48 results about "Performance tuning" patented technology

The invention provides a performance optimization method and system for aggregate communication operation, electronic equipment and a medium, and the method comprises the steps: executing a closed-loop optimization process taking meta-heuristic search as a core through determining a multi-dimensional parameter space for describing a hierarchical protocol strategy of the aggregate communication operation, a complex mapping relation between different strategy configurations in a parameter space and the actual performance of the system is fully explored; the target code of the candidate strategy is automatically generated and actually measured based on the historical performance data, the actually measured performance feedback is obtained through continuous iteration, and the target hierarchical protocol strategy of the set communication operation is obtained. According to the method, the centralization bottleneck problem in a direct protocol mode is effectively solved, and the communication efficiency and the resource utilization efficiency in a large-scale parallel system are remarkably improved. Moreover, the closed-loop optimization process gets rid of dependence on artificial experience, reduces the complexity and cost of manual adjustment and optimization in a multi-factor coupling environment, and improves the adaptability, performance and transportability of integrated communication operation in different platforms and tasks.

The invention provides a server control method, device and equipment and a readable storage medium, and the method comprises the steps: deploying a decision engine at a firmware layer of a system stack, and carrying out the data driving based on the current operation state of a server and a historical performance adjustment scheme to automatically generate a plurality of candidate performance adjustment schemes; the performance expression and the security risk of the server under the candidate performance adjustment scheme are further evaluated, and the decision engine combines the estimated performance and the security risk of the candidate performance adjustment scheme to select a target performance adjustment scheme considering both the performance and the operation security; the control parameters of the server component are directly written through the specified interface of the firmware layer, operation system scheduling and software stack transfer links are bypassed, full-process autonomous closed-loop control from performance optimization to safety protection is realized on the premise of ensuring safe operation of the system, the performance adjustment efficiency and the response speed of the server are improved, and the service life of the server component is prolonged. And the performance adjustment cost is reduced.

The invention discloses an application program resource monitoring method and device, electronic equipment and a storage medium, and relates to the technical field of computers, and the application program resource monitoring method comprises the steps that an out-of-band monitoring module is adopted to collect application program resource data, isolation of a monitoring channel and a service channel is achieved, and the service channel is monitored. Meanwhile, a resource use model of a time dimension is constructed based on real data to simulate production environment resource behaviors, and then performance pressure measurement and fault reproduction analysis are supported, so that the problems that in an existing monitoring and control method, due to the fact that an internal agent of a virtual machine is adopted and the virtual machine is not isolated from a service channel, monitoring data distortion is caused, application program operation is disturbed, and the service performance is affected can be solved. According to the technical scheme of the invention, the technical problems in the prior art are solved, and the technical effects of guaranteeing the accuracy of monitoring data, avoiding the interference on the operation of the application program, improving the performance optimization effect and effectively supporting the performance pressure measurement and fault reproduction analysis of the cloud computing system are achieved.

The embodiment of the invention discloses an AI full stack performance tuning tool for top-down analysis. A front end, a middle end and a rear end are separated; the front end is used for visually displaying summarized information needing to be concerned on each level of the AI full stack and specific execution information of different granularities; the middle end is used for connecting the front end and the rear end and providing a plurality of configurable visual components; and the rear end is used for receiving the data files in various data formats, processing the data files and storing the processed data files in a plurality of mutually decoupled databases for the middle end to query and call after configuration. The method has the advantages that the working efficiency of developers of all levels can be improved; a developer can be assisted to realize more accurate and deep performance optimization; the personalized configuration requirements of different levels of development on the front-end interface can be met; the large model deployment and use cost can be reduced, and the utilization rate of the computing power cluster system and the computing power chip is improved.

The invention discloses a distributed storage engine performance optimization method and device, equipment and a storage medium, and relates to the field of data storage, and the method comprises the following steps: obtaining a target search domain and a particle swarm individual number; initializing a target particle swarm based on the target search domain and the particle swarm individual number, performing performance optimization on a target distributed storage engine by using a parameter combination scheme corresponding to each target particle in the target particle swarm, and determining a multi-target function value corresponding to each parameter combination scheme based on an optimization target; the optimization target comprises the storage performance and the CPU utilization rate; and based on the Pareto dominance relationship and the multi-objective function value corresponding to each parameter combination scheme, determining a target optimal parameter combination scheme so as to perform performance optimization on the target distributed storage engine based on the target optimal parameter combination scheme. According to the method and the device, the performance of the distributed storage engine is optimized.

