33results about How to "Improve data access" patented technology

A system and method for processing machine learning techniques (such as neural networks) and other non-graphics applications using a graphics processing unit (GPU) to accelerate and optimize the processing. The system and method transfers an architecture that can be used for a wide variety of machine learning techniques from the CPU to the GPU. The transfer of processing to the GPU is accomplished using several novel techniques that overcome the limitations and work well within the framework of the GPU architecture. With these limitations overcome, machine learning techniques are particularly well suited for processing on the GPU because the GPU is typically much more powerful than the typical CPU. Moreover, similar to graphics processing, processing of machine learning techniques involves problems with solving non-trivial solutions and large amounts of data.

An embodiment of the invention provides a data storage system, a data storage method and a PCM bridge. The system comprises a processor, a memory controller, the phase change memory PCM bridge and a phase change memory PCM. The processor is connected with the PCM bridge through the memory controller. The PCM bridge is connected with the PCM. Small nonvolatile data are stored in the PCM. The PCM bridge is used for receiving first operation signals which are used for reading or writing small data and are from the CPU. The PCM bridge is also used for converting the first operation signals into second operation signals capable of being used for reading or writing the PCM so as to read the small data in the PCM or write small data in the PCM through the second operation signals. According to the data storage system, the data storage method and the PCM bridge, the PCM can be used for reading or writing in a bitwise mode, so that operation efficiency is improved, system service life is prolonged, and the capacity of data access is quickened.

A control method for flash memory based on variable length ECC is provided in the present invention. A first channel of the flash memory is set to have a first ECC with a first length based on the size of data page and the length of first management data; and a second channel of the flash memory is set to have a second ECC with a second length based on the size of data page and the length of second management data. The first ECC and the second ECC are designated with different identification codes respectively, wherein the first length is shorter than the second length.

An apparatus for improving the data access reliability of flash memory is provided, including an instruction register, an address register, a flash memory control circuit, a data register, an encoder, an error correction code (ECC) generator, a signal converter, a comparator, an arbitrator, and a decoder. The instruction register and the address register are connected to a flash memory respectively for storing the access instructions and the addresses. The flash memory control circuit is connected to both instruction register and address register for controlling the access to the flash memory. The data register is connected to flash memory control circuit for loading data to be written to the flash memory. The encoder encodes the written data, and the ECC generator generates an ECC, which is written to the flash memory through the signal converter. The comparator and the arbitrator provide the comparison with ECC and informing decoder f suspicious bit values when data is read from the flash memory. The decoder is for correctly decoding data and avoiding the suspicious bit values to be read. Thus, the object improving the data access reliability of flash memory is achieved.

The invention provides a Hadoop-based distributed retrieval method for power transmission line monitoring videos. The method comprises the steps of based on a Hadoop platform, dividing massive video data storage into different video clips and storing the video clips in Blocks of Hadoop; extracting key frames under the assistance of OpenCV; performing similarity comparison on the key frames and a sample picture submitted by a user to obtain N key frames most similar to the sample picture, and obtaining the video clips corresponding to the key frames; and displaying a retrieval result in a Web interface, thereby finishing video data retrieval work. The distributed retrieval method provided by the invention has the advantages of quickness, high efficiency and expansibility in processing large-capacity data, and solves the high-performance demand problem of power transmission line monitoring video retrieval of large data volume.

The invention provides an intelligent fire inspection method. The method comprises the steps that a database is created at a server; an inspection task list is created according to location information and personnel information, wherein the inspection task list comprises at least one inspection task; task information of the inspection task list is sent to at least one mobile terminal corresponding to the inspection task list; the server receives a piece of equipment address information of the mobile terminal and sends the inspection task corresponding to the equipment address information to the mobile terminal; the server receives feedback information of the inspection task; the database is updated according to the feedback information, and the feedback information is stored. Through the intelligent fire inspection method, fire information can be managed in a unified mode, real-time updating of data is realized, and inspection personnel can be dynamically and effectively managed and supervised; the method has the advantages that information management is convenient, updating speed is high, accuracy is high, and resources are saved.

