the register addressing method is applied to the target processor and comprises the steps that a register addressing instruction for a target register in the target processor is obtained, the register bit width and the operation data bit width of the target processor are different, and the register addressing instruction comprises an operation bit of the register bit width; determining a target operation bit from the operation bits of the register bit width based on the operation data bit width; and based on the target operation bit, executing an addressing operation on the target register. According to the method, the addressing of relaxation expansion is realized, and the capacity expansion of the operation bit is realized without adding additional software instructions, so that the number of software instructions is reduced, the addressing speed and efficiency of the register are improved, the complexity of a compiler and a runtime system is simplified, the data processing efficiency and the overall performance of the target processor are improved, and the addressing efficiency is improved.

The invention discloses a machine learning platform and method based on workflow, electronic equipment and a storage medium, and the platform comprises a management scheduling system, a workflow engine system and an operator runtime system. A user is allowed to construct a visual business process in a dragging and connecting mode, and each component corresponds to a specific AI algorithm processing unit written by Python; the platform supports the realization of multi-node business logic by using an operator concept by referring to a BPMN specification, automatically analyzes a flow chart designed by a user into a directed acyclic graph, converts the directed acyclic graph into an API calling sequence for a Python runtime system, disassembles a complex business system into a multi-branch and multi-step configurable process, and executes local logic by each node to realize multi-node business logic. The whole process is visually arranged, the development threshold is remarkably reduced, and the execution efficiency and the system flexibility are improved.

A human-machine interface (HMI) development and runtime system leverages generative artificial intelligence (AI) to create HMI dashboards or graphical interfaces without the need for manual coding based on natural language prompts submitted by the user, which specify the information the user wishes to see. In one or more embodiments, the system can support industry-specific prompt engineering services that assist a user in generating a customized HMI dashboard that satisfies specified criteria using natural language prompts that describe functional and visual requirements of the dashboard. The system can make use of a generative AI model and associated neural networks to generate suitable dashboards in accordance with functional requirements provided to the system as intuitive natural language inputs (e.g., spoken or written natural language text).

The invention relates to the technical field of compatibility testing, and discloses a multi-platform compatibility automatic testing method, which comprises the following steps of: establishing a three-dimensional heterogeneous association graph of three types of nodes, namely user behavior sequence data, system resource interaction data during application operation and test equipment bottom layer platform feature data through a time sequence alignment algorithm; an enhanced graph neural network fusing a dynamic attention mechanism and time sequence graph convolution accurately identifies a high-risk behavior sequence; constructing an operation atom-parameter matrix-expected result ternary structure case library based on the risk sequence, and deploying an edge-cloud collaborative shadow test architecture to realize millisecond-level preliminary verification and depth test; a backbone network + platform adaptive branch modular design is adopted, and new platform risk identification is adaptively adjusted in combination with a memory replay mechanism, so that the test efficiency and accuracy can be improved, the problems of resource waste and misjudgment in multi-platform compatibility test are effectively solved, and accurate positioning and efficient repair of compatibility problems are realized.

The invention belongs to the field of GPU computing performance testing, and provides a GPU flow scheduling performance measurement method and device.The method comprises the steps that in a pre-established basic execution environment, N null-kernel functions are submitted to a reference flow in sequence, and the reference flow executes the received null-kernel functions and measures the reference execution duration; alternately submitting the N null kernel functions to the first test flow and the second test flow, introducing a cross-flow synchronization event, and measuring the total execution duration; and according to the reference execution duration and the total execution duration, calculating to obtain synchronization time consumption between the streams. According to the method, the influence of a tracking tool and the like on the performance of the measured object is eliminated, the method has obvious advantages in accuracy and stability, high-precision and low-interference GPU scheduling performance measurement is realized, and a quantitative basis is provided for performance optimization of a system during GPU driving and operation.

The invention discloses a component arrangement method, device and equipment of a runtime system and a storage medium, relates to the technical field of component arrangement, and discloses the component arrangement method of the runtime system, which comprises the following steps: acquiring configuration file information of a to-be-arranged component in the runtime system, the configuration file information comprises a dependency interface and a provision interface of the to-be-arranged component; and arranging the to-be-arranged component according to the dependency interface and the provision interface to obtain a component dependency relationship network so as to realize the function of the runtime system based on the component dependency relationship network. According to the invention, the system expansibility during operation is improved.

A method includes receiving multiple security framework requirements, mapping the security framework requirements to a predicate, mapping the predicate to a system-specific implementation, evaluating, using a runtime system, the target system by analyzing a multitude of build files using the system-specific implementation, and presenting a report indicating whether the security framework requirements are satisfied.

