46results about How to "Improve memory utilization" patented technology

The invention provides a shared memory-based data interaction method for an embedded virtualization platform, which comprises the following steps of: dividing a shared memory area, including a management information area and a shared data area, initializing the management information area, including setting a magic number, configuring a mutual exclusion access flag and recording use information of the shared data area; acquiring a physical initial address of the shared memory, verifying the availability of the management information area, and ensuring that the management information area is correctly initialized; a proper data transmission mode is selected according to the data size, small data volume can be directly transmitted through an inter-core communication register, and large data volume is transmitted through a shared memory; the Szymanski algorithm is used to realize mutual exclusion access control, and the consistency of data is ensured; after data interaction is finished, if shared memory resources which do not need to be used any more need to be released, use information in the management information area is updated; and other related operation areas or virtual machines are notified according to needs, and state changes of data interaction are notified to the operation areas or the virtual machines.

The invention discloses a voxel map plane merging method, system and device and a storage medium, and the method comprises the steps: S1, carrying out the voxelization of a point cloud, detecting a plane in a voxel, and recording the feature information of the plane; s2, classifying the planes in the voxels according to the plane feature information, and calculating a hash key value; and S3, combining the plurality of planes according to the hash key values and the position relationship of the different planes. According to the method, by calculating the hash key values of the voxel planes, the voxel planes with similar geometric attributes can be quickly merged into the same hash bucket and are uniformly processed, so that the real-time performance and the resource utilization efficiency of the system are greatly improved, and particularly, when a large-scale environment is processed, the real-time performance of the system is greatly improved. The defects of a traditional fixed voxel size method can be more effectively overcome.

The invention relates to the technical field of computers, and discloses a memory management method and device, computer equipment and a storage medium, which are applied to a baseboard management controller, the method comprises the following steps: acquiring historical memory use data of the baseboard management controller; generating a memory usage data prediction model based on the historical memory usage data; in each preset period, inputting the memory usage data of the previous preset period of the substrate management controller into the memory usage data prediction model, and obtaining a predicted value of the memory usage data of the current preset period; based on the predicted value of the memory usage data, pre-allocating memory resources; acquiring load information of the substrate management controller; and adjusting the pre-allocated memory resources based on the load information. According to the load information of the substrate management controller, the memory resources are dynamically allocated, and the use efficiency of the memory resources is improved.

The invention belongs to the field of RDMA and cloud memory pooling, and provides a cloud memory pool priority flow control method and system based on RDMA, a cloud server is used as a cloud cloud memory pool of a plurality of local nodes, the local nodes read and write memories in a cloud memory pool server through RDMA, the memory capacity of the local nodes is logically expanded, and the priority flow control of the cloud memory pool is realized. The memory demand and the memory utilization rate of each local node on the cloud memory pool are monitored through a dynamic regulation and control algorithm, the memory capacity allocated to each local node by the cloud memory pool is dynamically regulated and controlled, the memory utilization rate is improved, and meanwhile, by maintaining a table containing the priority of each local node, when network congestion occurs, the network congestion rate is improved through priority scheduling. Network congestion control is realized, high-speed and stable data transmission and remote memory access are provided for high-priority nodes while the network throughput is ensured, and the processing speed of important transactions is ensured.

The Android system memory management optimization method is characterized by comprising the following steps: (1) memory fragment arrangement: (1-1) fragment detection and classification; (1-2) fragment merging; (1-3) rearranging memory blocks; (2) memory leak detection: (2-1) reference chain analysis; (2-2) garbage collection monitoring; (2-3) memory leak repair suggestion; (3) memory allocation optimization: (3-1) memory pool management; (3-2) dynamically adjusting the size of a memory block; (3-3) multiplexing the memory blocks; (4) a memory allocation strategy under multi-task processing: (4-1) task priority evaluation; (4-2) memory sharing; and (4-3) a memory recovery strategy. According to the method, the memory utilization rate of the Android system is improved, memory leakage is reduced, the memory allocation efficiency is improved, and memory allocation under multi-task processing is optimized.

