58results about How to "Reduce access latency" patented technology

The invention relates to the technical field of computer storage, and discloses a computer memory data processing system, which comprises the following modules: an intelligent cache prefetching module, which is used for predicting a future memory access address based on deep learning and reinforcement learning technologies and loading data into a cache in advance; the heterogeneous storage management module is used for dynamically classifying the data based on the data popularity value and realizing migration of the data among different storage media; and the storage consistency guarantee module is used for managing the data consistency among different storage layers through the improved MES I protocol. Cache prefetching is optimized by introducing deep learning and reinforcement learning technologies, and by predicting a future memory access address in real time and loading data in advance, the system effectively reduces the cache missing rate, improves the cache hit rate, remarkably reduces the memory access delay, and improves the memory access efficiency. And the system can keep more efficient data access performance in a large-scale data processing scene.

The invention provides a many-core cache consistency system, a many-core cache consistency method, electronic equipment, a storage medium and a product, and the system is based on a network-on-chip architecture, and a plurality of nodes, corresponding routers and distributed directories deployed in the routers are introduced. Each node comprises one or more processor cores and a distributed last-stage cache, and a memory address is divided into a plurality of address blocks and is associated with the distributed last-stage cache through a dynamic address mapping rule. The distributed directories and the distributed last-stage caches are in one-to-one correspondence and used for recording the states of cache lines, the router is responsible for receiving and responding to requests from local or remote nodes, querying the distributed directories and updating the states of the cache lines, and efficient management of cache consistency is achieved. The technical scheme belongs to the technical field of communication, effectively solves the problems of high load and communication delay of a main node in a traditional cache consistency protocol, and improves data consistency and access efficiency in a multiprocessor system and a large-scale parallel computing environment.

The invention relates to the technical field of data storage management, in particular to a data storage method based on a large model. The method comprises the following steps: performing input class cluster fragmentation optimization on a target pre-training large model to obtain storage modulus fragmentation data; constructing an intelligent fragmentation storage strategy by using the storage modulus fragmentation data; obtaining target model training data; performing intelligent fragmentation storage on the target model training data by using an intelligent fragmentation storage strategy, and performing redundancy compression optimization to obtain dynamic redundancy storage index data; and training data storage space optimization is carried out according to the dynamic redundancy storage index data, and an optimized data storage strategy is obtained. According to the method, intelligent fragmentation storage and refined data redundancy compression are performed on the data, so that efficient and stable model training data access is realized, the problem of low data storage and access efficiency in a large model training process is solved, and the model training efficiency and reasoning real-time performance are remarkably improved.

The invention discloses a data storage system of a digital country based on cloud computing, which comprises a client, a data transmission module and a cloud computing center, relates to the technical field of data storage, stores hot data on a high-performance hot data storage layer, ensures that frequently accessed data can be quickly responded, improves the system performance, and at the same time, stores the hot data on a high-performance hot data storage layer. The cold data is stored in a low-performance cold data storage layer, the storage cost is reduced, effective balance of performance and cost is achieved, the data is automatically migrated among different storage layers along with change of data access frequency, so that the performance and the cost are balanced, the frequently accessed data is migrated into a cache, the access speed of the data is increased, and the data access efficiency is improved. And establishing a relationship between a time interval and a data type of historical data, predicting data which are possibly accessed in a given time interval, adding the data into the cache, dynamically adjusting the cache, and ensuring the validity of the cache.

The invention discloses a cache management method, electronic equipment, a storage medium and a program product, and relates to the technical field of data storage, and the method comprises the following steps: obtaining current operation data of a storage system in a current period, determining current access popularity corresponding to a plurality of data blocks according to the current operation data, and storing the data blocks according to the current access popularity. The resource allocation of the target cache is dynamically adjusted, the problem that the data access efficiency of a storage system in the related technology is low is solved, and the technical effect of improving the data access efficiency of the storage system is achieved.