The invention discloses a multi-operator coarse-grained parallel scheduling method based on a heterogeneous data stream architecture, which belongs to the field of runtime systems, and comprises the following steps of: dividing a computing load of an operator with in-degree of 0 in a computing graph to be executed into a plurality of sub-tasks Task and packaging the sub-tasks into independent Task objects; constructing four priority queues which are respectively used for storing Task objects which have different preferences on four functional parts in the processing unit, putting index Task objects of operators into the corresponding priority queues according to the preferences of the operators on the functional parts and the number of instructions of the operators, and sorting the index Task objects, k (K < = 4) Tasks with complementary resource requirements are selected from the four priority queues to form a new composite operator, and the new composite operator is executed by the processing unit array; according to the method, the new operator is generated by identifying the preference of the operator function part and dynamically combining the Task, so that the utilization rate of the internal function part of the PE can be fully improved, and the operation efficiency of the whole data flow program is accelerated.

The disclosed embodiments provide a memory allocation method, device, and storage medium based on a runtime system. The method determines the first data size and the first pointer layout type of the object to be allocated; based on the first data size and the first pointer layout type, determines a target memory block for storing the object to be allocated from pre-configured memory blocks of different types, where any type of memory block is used to store objects with the same data size and pointer layout type; allocates available memory space for the object to be allocated in the target memory block, and stores the object to be allocated in the available memory space. By configuring different types of memory blocks to store objects of different data sizes and different pointer layout types, there is no need to use an additional object header or a global bitmap to store the pointer layout information of the object, thus reducing additional memory overhead and CPU overhead of encoding and decoding, improving memory allocation efficiency, and improving the efficiency of subsequent garbage collection processes that require obtaining pointers in objects.

Systems, apparatuses, and methods for updating and optimizing task scheduling policies are disclosed. A new policy is obtained and updated at runtime by a client based on a server analyzing a wide spectrum of telemetry data on a relatively long time scale. Instead of only looking at the telemetry data from the client's execution of tasks for the previous frame, the server analyzes the execution times of tasks for multiple previous frames so as to determine a more optimal policy for subsequent frames. This mechanism enables making a more informed task scheduling policy decision as well as customizing the policy per application, game, and user without requiring a driver update. Also, this mechanism facilitates improved load balancing across the various processing engines, each of which has their own task queues. The improved load balancing is achieved by analyzing the telemetry data including resource utilization statistics for the different processing engines.

Techniques are provided for resource scheduling of multiple application types in a heterogeneous endpoint device. One method comprises monitoring an execution of applications on a heterogeneous endpoint device, wherein the applications are of different application types and the heterogeneous endpoint device comprises multiple application runtime systems each supporting a different one of the different application types; and initiating an adjustment of an allocation of resources, provided by the heterogeneous endpoint device, assigned to at least one of the applications based on: (i) a current resource utilization of one or more of the applications, (ii) a priority assigned to one or more of the applications and / or (iii) a resource threshold value assigned to one or more of the applications. A global scheduling engine may: (i) monitor multiple endpoint devices and / or (ii) deploy one or more applications to at least one endpoint device based on resource consumption information from the endpoint devices.

The present application belongs to the technical field of embedded processor firmware, and particularly relates to an efficient firmware running system for a neural network processor, aiming to solve the problems of slow firmware deployment speed, large occupied space and low processing efficiency of files generated after compilation by a neural network compiler of the existing neural network processor. The system comprises: a file preprocessing unit configured to obtain files generated after compilation by a neural network compiler and preprocess the files to obtain preprocessed files; an embedded hardware platform configured to store the preprocessed files to a set memory address; and an efficient firmware configured to read and parse the preprocessed files, obtain operator stream files corresponding to a neural network model, and sequentially execute each operator through a pre-built operator execution efficient optimization method until a complete calculation of the neural network model is completed. The present application improves the processing efficiency of files generated after compilation by a neural network compiler, improves the firmware deployment speed, and reduces the occupied space.

The invention provides a PLC program execution system, and relates to the technical field of programmable logic controllers, the PLC program execution system comprises a runtime system and an execution mechanism, the runtime system is used for storing data and carrying out network interaction, and a unified interaction interface is arranged in the runtime system; the execution mechanism is connected to the runtime system through the interactive interface and used for accessing a memory and a network through the interactive interface and selecting one implementation mode to execute a PLC program, and the implementation modes of the execution mechanism comprise an interpretation type actuator, a compilation type actuator and a JIT type actuator. Through the execution mechanism in the PLC program execution system, one implementation mode can be selected from multiple implementation modes of the execution mechanism to execute the PLC program, the implementation mode is flexible and selectable, the flexibility of the PLC program execution system is improved, and the problems that in the related technology, a user cannot select the execution mode of the PLC program, and flexibility is lacked are solved.