The invention relates to a task processing method and device, computer equipment, a computer readable storage medium and a computer program product. The method comprises the steps that in response to starting of a to-be-executed task, the memory requirement of the to-be-executed task is determined; distributing a target memory for the to-be-executed task according to the memory requirement and the memory operation strategy, and executing the to-be-executed task through the target memory; and in the process of executing the to-be-executed task, according to the memory state information of the target memory, carrying out dynamic allocation updating on the target memory until the to-be-executed task is executed. By adopting the method, the memory utilization rate can be improved.

The invention relates to the technical field of computers, and particularly provides a data stream cardinal number estimation method supporting element-level deletion operation, which comprises the following steps of: in response to updating operation, updating a bucket data sequence stored in a virtual bucket; wherein the number of the virtual buckets is multiple, each virtual bucket comprises a guard unit and a plurality of conventional units, the guard unit is used for recording a minimum sampling value, and the conventional units are used for storing fingerprints; obtaining the number of non-empty units, and calculating a bucket level cardinal number of each virtual bucket according to the number of the non-empty units and the updated bucket data sequence; wherein the number of the non-empty units is the number of the conventional units, in which the fingerprints are stored, in the virtual bucket; and accumulating the bucket-level cardinal numbers of the virtual buckets, and calculating the global data cardinal number of the data stream. According to the method, the target data is dynamically updated and the global data cardinal number of the data stream is calculated based on the virtual bucket, the memory utilization rate is high, and calculation is fast and accurate.

The invention relates to the technical field of electric power acquisition terminal upgrading, and discloses an upgrading method for an electric power acquisition terminal, which comprises the following steps: equally dividing an upgrading file into a plurality of upgrading sub-files; for each intelligent electric meter, executing a communication error rate statistical strategy, and calculating the communication error rate of each intelligent electric meter; equally dividing 24 hours in a day into a plurality of divided time periods; according to a radius division strategy, classifying the intelligent electric meters according to the distance between the intelligent electric meters and the to-be-upgraded terminal to obtain a plurality of time period electric meter sets; wherein each time period electricity meter set corresponds to one division time period, and the intelligent electricity meters in the time period electricity meter set send the same upgrading sub-file to the terminal to be upgraded in the corresponding division time period; sending the upgrading sub-file to the to-be-upgraded terminal through the intelligent electric meter in each divided time period; executing an upgrade ending prediction strategy, and predicting a moment when the to-be-upgraded terminal finishes upgrading; and the plurality of standby intelligent electric meters improve the transmission success rate of the upgrade sub-files.

The invention discloses a multi-node high-performance data sharing index method based on a CXL memory architecture. A main server node is used for constructing a CXL data index area and a data storage area. After each server node is connected with the CXL memory device, the CXL memory device is mounted to the server node in an NUMA binding mode, and a sharing area is appointed, so that sharing access of multiple nodes to CXL physical memory data is realized. After the main server node is connected to the CXL memory, the main server node constructs a CXL data index area and a shared storage area, and supports multi-node data storage and query through a Hash index table. When a request is queried, the position of data in the hash table is positioned through the hash function, query and matching are carried out in the data hash index table, and finally the memory area where the data are located is positioned. And after the data is found, the server node loads the data from the CXL memory to the local memory for use, so that cross-server node data sharing is realized.

The present invention discloses a method for detecting tumor multi-genes based on hyperspectral images, which belongs to the technical field of mutation sample detection, and specifically includes: obtaining tumor tissue sections of different carcinogenic driver gene mutation states after H&E staining; photographing tumor tissue sections by a hyperspectral camera, obtaining hyperspectral images of tumor samples and preprocessing; building a global feature extractor based on S3Anet; building a tumor multi-gene detection model and using multi-task learning training; collecting tumor samples to be detected, and obtaining hyperspectral data of the samples to be detected by a hyperspectral camera; inputting the hyperspectral data into a trained S3ANet model for feature extraction, and then inputting it into a trained multi-gene detection model to obtain the gene detection results of the tumor samples. The beneficial effects of the present invention are: the present invention effectively reduces the dependence on traditional pathological diagnosis experience and gene sequencing, significantly improves the accuracy, reliability and generalization ability of detection, and reduces the detection cost.