The invention provides a CDN (Content Delivery Network)-based file resource management system and method. The file resource management method comprises the following steps: carrying out network configuration of a CDN (Content Delivery Network) according to a resource management demand set by a user; the CDN screens target edge nodes corresponding to the users according to the file resource uploading parameters of the users, and determines the target edge nodes corresponding to the users; the CDN network sends and caches the content information uploaded in real time of the file resource to a target edge node in the CDN network, and the target edge node performs data sharing with other edge nodes of the CDN network; and monitoring the operation parameters of the CDN network in real time, and giving an abnormity alarm when the CDN network is abnormal. The system comprises modules corresponding to the steps of the method.

The invention discloses a multi-level cache system and method based on REDIS. The multi-level cache system comprises N Node nodes, a plurality of application programs and proxy services are deployed on each Node node, and a first-level cache is configured on a driving layer closest to the application programs; simultaneously deploying a proxy cache database proxy service on a server where the application program is located, and configuring a second-level cache; and finally, the cache database Redis itself is used as a third-level cache. According to the method, by reasonably distributing cache responsibilities of different levels, the read load of the Redis database is reduced, the cache hit rate is increased, the network I / O overhead is reduced, and the overall performance and stability of the system are improved.

The invention discloses a cache request processing method, equipment and a medium, belongs to the technical field of data processing, and is used for solving the technical problem that in the prior art, after the size of a cache line is reduced, a system needs to process a cache request more frequently, and the complexity and overhead of cache management are increased. The method comprises the steps that n block requests are combined into a total request, and n is a positive integer; determining whether a target data block corresponding to any one request in the total request is cached and hit; if the cache of the target data block is hit, obtaining cache data corresponding to the target data block from a cache; and if the target data block cache is missed, sending a missed request to the L2 cache, distributing a target cache space for the target data block in the L1 cache, and obtaining corresponding cache data from the target cache space. According to the method, the size of the cache line is reduced, the frequency of processing the cache request is reduced, and the complexity and overhead of cache management are reduced.

The embodiment of the invention provides a processor cache structure, a processor and a data caching method, and relates to the technical field of computers, the processor cache structure comprises a first-level data cache, a first-level instruction cache, a second-level private cache, a second-level private prefetching cache and a cache consistency control unit; the primary data cache is used for storing data accessed by the current core of the processor for multiple times; the first-level instruction cache is used for storing instructions executed by the current core for multiple times; the second-level private cache is used for storing data read by the current core from the main memory or the high-level cache; the second-stage private prefetch cache is used for calculating a prefetch address by adopting a preset algorithm and sending a prefetch request to the second-stage private cache, so that the second-stage private cache prefetches a cache line according to the prefetch address; and the cache consistency control unit is used for monitoring cache operation of each core of the processor. The scheme does not need to wait for other cores to release cache resources, so that the data access delay is remarkably reduced, and the overall performance of the processor is improved.

The invention discloses a CXL control system based on an fpga pcie controller and application, and relates to the technical field of control systems. The PCIe PHY module comprises a PCIe PHY module, a cxl arb mux layer, a cxl io ctrl layer, a cxl mem / cache layer, a DDR controller module, an SRAM (Static Random Access Memory) module and a clkrstngen module. According to the method, the conflict between logic resources and a time sequence is effectively solved, and on the basis of realizing high bandwidth, the CXL access delay is effectively reduced by optimizing the CXL function and the time sequence, so that various CXL application scenes are effectively realized; the programmable characteristic of the FPGA is fully utilized, Flex Bus logic resources are optimized, and the utilization rate of the logic resources is reduced, so that the fpga time sequence requirement is met, DDR expansion and SRAM expansion are achieved, and the speed and function requirements of different application scenes are met.

The invention relates to the technical field of distributed storage, and discloses a distributed storage system with infinitely extensible capacity and performance and a storage method thereof.The distributed storage system comprises a special storage device cluster and further comprises a mobile device node pool composed of mobile devices in a company internal local area network; the dynamic node management and self-adaptive expansion mechanism is located in the data scheduling and management module and is closely integrated with the data scheduling and management module; and a data transmission protocol and an encryption mechanism. Through the synergistic effect of the data caching module and the preloading engine, the system significantly reduces the direct access to the special storage device cluster, and optimizes the data flow path. The data scheduling and management module dynamically adjusts a cache strategy, so that common data can be quickly read from a cache, and meanwhile, a preloading engine loads possibly needed data in advance according to a historical access mode. According to the mechanism, the data access delay is reduced, and the response speed and the overall performance of the system are greatly improved especially in the scene of high-concurrency requests.