This invention discloses a software code optimization method and system based on big data. The method includes the following steps: constructing a code optimization knowledge base; acquiring the target code to be optimized and its corresponding estimated runtime data features; parsing the target code, extracting its structural features, and fusing the structural features with the estimated runtime data features to generate a target code fingerprint; performing similarity matching between the target code fingerprint and historical code fingerprints in the code optimization knowledge base to determine target historical code fingerprints that match the target code fingerprint; and determining a target optimization strategy for the target code based on the optimization strategy corresponding to the target historical code fingerprint. This invention can improve development efficiency, reducing the performance tuning work of senior architects, which can take hours or even days, to automated recommendations at the minute level, greatly improving software iteration efficiency.

A method may include obtaining multiple tunable parameters associated with a data transform accelerator operable to perform data transform operations. The method may also include configuring a resource configuration vector based on the multiple tunable parameters. The method may further include obtaining a target performance metric. The method may also include measuring one or more performance metrics associated with the data transform accelerator. The method may further include automatically tuning at least one tunable parameter of the multiple tunable parameters to obtain tuned parameters in response to a performance metric of the one or more performance metrics failing to satisfy the target performance metric. The method may also include updating the resource configuration vector in view of the tuned parameters.

A container-based orchestration system includes a master node and a plurality of worker nodes. The master node can receive, from each agent executing on a corresponding worker node, node characteristics associated with the worker node. The master node can determine, for each worker node, one or more parameters corresponding to the node characteristics associated with the corresponding worker node and a node profile of the worker node and provide the parameters to the agent executing on the corresponding worker node. The agent configures the worker node in accordance with the parameters. In response to receiving a request to deploy a pod to a worker node, the master node can select a worker node to receive the pod based on the node characteristics and the pod characteristics. The agent can configure the selected worker node to execute workloads of the pod in accordance with the one or more parameters.

This invention discloses a system performance optimization method and related equipment based on layered decoupling. When a kernel service node detects a cross-process service call, it obtains the call context information, filters target communication data that meets key service filtering conditions from the call context information to identify the current scenario and determine the current scenario identifier data. The hardware abstraction layer service performs performance parameter matching on the current scenario identifier data and performs business logic processing based on the parameter matching results to generate performance tuning data. The kernel service node then executes the performance optimization processing corresponding to the performance tuning data. This invention enables system performance optimization based on layered decoupling, significantly improving performance optimization efficiency. Through the layered decoupling architecture, it greatly reduces the difficulty and workload of code refactoring and adaptation, improves the maintainability and upgrade efficiency of the performance optimization framework, and can be widely applied in the field of system performance optimization technology.

The application discloses a kind of Xinxin terminal compatibility mutual authentication and performance tuning pedestal system, it is related to Xinxin terminal compatibility mutual authentication and performance tuning pedestal technical field, for solving the problem that address book module, service number module and workbench module under multiple Xinxin operating system and domestic CPU architecture are difficult to be uniformly carried, multi-security domain service routing error and terminal performance and compatibility are difficult to collaborative optimization;Around Xinxin terminal pedestal, collaborative work such as dependent reconstruction module, mobile application interface architecture rendering scheduling module is constructed, based on environmental compatibility score, Xinxin terminal pedestal construction and dependent reconstruction are automatically completed, under the premise that unlicensed personal information in public places is not collected, in combination with business module loading sequence optimization, interface scaling and rendering priority control and performance score and Xinxin compatibility test matrix closed loop optimization, improve the unified interface experience and running stability under multiple operating systems, multiple CPU architecture and multiple security domain deployment.

A system for incorporating user feedback in a large language model (LLM) based retrieval augmentation (RAG) tools includes an application for incorporating user feedback (UFA). The UFA receives an input to the LLM from a host device user, generates an LLM output as a response to the input, causes a system user to verify the LLM output and provide user feedback via a human-machine interface (HMI) of the host device, and in response to the user feedback, engages a performance tuning optimizer that modifies the LLM output by: revising domain knowledge, revising system prompts, and performing constrained regeneration of the LLM output. The performance tuning optimizer progressively reduces computational resource utilization, progressively increases computational efficiency, and progressively reduces reliance on human validators and user feedback over time, and wherein the LLM output is a command to one or more systems of the host device.