This application provides a method, system, device, and medium for virtual terminal resource preloading and caching management based on the ISOBUS standard. This application innovatively constructs a multi-level caching architecture and combines a Markov chain-based resource prediction engine with adaptive memory management technology to achieve intelligent preloading and efficient caching of interface resources. Addressing the pain points of existing ISOBUS VT systems, such as high resource loading latency, low cache hit rate, and insufficient memory utilization, this invention provides a targeted technical solution. Experimental data verifies that this technical solution can reduce resource loading time by 50% to 70%, increase cache hit rate to over 92%, improve memory utilization by 40%, and reduce system crash rate by over 80%.

The application discloses a memory management method and electronic equipment, and relates to the technical field of computers, including any controller of a target cluster, based on the current memory management parameter of the controller, performing corresponding memory configuration on the controller, and running a preset test case on the controller to obtain the performance index of the controller under the current memory configuration, determining the fitness function value of the controller under the current memory configuration, and then determining the self historical optimal memory management parameter of the controller; if the current iteration number does not reach the iteration number threshold, updating the memory management parameter based on a target update speed, and continuing iteration based on the updated memory management parameter; if the current iteration number reaches the iteration number threshold, performing corresponding memory configuration on the controller based on the self historical optimal memory management parameter of the controller. The application realizes dynamic management of the memory of the controller, and achieves the technical effect of improving the memory utilization rate.

The invention discloses a strategy optimization method for cache garbage collection, relates to the technical field of computer cache management, and is used for solving the problem of low cache collection efficiency. According to the method, a cache storage area is abstracted into a plurality of cache partitions, a partition state vector composed of capacity, access and fragmentation degree operation indexes is constructed for each partition, and a strategy evaluation thread periodically reads the partition state vectors to calculate garbage recycling pressure and recycling priority. The upper limit of the number of scanning objects and the upper limit of time slices are distributed for the target cache partition by combining the idle time proportion of the processor, and a recycling execution thread conducts incremental recycling in a batch mode according to a strategy list and records the recycling progress; after recovery, index weight and scanning configuration are adjusted according to release capacity, fragment degree change, hit times and back-to-source request change, so that garbage recovery gives consideration to release space and overhead, the hit rate and the memory utilization rate are improved, and pause and invalid recovery caused by centralized recovery are reduced.

The invention relates to an intelligent memory management system and management method based on a Java virtual machine, the system comprises a memory monitoring system, an intelligent memory management system and a manual intervention management system, the memory monitoring system is respectively connected with the intelligent memory management system and the manual intervention management system; the detected memory problem is given to a manual intervention management system, and the memory without the problem is managed by a memory intelligent management system. By dynamically monitoring, optimizing and managing the memory use of the Java application program, the purposes of efficiently utilizing the memory, reducing the memory leakage risk, improving the application performance and the like are achieved.

The invention discloses a GPU memory optimization method and system, and the method comprises the steps: obtaining the real-time change characteristics of a task load through obtaining the dynamic data flow information of the task load and an intermediate data generation rule; according to the real-time change characteristics of the task load, an optimization algorithm based on gradient descent is adopted to determine the distribution proportion and the reserved space of a memory pool, and adjusted memory pool configuration is obtained; extracting available memory block information from the adjusted memory pool configuration, and compressing the intermediate data with low access frequency by adopting a Huffman coding algorithm to obtain a compressed intermediate data storage unit; and if the access frequency of the compressed intermediate data storage unit is lower than a preset threshold value, temporarily storing the compressed intermediate data storage unit to a low-speed storage area, and reallocating memory resources according to the task priority to obtain an optimized memory allocation scheme. According to the method, the memory utilization rate is remarkably improved, the access delay is reduced, and efficient memory resource management is realized.