The invention discloses a solid state disk storage wear leveling control method and device, and relates to the field of solid state disk wear leveling control, and the method comprises the steps: screening out a plurality of first target blocks which are not accessed in a solid state disk within a preset time threshold value, and determining the corresponding access times; obtaining storage data and residual capacity of the first target block with the access times smaller than a first preset threshold value; integrating all the target blocks according to the storage data and the residual capacity to obtain a plurality of second target blocks; and migrating the data in the second target block to a third target block of which the access frequency is between the first preset threshold value and the second preset threshold value. By implementing the method, data storage can be dynamically adjusted according to the access frequency of the data block, cold data which is not accessed for a long time is integrated and migrated, the problem that local abrasion is aggravated due to the access frequency difference is solved, and the service life of the solid state disk is prolonged.

This invention discloses a cache adjustment method, apparatus, device, and computer-readable storage medium, comprising: collecting access characteristic data from the client layer, object storage daemon layer, and device layer respectively; determining the object access mode of the system based on the access characteristic data; and adjusting the cache priority of each access task, the cache capacity ratio between the client layer and the object storage daemon layer, and the data residence time in the cache layer according to the object access mode; wherein the cache layer includes the client layer and the object storage daemon layer. This invention improves cache hit rate and overall read / write throughput performance, and reduces access latency.

The invention discloses a remote storage access method and device, a medium and equipment. The method comprises the following steps: a computing server sends a remote memory access request through an internet access; the remote memory access request reaches the network switch through the network link; the network switch sends the remote memory access request to a specified storage server; and the storage server bypasses a central processing unit of the storage server to directly read and write the solid-state memory according to the remote memory access request.

The invention provides a cloud-side collaborative content caching and user association method and system, and relates to the technical field of wireless communication networks, and the method comprises the steps: building a weighted delay minimization problem with minimization of user content acquisition delay and association base station switching delay as a target; and dividing the constructed weighted delay minimization problem into a long-term user association and content caching problem under a long-term time scale and a short-term user association problem under a short-term time scale to obtain a content caching strategy and a user association strategy so as to minimize the content transmission delay between the user and the base station. According to the invention, the content transmission delay can be reduced on the basis of reducing the backhaul link burden, and the network performance is improved.

The invention discloses a storage-level memory, a data processing method and a processor system, and relates to the field of computers. A plurality of memory particles included in the storage class memory are divided into at least one group. Each group comprises a first type of memory particles and a second type of memory particles. The second type of memory particles are used for storing operation error correction codes, and the first type of memory particles are used for storing data and re-error correction codes. The operation error correction codes are used for performing first-level memory error correction on the data stored in the first-class memory particles in the same group. The re-error correction code is used for performing secondary memory error correction on the data stored in the first type of memory particles when the primary memory error correction fails. And the access delay is as low as possible when the high reliability of the data stored in the storage-level memory is ensured through a two-stage memory error correction mechanism. Besides, the memory particles of the storage-level memory are arranged, so that the bit width of the storage-level memory meets the memory bit width indicated by the DDR protocol, and the storage-level memory is compatible with the DDR protocol.

This application proposes a method, apparatus, electronic device, and storage medium for arranging vector elements. The method includes: acquiring vector elements to be processed, and determining the levels of each switching node (SN) unit in a butterfly network based on the number of vector elements to generate the butterfly network; wherein the butterfly network is used to arrange the vector elements; determining the input-output mapping relationship between the levels of each switching node (SN) unit in the butterfly network; determining the target state of the SN unit in each level based on the input-output mapping relationship; arranging the vector elements based on the target state, and performing data processing based on the arranged vector elements. Therefore, this scheme enables the arranged vector elements to be transmitted non-blockingly in the butterfly network, thereby enabling the mapping of complex and variable large model workloads onto vector hardware to achieve efficient data processing execution.