A system for incorporating user feedback into large language model (LLM)-based retrieval augmentation (RAG) tools includes a user feedback incorporation (UFA) application. The UFA receives input into the LLM from a user of a host device, generates an LLM output in response to the input, prompts a system user to verify the LLM output and provide user feedback via a human-machine interface (HMI) of the host device, and, in response to the user feedback, deploys a performance tuning optimizer that modifies the LLM output by revising domain knowledge, revising system requests, and performing constrained regeneration of the LLM output.The performance tuning optimizer progressively reduces the use of computing resources over time, progressively increases computing efficiency, and progressively reduces the reliance on human reviewers and user feedback, with the LLM output being a command to one or more systems of the host device.

The invention relates to a mercuric chloride NPU operator automatic tuning method and system based on fusion operator performance modeling, and the method comprises the steps: firstly constructing an end-to-end performance analysis model of a fusion operator for collaborative execution of a Cube core and a Vector core, and employing two modes to accurately predict execution delays under different parameter configurations through dynamic discrimination of a calculation speed relation between double cores; on the basis, a heuristic design space exploration algorithm with minimization of data movement amount as guidance is provided, and the algorithm is combined with the multilevel storage characteristic of the mercuric chloride NPU, adopts a two-stage strategy of block size optimization and buffer area allocation collaboration, and automatically searches for optimal parameter configuration meeting on-chip storage capacity constraints. According to the method, automatic performance adjustment and optimization of the fusion operator are realized, compared with traditional manual adjustment and optimization or compiler default configuration, near-optimal configuration can be found within a few minutes, the average performance prediction error is lower than 5%, the maximum performance speed-up ratio of part of operators can reach 1.46 times, and the efficiency of operator development and deployment on the mercuric chloride NPU is remarkably improved.

The invention relates to the technical field of GPU (Graphics Processing Unit) performance analysis, in particular to a performance analysis and tuning system based on a memory access intensive program, which comprises a GPU, a test case corresponding to a to-be-analyzed program, a processor and a memory in which a computer program is stored. The method comprises the following steps of determining an ideal bandwidth when the GPU runs the test case according to a memory access parameter corresponding to the test case and a thread starting rate corresponding to the GPU, comparing the ideal bandwidth with an actual bandwidth when the GPU runs the test case, determining whether adjustment and optimization are needed or not, and after it is determined that adjustment and optimization are needed, starting the GPU according to the adjustment and optimization. According to the method, whether parameter tuning is needed or not is judged according to the first limiting condition, the tuning target and the tuning limiting condition are provided, and the tuning direction is provided for performance tuning, so that the accuracy and efficiency of performance analysis and tuning are improved.

The invention discloses a Spark job performance adjusting and optimizing method and device and a storage medium, relates to the technical field of data processing, and discloses a Spark job performance adjusting and optimizing method which comprises the steps that in response to a performance diagnosis request for a target Spark job, multi-dimensional operation data of the target Spark job is acquired; determining a target data type and a data extraction rule corresponding to each diagnosis scene according to a plurality of preset diagnosis scenes; according to the target data type and a data extraction rule, extracting input data required by each diagnosis scene from the multi-dimensional operation data; based on the input data, performing diagnosis analysis on the corresponding diagnosis scenes to obtain diagnosis results of the diagnosis scenes; and summarizing the diagnosis results to recognize the performance bottleneck of the target Spark job, and generating a final diagnosis report containing the optimization scheme, thereby improving the recognition accuracy and diagnosis flexibility of the performance bottleneck of the Spark job under the influence of multiple factors.

The application provides a performance optimization method and system of ultra-high temperature cementing cement, relates to the field of artificial intelligence, and comprises the following steps: acquiring performance influence characteristics and compressive strength sample data set of the ultra-high temperature cementing cement; taking the performance influence characteristics of the ultra-high temperature cementing cement as input characteristics, taking the compressive strength of the ultra-high temperature cementing cement as an output target, determining an optimal model from a set machine learning model library according to the influence coefficient of the input characteristics on the output target; determining the influence coefficient of each performance influence characteristic on the compressive strength; screening target characteristics with an influence coefficient greater than a set influence threshold, and generating an optimization scheme of the ultra-high temperature cementing cement according to the influence coefficient of each target characteristic. The application can optimize the mechanical properties of the ultra-high temperature cementing cement based on the optimization scheme, so as to ensure that the ultra-high temperature cementing cement can adapt to different application scenarios and meet the application requirements in normal temperature, high temperature, ultra-high temperature and other scenarios.