The invention discloses an efficient mass data indexing method in cloud computing. According to the method, the node splitting history is recorded by use of a log on the basis of a distributed B-tree and the distributed B-tree is efficiently and simultaneously accessed based on the node splitting history, and therefore, the accessing and indexing efficiency of mass distributed data in the cloud computing environment can be effectively improved. The efficient mass data indexing method in the cloud computing is superior to an affair method because distributed locks need to be added to all nodes in the path from the root to a leaf node to be accessed in the affair method, and the efficient mass data indexing method only needs to lock the leaf node to be accessed; the efficient mass data indexing method is superior to a link method because the number of nodes to be traversed in the efficient mass data indexing method is smaller than that of the link method.

Methods and systems for use in connection with an asymmetrical digital subscriber line (ADSL) system being available to provide voice service and data service to the customer's premises. One service of the voice service or the data service is in use at the customer's premises, and the exemplary methods and systems detect a request for provision of the other service. In response to the detection, impact information may be provided on performance of the one service in light of concurrent use of the other service. A list of modes of operation of the other service is provided. The subscriber may consider the impact information and the list of modes and respond with a selected mode of operation of the other service. The selected mode is detected and the other service is operated pursuant to the selected mode to the customer's premises.

A means of power management and control via a reverse link of a CDMA or LTE network is disclosed. The protocol is set to superimpose itself in an agnostic way to the power management as one more layer of freedom is added. The adjustment depending on the power control is then derived by mismatch if too much power or antenna tuning to enhance the power. This is also following a table in the device that will adjust ECC in case of a MIMO transmission such as LTE as well as SAR. This will generate also a mean of a soft-handoff and load control on the base station.

Systems and methods self-organize a multifunctional power and energy control and management system by integrating multiple backplane based modules through module descriptions. Dynamic data table structures may be configured based on information provides with the module descriptions and provide for improved data accessing, storing, and updating.

The invention provides a performance improvement method of a stacked DRAM cache, and provides a row buffer manager to help to shorten the access delay of the stacked DRAM cache and accelerate data access, so that the performance is improved. The method comprises the following steps that S1, a row buffer manager is provided and comprises a row state table, the row state table comprises a pluralityof row state entries, and each row state entry comprises an activation bit, a label value sequence, a label value filling bit and a waiting request number so as to describe the state of memory row data; S2, the row buffer manager is connected with the data access request queue, andinformation of a row state table is updated by the row buffer manager according to the data access request arriving inthe data access request queue; S3, the row buffer manager is connected with the cache controller, and the row buffer manager receives a command from the cache controller and updates information of the line state table; and S4, the row buffer sends a control command to the stacked DRAM cache through the cache controller according to the information of the row state table.

Methods and systems for use in connection with an asymmetrical digital subscriber line (ADSL) system being available to provide voice service and data service to the customer's premises. One service of the voice service or the data service is in use at the customer's premises, and the exemplary methods and systems detect a request for provision of the other service. In response to the detection, impact information may be provided on performance of the one service in light of concurrent use of the other service. A list of modes of operation of the other service is provided. The subscriber may consider the impact information and the list of modes and respond with a selected mode of operation of the other service. The selected mode is detected and the other service is operated pursuant to the selected mode to the customer's premises.

The invention relates to a data interaction method and device, a storage medium and electronic equipment, a data interaction instruction sent by a terminal can be received through a server, and the data interaction instruction is used for indicating that target data corresponding to an image application program on the terminal is subjected to data interaction between the image application program and a GPU; when it is determined that the data interaction instruction is an asynchronous operation instruction, first response information is sent to the terminal, the first response information is used for indicating the terminal to allocate a first memory on a memory of the terminal, and the target data is stored in the first memory; receiving the target data and a data writing instruction sent by the terminal in response to the first response information; and writing the target data into the GPU according to the data interaction instruction and the data writing instruction.

The invention belongs to the technical field of Internet of Things data access, and particularly relates to a Netty-based mass data access method and system in Internet of Things application, and the method comprises the steps: after a Netty server of an access layer node receives a terminal message, identifying the type of the message, and storing the message in a local blocking queue through thread lock-free scheduling; the exclusive consumption thread of the blocking queue pulls the message data from the queue, and if the message is core data, the thread calls a load balancing module to select the message data for forwarding; otherwise, the thread sends the message data to the kafka forwarding exclusive Topic partition through the local Kafka client; the Netty server side of the data analysis layer node receives or pulls the message data from the Kafka, converts the message data into structured data, and sends the structured data to the Kafka business exclusive Topic partition. According to the invention, the advantage that Netty can undertake hundreds of thousands of terminal communication is fully exerted, and service decoupling is realized.