The application relates to the technical field of vehicles, in particular to a vehicle-mounted video sharing method based on a PCIe bus, an automobile central platform, an electronic device and a computer readable storage medium. The method comprises the following steps: a plurality of system-level chips are connected in communication through a PCIe switch; in each system-level chip, an independent shared memory space is respectively allocated for each camera; when any system-level chip acquires video data, the video data is stored in the local shared memory space, and the video data is transmitted to a PCIe domain space through direct memory access after address mapping; other system-level chips acquire the video data from the PCIe domain space and store the video data in the local shared memory space; and in response to a reading request of the video data of an application program running on any system-level chip, the video data is read from the local shared memory space by the application program.

The present invention discloses a method and system for loading and running program contents from server memory on demand, wherein the client intercepts the page fault interruption caused by the mapping of the target program, and the mapping has been established before running to the target program entry address; the client sends the request for the page fault interruption to the server, and the request includes the name of the process where the page fault occurs and the address of the current page fault; the server obtains the corresponding page from its memory according to the request, and returns the page to the client. Compared with the prior art, the present invention controls the client to load the code and data required for program operation from the server memory into the local memory on demand, and the whole process does not need to consume local storage resources, and directly obtains the required code and data from the server's memory, reducing data access delay.

The application discloses a file processing system based on Internet of Things, and relates to the technical field of Internet of Things, which comprises a processing carrier and a plurality of terminals, the processing carrier is used for sending protocols to each terminal, and sharing data and files within a specified range; after accepting the protocols, the plurality of terminals mirror backup internal files and data, each terminal acquires a plurality of sets of marking information, each set of information represents the file and data classification of the terminal, and each terminal marks each working state of itself to acquire working information Z; each terminal comprises a processing module, a receiving module and a mutual inductance module. When working, each terminal in the application collects data or files generated during its own work, classifies and assigns weights to the data or files, and transmits the files and their attached information between the remaining terminals and the processing carrier, so that the working state change speed between different terminals can be improved.

Embodiments of this application provide an information transmission method and an apparatus. The method may include: A terminal device receives area identification information from a network device through a first frequency resource; and the terminal device performs an area update procedure or receives system information based on the area identification information through a second frequency resource. According to this application, transmitting common information on a WUR link can prolong working duration of the terminal device on the WUR link, improve energy saving benefit, and reduce an access delay of the terminal device. The method provided in embodiments may be applied to communication systems such as 5G or NR, LTE, V2X, D2D, M2M, MTC, and the internet of things.

The embodiment of the invention relates to the technical field of computer data processing, in particular to a floating point variable processing method and device, computer equipment and a storage medium, and the floating point variable processing method comprises the steps: obtaining a to-be-processed floating point variable from a coprocessor based on a preset floating point instruction, and converting the floating point variable into a fixed point variable; mapping the to-be-processed floating point variable to a corresponding coordinate interval in the preset coordinate space according to the quadrant of the to-be-processed floating point variable in the preset coordinate space, and determining the initial direction rotation amount of the fixed point variable in the rotation branch and the initial plane branch movement amount corresponding to the plane branch; performing iterative calculation on the fixed-point variables based on the initial direction rotation amount and the initial plane branch movement amount until iteration meets calculation requirements, and obtaining each output fixed-point amount; and performing reflection on each output fixed-point quantity based on the coordinate interval, converting the output fixed-point quantity into a target floating-point quantity and writing the target floating-point quantity back to the target module. And the floating point variable processing method with higher efficiency is provided.

This disclosure relates to the field of semiconductor technology, providing a semiconductor device to address the technical problem that high-bandwidth memory technology still has significant room for improvement. The semiconductor device includes a chip stack structure, comprising: a first semiconductor chip and a second semiconductor chip stack structure stacked along a first direction; the first semiconductor chip includes a controller, a computing module, and an interface module; the controller is configured to: perform a first access operation on stored data in the second semiconductor chip stack structure and send the stored data to the interface module to allow external devices to access the stored data in the second semiconductor chip stack structure through the interface module; and perform a second access operation on the stored data in the second semiconductor chip stack structure and send the stored data to the computing module to allow the computing module to perform computational operations on the stored data. This improves integration and shortens data transmission paths, reducing bandwidth consumption and energy loss.