This application provides a license plate recognition method, apparatus, electronic device, and storage medium. The method includes: sliding a sliding window of preset length and width along the long side of a preset sliding area in a first image, acquiring multiple sliding window results generated during the sliding process; determining the cutting position of the image to be processed based on the multiple sliding window results and the cutting position; determining whether the license plate to be recognized is a double-layer license plate based on the cutting position; if the license plate to be recognized is a double-layer license plate, cutting and stitching the image to be processed according to the cutting position to obtain a second image; and inputting the second image into a license plate recognition model to obtain a first recognition result. In this technical solution, the electronic device only needs to store a single-layer license plate recognition model to simultaneously recognize both single-layer and double-layer license plates, improving memory utilization and reducing the computing power used.

Embodiments of the present disclosure disclose an adaptive memory conversion method and device for a distributed file system. A specific implementation of the method comprises: dividing a metadata set in a target directory to obtain a metadata subset group; setting a corresponding memory data structure form for each metadata subset; determining a target metadata subset; executing a corresponding access processing flow on the target metadata subset to obtain and output a final access result; obtaining access feature sets and system resource states of each metadata subset in the metadata subset group in real time to obtain a directory index set; determining memory data structure form conversion instructions corresponding to metadata subsets to be converted in the metadata subset group to obtain a conversion instruction set and a to-be-converted metadata subset group; and performing data reconstruction on the to-be-converted metadata subset group to obtain a converted metadata subset group. The implementation can improve memory utilization, reduce redundant memory occupation, and reduce unnecessary data access overhead.

This application provides a dynamic memory management method, apparatus, electronic device, and storage medium, relating to the field of computer technology. It generates dynamic memory resource demand information by acquiring at least two performance indicators of the system in real time, thereby determining the resource allocation strategy for a first type of memory management unit used for data compression and storage and a second type of memory management unit used for data exchange to non-volatile storage media. Based on this strategy, the operating parameters of both units are configured, and the system memory resources are scheduled and managed. Therefore, it solves the problems in existing technologies where static allocation strategies are used, a dynamic coordination mechanism between the two types of memory management units is not established, and resource allocation cannot be adjusted in real time according to system status. This achieves the technical effects of dynamically adapting memory resources, improving system response efficiency, increasing memory utilization, and optimizing overall system performance and stability.

The present application discloses a memory management method for a multi-core system and a multi-core system, device, and medium, and relates to the field of memory management technology. The method includes: the main core unit identifies the initial memory management priority of the main core unit and each slave core unit as the second level; the memory management priority also includes: the first level higher than the second level and the third level lower than the second level; judging the system operation status; updating and determining the current memory management priority of the main core unit and each slave core unit according to the system operation status and the second level; the memory manager performs memory scheduling and allocation processing according to the memory scheduling allocation request sent by the applicant, the system operation status, the type of the applicant, and the current memory management priority of the applicant, and allocates idle memory units to the applicant so that the applicant can execute tasks in the task queue; the applicant is the main core unit or the slave core unit. It can improve the system memory utilization and ensure that the multi-core system runs efficiently and correctly.

The embodiment of the present application discloses a memory management method, device, limited device, terminal, storage medium and program, wherein the memory management method is applied to a resource limited device, comprising: dividing a memory region managed by a virtual machine into a first memory region, a buffer common area and a second memory region; wherein the first memory region and the second memory region are two end used memory regions of the memory region managed by the virtual machine; the buffer common area is an intermediate unused memory region of the memory region managed by the virtual machine; the first memory region and the second memory region allocate memory space from the buffer common area on demand; the first memory region is used to store a management virtual machine stack and an indirect access variable stack; and the second memory region is used to store application RAM data. The technical scheme of the embodiment of the present application can effectively improve the memory utilization of the operating system of the resource limited device.