The invention discloses a multi-source alarm processing method and device based on a unified alarm platform, a multi-source alarm system and related equipment. The method comprises the following steps: acquiring alarm data of a multi-source alarm event; analyzing the alarm data of the multi-source alarm event to obtain alarm information of the multi-source alarm event; and according to the alarm information and the business rule of the multi-source alarm event, assigning the multi-source alarm event to the corresponding alarm receiving and processing terminal device for processing. According to the method, the alarm data of the multi-source alarm event can be obtained. Then, the alarm data is analyzed, and alarm information can be obtained. Then, according to the alarm information and theservice rule, the multi-source alarm event is allocated to the corresponding alarm receiving seat or alarm processing seat, so that alarm receiving seat personnel or alarm processing seat personnel can correspondingly process the alarm event. The method can support access of multi-source alarm data, the alarm information response processing efficiency is improved, and the alarm response time is shortened.

A memory architecture includes K first wordlines, M groups of second wordlines, a memory cell array and M switch circuits. K and M are positive integers. Each group of second wordlines includes a plurality of second wordlines. The memory cell array includes M memory banks. The M memory banks are coupled to the M groups of second wordlines respectively, and receive independent M sets of second wordline signals through the M groups of second wordlines respectively. M switch circuits are disposed in correspondence with the M memory banks respectively. Each switch circuit selectively couples the K first wordlines to a corresponding memory bank so that the corresponding memory bank receives a shared set of first wordline signals through the K first wordline. Each memory bank performs a data access operation according to the received set of first wordline signals and a corresponding set of second wordline signals.

The invention discloses a distributed storage system and a cache layer thereof. The distributed storage system comprises the cache layer and a rear-end storage layer used for data persistent storage. The cache layer comprises a plurality of cache nodes, a plurality of object cache services OCSs run on the plurality of cache nodes, the plurality of OCSs form a plurality of cache groups, and each cache group independently processes different data. By applying the technical scheme provided by the embodiment of the invention, since no data cross exists between the cache groups, only the data consistency between the OCSs in the same cache group needs to be maintained, so that the maintenance cost of the data cache consistency is reduced, the cache data synchronization process is simpler, the system performance is improved to a certain extent. Data access can be performed in a low-delay and high-concurrency manner through each cache group, and a high-performance storage service is provided. The invention further discloses a data access method and a device applied to the distributed storage system and a computer storage medium, and the corresponding technical effects are achieved.

A memory architecture includes K first control lines, M groups of second control lines and a memory cell array. K and M are positive integers. Each group of second control lines includes at least one second control line. The memory cell array includes M memory banks. Each memory bank is coupled to the K first control lines. The M memory banks are selected according to M bank select signals respectively so as to receive a shared set of first control signals through the K first control lines. The M memory banks are coupled to the M groups of second control lines respectively, and receive independent M sets of second control signals through the M groups of second control lines respectively. Each memory bank performs one of a column select operation and a sense amplification operation according to the set of first control signals and a set of second control signals.

The invention discloses a data access method and system of a storage gateway system. The method comprises the following steps: performing metadata operation through a file protocol module in the storage gateway system and a cache file subsystem in the storage gateway system, wherein the cache file subsystem comprises a metadata cache module and a data cache module; and based on the metadata in the metadata cache module after the metadata operation, performing data access through the file protocol module and the cache file subsystem. By means of the method, the data access speed of the storage gateway can be increased.

The invention provides a method for improving the performance of a stacked DRAM cache, and proposes a line buffer manager to help shorten the access delay of the stacked DRAM cache and accelerate data access, thereby improving performance. The present invention includes the following steps: S1, propose a line buffer manager, which includes a line state table, the line state table includes a plurality of line state entries, and each line state entry includes an activation bit, a tag value sequence, a tag value filling bit and a waiting request In order to describe the state of a memory row data; S2, the row buffer manager is connected with the data access request queue, and the row buffer manager updates the information of the row status table according to the data access request arriving in the data access request queue; S3, The row buffer manager is connected to the cache controller, and the row buffer manager receives the command from the cache controller and updates the information of the row status table; S4, the row buffer sends the stacked DRAM buffer through the cache controller to the stacked DRAM cache according to the information of the row status table. control commands.