The embodiment of the present application provides a performance tuning method and device, wherein the method comprises the following steps: obtaining historical performance data generated in a process of a host task carried by a storage system; setting the host task to be stored by a second node when a first node is abnormal; analyzing the historical performance data and an initial performance index input by a user to generate a predicted performance index after the first node returns to normal; selecting a target performance index from the predicted performance index to determine a performance tuning configuration parameter; and finally executing a data synchronization task by using the performance tuning configuration parameter to realize synchronization of difference data between the first node and the second node, so that the purpose of intelligently tuning performance in a storage dual-active fault recovery scene is achieved, thereby realizing the technical effects of optimizing resource allocation and improving data synchronization efficiency, and solving the problem that the resource allocation mechanism of host services and synchronization tasks is not coordinated enough and resource utilization is difficult to guarantee in the related art.

The application discloses a kind of based on big language model's chart alternative text quality automation evaluation method and system, it is related to big language model field and chart alternative text field, including: by demand investigation, obtain chart alternative text quality evaluation scheme;Artificially construct chart alternative text evaluation standard data set carries out data preprocessing, based on GLM4-9B model carries out Finetune training and performance tuning.The chart alternative text quality automation evaluation method proposed in the application is based on chart accessibility reading criteria to design evaluation scheme, and the prediction score of the transfer model obtained by training is highly correlated with the artificial evaluation score, which can provide high-reliability automated chart alternative text quality evaluation results and can be applied to text quality evaluation in the field of chart alternative text.

This application provides a performance analysis method, apparatus, electronic device, chip, and medium, relating to the field of chip design and development. The method includes: generating at least one set of test configuration schemes based on the performance requirements of the module under test and a preset performance scenario; performing performance simulation of the module under test under the preset performance scenario in parallel based on the at least one set of test configuration schemes to trigger at least one target performance event and acquire at least one performance data point through a preset performance monitoring unit; quantifying and analyzing the at least one performance data point to obtain performance analysis results under the preset performance scenario; and updating at least one set of test configuration schemes based on the performance bottleneck when the performance analysis results show a performance bottleneck, thereby determining the optimal configuration parameters of the optimal test configuration scheme based on the updated at least one set of test configuration schemes. This improves performance tuning efficiency, reduces development cycle and labor costs, and ensures that the module under test achieves optimal resource allocation while meeting performance targets.

A system for incorporating user feedback into a large language model (LLM)-based retrieval enhancement (RAG) tool includes an application for incorporating user feedback (UFA). The UFA receives an LLM input from a host device user, generates an LLM output as a response to the input such that the system user verifies the LLM output and provides user feedback via a human machine interface (HMI) of the host device, and in response to the user feedback, enables a performance adjustment optimizer to adjust performance of the host device. The performance adjustment optimizer modifies the LLM output by revising domain knowledge, revising system cues, and performing constrained regeneration of the LLM output. Over time, the performance adjustment optimizer gradually reduces computing resource utilization, gradually increases computing efficiency, and gradually reduces dependency on human verifier and user feedback, and wherein the LLM output is a command to one or more systems of the host device.

The invention discloses a USB (Universal Serial Bus) network card performance tuning method based on interrupt binding, which relates to the technical field of computer system performance optimization and comprises the following steps: determining a high-performance CPU (Central Processing Unit) core as an anchor point core and closing interrupt automatic load balancing; analyzing the data packet in real time on a USB driving layer and distributing semantic tags based on a preset rule; dynamically deciding an interrupt triggering strategy according to the semantic label, carrying out aggregation waiting on the video stream, and immediately triggering interrupt when a high-priority packet is identified; dynamically binding the interrupt to an anchor core or other cores according to the semantic tag and the system load, and associating a corresponding network queue; monitoring the load of the anchor point core, and switching the specified data flow from an interruption mode to a controlled polling mode when the load exceeds a threshold value; according to the method, the balance of high throughput, low delay and low CPU (Central Processing Unit) occupancy is realized through semantic perception and dynamic intelligent scheduling.

The application discloses a page response performance optimization method and device, equipment, a storage medium and a computer program product. The method comprises the following steps: obtaining page access data of a user; determining an important element position area in a to-be-accessed page based on the page access data, and extracting a page important element set in the important element position area; obtaining pixel change information corresponding to the page important element set, and calculating a page response time delay according to the pixel change information; and formulating a page optimization scheme based on the page response time delay and an access condition corresponding to the page access data. The application improves the efficiency of calculating page performance data.