The embodiment of the present application provides a method and device for managing a memory pool, a storage medium, and an electronic device, wherein the method includes: obtaining the operation requirements of a distributed storage system, and dividing the storage resources associated with the distributed storage system into large pages according to the operation requirements to obtain multiple target memory pools; wherein each of the multiple target memory pools corresponds to a class of memory object types, and the class of memory object types includes at least one memory object; recording historical data of using multiple target memory pools within a first preset period; constructing an intelligent prediction model through historical data and a preset initial prediction model, and using the intelligent prediction model and the current state data of the distributed storage system to predict the adjustment requirements within a second preset period; and dynamically managing multiple target memory pools based on the adjustment requirements. The problem of high overhead and low efficiency of memory allocation, and inability to achieve dynamic management and optimization of memory resources, is solved.

The present application discloses a computing power scheduling method and related devices, the computing power scheduling method comprising: obtaining the type of task to be created, and a first computing power card subset containing tasks of the first type in the computing power card set; wherein the type of task to be created is the first type; in response to the first computing power card subset including a target computing power card that meets the preset conditions, creating the task to be created on the target computing power card. Through this design, tasks of the same type can be created on the same computing power card as much as possible, which not only improves the memory utilization rate and CNN computing power utilization rate of the computing power card, thereby improving the remaining resource utilization rate of the computing power card, but also saves the additional memory consumption required for the first task creation.

The invention discloses a voxel map plane merging method, system and device and a storage medium, and the method comprises the steps: S1, carrying out the voxelization of a point cloud, detecting a plane in a plurality of voxels, and recording the feature information of the plane; wherein the plane is a flat surface with a certain area; s2, classifying the planes in the voxels according to the plane feature information, and calculating a hash key value; and S3, according to the position relationship between the hash key values and the planes, merging the plurality of adjacent planes. According to the method, by calculating the hash key values of the voxel planes, the voxel planes with similar geometric attributes can be quickly merged into the same hash bucket and are uniformly processed, so that the real-time performance and the resource utilization efficiency of the system are greatly improved, and particularly, when a large-scale environment is processed, the real-time performance of the system is greatly improved. The defects of a traditional fixed voxel size method can be more effectively overcome.

The invention discloses a matrix multiplication optimization method and system oriented to an RISC-V architecture, and the method comprises the steps: detecting the hardware characteristics of a target RISC-V platform, and adaptively configuring optimization parameters according to a detection result; performing memory layout and data rearrangement on the input matrix B to realize memory optimization; a multi-level partitioning strategy is adopted to adapt to a memory hierarchical structure, and multi-level partitioning calculation is achieved; carrying out RVV vectorization core calculation by using RISC-V vector instruction extension to realize data-level parallel optimization; the instruction delay is hidden through loop expansion and a software assembly line, and a data prefetching mechanism is adopted to realize instruction-level parallel optimization; performing multi-core parallel computing by using an OpenMP multi-thread technology to realize thread-level parallel optimization; and calculating residual elements by using a vector register to realize tail processing optimization. According to the method, the cache hit rate and the calculation performance degree are improved, the hardware potential is fully excavated, and the method is suitable for edge devices with high requirements.

The application discloses a model inference acceleration method, system, device and medium of a vehicle-mounted edge device, the method comprising: by analyzing the inference task request and dynamically hot loading the model weight, mapping the pre-filling stage and the decoding stage to different CUDA streams for parallel processing, and introducing a forced preemption mechanism to ensure the real-time performance of high-priority tasks. Through paging memory management and KV Cache hierarchical replacement strategy, the memory resources are optimized, and combined with power consumption monitoring and bottom layer hardware instruction set optimization, the technical problems of low resource utilization, high-priority task delay and excessive power consumption of the existing vehicle-mounted edge device in multi-task concurrency are solved. The application can effectively improve the inference throughput and system stability of the model in the vehicle-mounted edge device, solve the technical problems of low resource utilization, high-priority task delay and excessive power consumption of the existing vehicle-mounted edge device in multi-task concurrency, and is suitable for complex computing scenarios of intelligent vehicles.