This disclosure provides a virtual machine runtime system, method, processor, and electronic device, relating to the field of virtual machine technology. The system includes: a virtual machine monitor, which runs at a first exception level and is used to create virtual machines using a virtualization environment provided by a microkernel, and to handle interrupt events associated with the virtual machines according to notifications from the microkernel; a virtual machine, which runs at a second exception level; and a microkernel, which runs at a third exception level and is used to provide the necessary virtualization environment for the virtual machine monitor to create the virtual machines, and to intercept interrupt events associated with the virtual machines and notify the virtual machine monitor to handle the interrupt events; the privilege levels of the first exception level, the second exception level, and the third exception level increase sequentially. This disclosure aims to solve the problems of centralized privileges and high resource consumption in existing virtual machine runtime systems.

The invention discloses a method and an electronic device for managing an executable program. The method comprises the steps that in response to first mounting of a system partition, the execution permission of a trusted partition is set to be executable, the read-write permission of the trusted partition is set to be read-only, so that an executable partition is obtained, and a trusted execution program and a trusted dynamic link library on which the trusted execution program depends are written in the trusted partition; setting execution permissions of other partitions except the trusted partition to be non-executable so as to obtain non-executable partitions; and forbidding to modify the read-write permission of the executable partition, and forbidding to modify the execution permission of the non-executable partition. According to the method, under the condition that system resources during operation are not consumed, execution of non-trusted programs can be effectively limited, it is ensured that only trusted programs can be executed in a system, and meanwhile the management process of executable programs is simplified.

The invention relates to the technical field of industrial control system security and industrial edge computing, and discloses a domestic industrial control security protection system with edge computing capability, which comprises a trusted state transition map generation module, a map embedding module, a map driving micro-isolation module, a mimicry verification module and a cloud edge collaboration module. The system firstly analyzes a control logic program of the industrial control equipment to construct a trusted state transition map comprising a security state, a legal instruction, a parameter space and a time sequence constraint, and the trusted state transition map and verification logic are solidified in a trusted execution environment together. During operation, the system performs micro-isolation on the real-time industrial control instruction by referring to the atlas. According to the method, an industrial control mechanism is combined with mimicry defense and trusted computing, the problems that a defense strategy of a traditional protection technology is fixed, and unknown attacks and time sequence cheating are difficult to resist are solved, and the edge protection capacity which is based on an endogenous mechanism, safe and dynamically evolved is achieved.

The invention discloses a software upgrading method and device and a storage medium, and belongs to the field of communication. The method comprises the steps that a second module code of a first software module is compiled in a runtime system to obtain a second shared library, the second shared library comprises data of multiple states, and the version of the second module code is higher than that of the first module code; pausing the runtime system to run the first module code based on the data of each state included in the first shared library; synchronizing data of at least one state included in the first shared library to the second shared library, wherein the at least one state is a state commonly included in the first shared library and the second shared library; and controlling the runtime system to run the second module code based on the data of each state included in the second shared library. According to the method, the software upgrading scene can be enriched.

The present disclosure provides a method for managing allocation of compute resources in a distributed database. Each request on every task is allocated a certain number of tokens when it arrives, such as by using a token bucket algorithm. This initial allocation may be the per-request maximum usage permitted in the system. This request is made to a central quota server in the region to coordinate allocations between tasks. As the task runs, the system evaluates how much of the limit the request is using, and adjusts its limit up or down to keep the quota allocation tight on the actual usage of the request. If there aren't enough tokens to keep the request active, the request can be evicted.

A memory anomaly detection method includes: sending a startup command to an artificial intelligence (AI) application runtime system to launch the AI ​​application; during the AI ​​application's operation, identifying anomaly operators based on the actual memory access information and memory allocation information of at least one operator among multiple operators; detecting the memory access behavior of the anomaly operators based on their input / output information during the AI ​​application's operation; and obtaining detection results. This method divides the memory anomaly detection process into two stages: delimiting operators and single-operator detection, transforming the entire network problem into a single-operator problem. The entire detection process requires no compiler or hardware support, is independent of source code, and is applicable to scenarios with only binary code, exhibiting high availability. Furthermore, this method transforms the memory pressure of performing memory anomaly detection on the entire network into the memory pressure of the delimiting operator stage within the entire network, minimizing the impact of the detection process on the overall network's memory.