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.

The present invention provides a method for accelerating hotspot data in a database and an electronic device, which are applied to the fields of computer technology and database technology. The acceleration method includes: in response to detecting a data write operation, determining the incremental data of the target row according to the pre-write data and post-write data of the target row targeted by the data write operation; when it is determined that the registered data of the target row already exists in the row buffer pool, writing the incremental data into the row version chain corresponding to the target row in the incremental buffer pool; in response to determining that the target row is hotspot data, when the complete row data of the target row does not exist in the row buffer pool, writing the post-write data as the complete row data into the row buffer pool, so as to calculate the version data of the target row to be read according to the row version chain and the complete row data in response to a read request. Therefore, the present invention can at least solve the technical problem of obvious access delay and achieve the technical effect of accelerating hotspot data and reducing access delay.

The invention is applicable to the technical field of Internet, and provides an Internet access method and an Internet access system, comprising the following steps: receiving an information access instruction, the information access instruction comprising access time, an access account and an access address; calling user historical data according to the access account and the access address, and determining a plurality of recommended access addresses according to the user historical data and the access big data; obtaining a predicted access track according to the access time, the access account and the access address, and caching data to be accessed in advance according to the predicted access track; determining a real access track of the user, and updating the predicted access track based on the real access track; and based on the real access condition of the pre-cached data, the pre-cached data is cleared in a differentiated manner. According to the method, the data to be accessed are cached in advance according to the predicted access trajectory, and the predictive caching strategy based on the user access behavior can significantly improve the cache hit rate and reduce the access delay.

The present application discloses a method, device, equipment and storage medium for accessing a streaming media server, including: obtaining a first access request from a client located in an edge area and sending the first access request to a first load balancer located in a central area through a proxy server located in the edge area, so that after receiving the first access request, the first load balancer allocates an edge streaming media server in the edge area to the client according to the correspondence between the proxy server and the edge area; wherein the edge area is a remote area of ​​the central area; and accessing the source station streaming media server in the central area by controlling the client to access the edge streaming media server in the edge area to which it is allocated. It can be seen that the present application enables the client to access the streaming media server nearby, thereby reducing the export bandwidth, reducing the access delay and improving the access efficiency.

The application realizes a satellite cloud-oriented distributed file system through a method in the field of distributed systems.The system is composed of five components: a metadata management module, a StorageEngine, a Policy, a client and a satellite topology service module; the metadata management module stores metadata, the Storage Engine is a data storage module, the Policy is a strategy module, the user uses the client to interact with the satellite cloud file system, the satellite topology service module is specially designed for the satellite cloud scene to provide the satellite cloud network topology information for the satellite cloud file system, and the modules communicate with each other through a special file system interface using RPC.The method provided by the application guarantees efficient reading and reliability of metadata for the satellite communication scene, maximally reduces the communication cost and energy loss of replica placement by using the network topology characteristics of the satellite cloud, and maximally reduces the access delay.

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.

A method for remote storage of node metadata partition, which binds independent logical acceleration pool for object storage device process based on mechanical hard disk to remotely store metadata. In the formatting stage, a globally unique identifier is generated, virtual block devices are created for database and log partition respectively, and acceleration pool is connected; continuous address space is divided into discrete storage objects according to preset object size, and a mapping relationship between logical offset and object and object internal offset is established; database label and log clipping record are written into corresponding objects of acceleration pool. When a write request is received, log entries are written into log storage objects according to the mapping, and after the memory table is flushed as a sorted string table, the database storage objects are written according to the mapping; when a read request is received, data is read from the acceleration pool according to the mapping relationship, and pre-reading cache is used for acceleration. The method can eliminate the single point failure risk of local high-speed storage medium, realize the pooling sharing and elastic expansion of storage resources, and guarantee the metadata access performance.