The application discloses a memory management method of a real-time operating system, a real-time operating system and a storage medium, and belongs to the technical field of memory management of real-time operating systems. The technical problem to be solved is that when a memory pool is not used, memory fragmentation is high and resource utilization is low. The technical solution is as follows: whether the memory pool is used is determined according to configuration; if the memory pool is not used, the following memory allocation strategy is enabled: when an idle thread is initialized, a dynamic memory monitoring mechanism is added; the dynamic memory monitoring mechanism is used to dynamically monitor memory usage of remaining task threads other than the idle thread; and according to the memory usage of the remaining task threads other than the idle thread, memory of the corresponding threads is dynamically recycled or released to the corresponding threads.

The invention discloses a memory allocation method and system, a medium and a product, which are applied to the technical field of storage, and the method comprises the following steps: monitoring a memory access condition of a running task to obtain memory space release information of the running task; determining a memory space physical address to be addressed based on the memory space release information; addressing the memory space physical address to be addressed to obtain a continuous addressed address; and allocating the memory space for the memory space application instruction based on the continuous addressing address and the unaddressed memory space physical address. In this way, the memory utilization rate can be increased, and therefore the real-time task processing performance is improved.

The invention relates to memory allocation, in particular to an operator memory allocation method for complex structure deep learning model reasoning, and the method comprises the steps: counting the types of operators in a calculation graph; determining the memory allocation priority of each operator; determining the life cycle of each operator; selecting an operator which is not subjected to memory allocation and has the highest memory allocation priority, and performing memory allocation based on a storage structure of a deep learning acceleration kernel NNA and the life cycle of each operator; s4 is repeated until memory allocation of all operators is completed; according to the technical scheme provided by the invention, the defect that the reasoning time is difficult to effectively shorten when the method is applied to a deep learning model with a complex structure in the prior art can be effectively overcome.

This invention discloses a GPU memory optimization method and system. By acquiring dynamic data flow information and intermediate data generation patterns of the task load, the real-time variation characteristics of the task load are obtained. Based on these characteristics, a gradient descent-based optimization algorithm is used to determine the memory pool allocation ratio and reserved space, resulting in an adjusted memory pool configuration. Available memory block information is extracted from the adjusted configuration, and a Huffman coding algorithm is used to compress low-access-frequency intermediate data, resulting in compressed intermediate data storage units. If the access frequency of the compressed intermediate data storage units is lower than a preset threshold, the compressed intermediate data storage units are temporarily stored in a low-speed storage area, and memory resources are reallocated according to task priority, resulting in an optimized memory allocation scheme. This invention significantly improves memory utilization, reduces access latency, and achieves efficient memory resource management.

The invention provides an end-side cloud collaborative big data analysis method and system, and the method comprises the steps: employing the hierarchical retrieval under end-side cloud collaboration, determining a retrieval starting level based on a deployment and control target type and retrieval task timeliness, and triggering continuous retrieval according to a preset collaborative strategy: carrying out the matching of an associated camera range according to the deployment and control target type and an incident site; in the terminal layer retrieval, each associated camera performs deployment and control target retrieval on historical feature data in a local short-term time window, and triggers edge layer retrieval when the historical feature data are not hit; according to edge layer retrieval, an edge box calls historical feature data in a middle-term time window uploaded by all associated cameras in an area, deploy and control target retrieval is carried out, and cloud layer retrieval is triggered when a target is not hit; and in the cloud layer retrieval, a cloud server screens feature data in a long-term window of a retrieval area from a cloud database according to a retrieval task and performs deployment and control target retrieval. The problems of high transmission pressure, low retrieval efficiency, resource waste and the like are solved.