Methods, systems, and devices for adaptive performance tuning for memory systems are described. A memory system may apply a programming mode ratio. For example, the memory system may write data to single-level cell (SLC) word lines and multiple level cell word lines according to a ratio. The memory system may determine the ratio according to operating parameters, such as a command queue depth, size of commands in the command queue, host delays, an amount of free space, or the like. The memory system may determine the ratio based on whether the operating parameters satisfy thresholds. That is, the memory system may compare each operating parameter to one or more operating parameter thresholds.

A material handling vehicle receives a wirelessly transmitted performance tuning profile and, in response, adjusts at least one operational capability. While the material handling vehicle operates in its work environment, the performance tuning profile can be dynamically updated based on multiple different factors, including operation, operator, environment, others, or combinations thereof. An additional aspect provides a graphical user interface that allows individuals (such as supervisors) to create a library of performance tuning profiles and to create and / or program a rule engine to automatically transmit appropriate performance tuning profiles to the corresponding material handling vehicle in response to the detection of a corresponding event associated with a programming rule.

The invention provides a performance parameter adjusting method, electronic equipment and a performance parameter adjusting system, and relates to the technical field of communication, target performance parameters of the equipment can be adaptively adjusted according to state data such as date information, environment temperature and operation information of a user, so that the equipment can obtain different performance parameters in different states, and the user experience is improved. Therefore, the operation performance of the equipment can be better exerted, the jamming phenomenon of the equipment is reduced, and the user experience is improved.

The invention discloses a credential terminal compatibility mutual authentication and performance tuning base system, and relates to the technical field of credential terminal compatibility mutual authentication and performance tuning bases. The method is used for solving the problems that an address book module, a service number module and a workbench module are difficult to bear in a unified manner under various credential operation systems and domestic CPU architectures, multi-security domain service routing is easy to make mistakes, and terminal performance and compatibility are difficult to collaboratively optimize. According to cooperative work of a credential terminal base construction dependency reconstruction module, a mobile application interface architecture rendering scheduling module and the like, credential terminal base construction and dependency reconstruction are automatically completed based on environment compatibility scores, and the reliability of the credential terminal base is improved on the premise that unauthorized personal information of public places is not collected. And in combination with business module loading sequence optimization, interface zooming and rendering priority control and performance scoring and credential compatibility test matrix closed-loop tuning, the unified interface experience and operation stability under multi-operation system, multi-CPU architecture and multi-security domain deployment are improved.

The application discloses a digital twin water conservancy super-fusion algorithm model and a hardware system performance optimization method, and belongs to the technical field of a digital twin water conservancy system. Real-time interaction data of a model is extracted, and a docking configuration system is built. The docking configuration system has an adaptive interface for external service and an interaction mechanism for internal processing. The adaptive interface is adapted to a heterogeneous platform, realizes dynamic acquisition and analysis of external data of the model to obtain input data, and the interaction mechanism is embedded with a hydrology-hydrodynamic river network multi-scale hierarchical model to obtain real-time model parameter values and broadcast focus point information according to the input data. Forecasting and early warning parameter calculation is performed based on the real-time interaction data of the model to obtain current output data, and the current output data is converted into display data for rolling display. The docking configuration system with the adaptive interface is built to adapt to data formats of the heterogeneous platform and break data islands, and a parameter self-adaptive adjustment mechanism combining particle swarm optimization and reinforcement learning is adopted to solve the problem of insufficient adaptability of static parameters.

This invention discloses a method, apparatus, device, and medium for intelligent predictive tuning of JVM based on a large model, relating to the field of artificial intelligence. The method includes: collecting multimodal runtime metric data of a target JVM instance; performing feature engineering on the multimodal runtime metric data to generate standardized feature vectors; inputting the standardized feature vectors into a pre-trained performance prediction model to obtain the predicted probability of a performance degradation event occurring in the target JVM instance within a preset future time window; in response to the predicted probability exceeding a preset probability threshold, generating at least one candidate JVM parameter optimization scheme that satisfies preset objective constraints through a multi-objective constraint optimization algorithm; selecting a target optimization scheme from the at least one candidate JVM parameter optimization schemes, and safely executing the target optimization scheme on the target JVM instance. This invention can significantly improve the accuracy and reliability of JVM performance tuning, effectively enhancing the control effect of JVM performance.