The invention discloses a plugin-based Widget implementation method and related equipment, relates to the field of mobile application development, and mainly aims to solve the problem that the layout of the current Widget function is too rigid and not intelligent enough. The method comprises the following steps: querying a plug-in component information list through a preset identifier to obtain key component information; obtaining a dual-environment view instance based on the key component information; the dual-environment view instance is integrated into a host Wi-get container to achieve a dual-environment operation system of Widget, and the dual-environment operation system comprises a host operation environment and a plug-in operation environment. The Widget implementation method and device are used for the Widget implementation process based on plug-in.

This disclosure provides a PLC program execution system, relating to the field of programmable logic controller (PLC) technology. It includes a runtime system and an execution mechanism. The runtime system stores data and performs network interactions, and has a unified interaction interface. The execution mechanism is connected to the runtime system through the interaction interface and is used to access memory and the network, selecting an implementation method to execute the PLC program. The execution mechanism can be implemented as an interpreted executor, a compiled executor, or a JIT executor. This disclosure, through the execution mechanism in the PLC program execution system, allows for the selection of an implementation method from multiple options to execute the PLC program. This flexible selection of implementation methods improves the flexibility of the PLC program execution system and solves the problem of insufficient flexibility in related technologies where users cannot choose the execution method of the PLC program.

The invention discloses a memory management method of a runtime system, a programmable logic controller and the runtime system, relates to the technical field of programmable logic controller management, and discloses the memory management method of the runtime system, which comprises the following steps: acquiring a memory configuration file when the runtime system is started, the memory configuration file comprises memory configuration information of a memory area in the runtime system under different scenes; and determining a target numbered memory block according to the memory configuration information and preset numbered memory block information, and performing memory management on a power-down retention variable in the memory configuration information according to the target numbered memory block. According to the method, the memory management accuracy of the system during operation is improved.

Systems and methods for generating and executing distributed code. The systems and methods include receiving serial code generated by a large language model (LLM) for vision applications and analyzing the serial code with a trained model to identify code dependencies and detect independent application programming interface (API) calls. The systems and methods further include transforming the serial code by incorporating program semantics that enable concurrent execution of the independent API calls and generating distributed code configured for execution on a container orchestration platform cluster, wherein the distributed code includes service calls that can be understood and executed by a runtime system.

According to some embodiments, systems and methods are provided including receiving a scenario including a plurality of steps; assigning each step to one of: a baseline phase, a bridge phase and a post-upgrade phase, the bridge phase including a transition from a first to a second runtime system; executing an automate for the scenario including a first pause at an end of the baseline phase and a second pause at an end of the bridge phase; initiating, during the bridge phase, the transition of the system from the first to the second runtime system in response to the first pause; resuming execution of the automate for the one or more steps assigned to the bridge phase during the transition; initiating the second pause; completing the transition to the second runtime system; and resuming execution of the automate for the post-upgrade phase. Numerous other aspects are provided.

This invention relates to a runtime server (100) for simultaneously executing multiple runtime systems (101) within an operating system (103) for a data processing device controlling an automated device based on a device control program. The runtime systems (101) are configured to execute the device control program in real time. The runtime server includes at least two runtime systems (101) for executing application modules (105) of the device control program, wherein at least one application module (105) for executing the device control program is installed on each runtime system (101), and each runtime system (101) has a data transmission interface for data transmission between runtime systems (101) and / or between application modules (105). 107), wherein each runtime system (101) has an I / O configuration (109) that defines an allocation between at least one variable of an application module (105) of the runtime system (101) and at least one hardware address of a hardware component (113) of the automation device (112) to be controlled; an I / O interface (111) having at least one I / O input (115) and / or I / O output (117) for exchanging data between at least two runtime systems (101) and the hardware component (113) of the automation device (112); and an I / O mapping intermediate layer (119) in which the I / O configurations (109) of at least two runtime systems (101) are mapped. The invention further relates to a computer program product (200) for executing a runtime server (100).

A method includes receiving, by a user interface layer in a client terminal, application code input by a user, processing the input application code by a client-as-a-service layer in the client terminal, and forwarding the input application code to be stored in a cloud-based repository. The method further includes synchronizing application code stored in a cache with application code stored in the cloud-based repository, and performing a runtime remodel on a runtime system based on the input application code.