The invention relates to a coach training service system based on Web services and integrated software. The system includes a remote Web server, a remote search server, a coach information managementserver, a data encoding server, a data management operation platform, a wireless router, a data acquiring terminal and a coach identity identification card. A method is also disclosed for using the system. The method herein includes system establishment, data acquiring setting and data management service. According to the invention, on one hand, the system and the method can identify and process all kinds of identity information and text and graphics information, and on the other hand, the system and the method can effectively make the inputting and acquiring work of identity identification data more easily and flexible, and effectively increase the need for performing data access and modifying work in a remote manner based on the identity identification card.

The invention discloses a temperature data generation method and device and a storage medium, and belongs to the technical field of photovoltaic power generation, and the method comprises the steps: determining the hour-level monthly average temperature in a preset time unit of each day based on the monthly average temperature and irradiation data of a preset reference region, the hour-level highest temperature and the hour-level lowest temperature are determined from the hour-level monthly average temperature; determining a daily maximum average temperature and a daily minimum average temperature based on the monthly average temperature, the hour-level maximum temperature, the hour-level minimum temperature and the irradiation data; and calculating the daily hour-level temperature based on the daily maximum average temperature, the daily minimum average temperature and the irradiation data. The hour-level temperature data can be obtained only through monthly average temperature and irradiation data, so that the problems that photovoltaic power generation simulation evaluation is not accurate enough and the efficiency is low due to the fact that the hour-level temperature data is difficult to collect and obtain are solved.

The invention provides a Hadoop-based distributed retrieval method for power transmission line monitoring videos. The method comprises the steps of based on a Hadoop platform, dividing massive video data storage into different video clips and storing the video clips in Blocks of Hadoop; extracting key frames under the assistance of OpenCV; performing similarity comparison on the key frames and a sample picture submitted by a user to obtain N key frames most similar to the sample picture, and obtaining the video clips corresponding to the key frames; and displaying a retrieval result in a Web interface, thereby finishing video data retrieval work. The distributed retrieval method provided by the invention has the advantages of quickness, high efficiency and expansibility in processing large-capacity data, and solves the high-performance demand problem of power transmission line monitoring video retrieval of large data volume.

The invention relates to a student card consulting service system based on Web services and integrated software. The system comprises a remote Web server, a remote retrieval server, a student information management server, a communication protocol server, a data management and manipulation platform and a data acquisition terminal, wherein the data management and manipulation platform is interconnected with the communication protocol server through a communication network, and is respectively intercommunicated with the remote Web server, the remote retrieval server and the student information management server through the communication protocol server. An application method includes the following three steps: system construction, data acquisition setting and data management services. According to the scheme of the invention, on the one hand, the identification and processing of various types of identification information, graphic and text information and other information can be effectively met, and on the other hand, the convenience and flexibility of identification data entry and collection operations can be effectively achieved, and the needs for directly and remotely performingdata access and modification operations based on an identification card can be effectively improved.

The invention provides an elastic Key-Value key-value pair data storage method based on RDMA and HTM, the design follows the server-client architecture mode, including: on the server side, an improved G-Cuckoo is provided in combination with a hash graph Hash data management mode, avoiding the problem of kick-out infinite loop between hash tables caused by searching for free positions during data insertion; analyzing the traditional network transmission of messages between the client and the server that requires back and forth round responses Key-value pair storage performance bottleneck problem, use Infiniband remote direct memory access RDMA technology, design RDMA network communication engine, receive data access requests and send back data request results; use hardware transactional memory HTM technology, realize two-stage lock protocol lock operation, guarantee Atomic characteristics of data operations; using key-value pairs for data self-verification check codes to ensure data consistency. The implementation of the present invention can greatly improve the basic operation speed of key-value pair data.

Embodiments of the present invention provide a data storage system, a data storage method, and a PCM bridge. The system includes: a processor, a memory controller, a phase-change memory PCM bridge, and a phase-change memory PCM. The processor is connected to the PCM bridge through the memory controller. , the PCM bridge is connected to the PCM, and non-volatile small data is stored in the PCM, and the PCM bridge is used to receive the first operation signal for reading or writing small data from the CPU through the memory controller, and convert the first operation signal A second operation signal capable of reading or writing to the PCM is formed to read small data in the PCM or write small data in the PCM through the second operation signal. The data storage system, the data storage method and the PCM bridge provided by the embodiments of the present invention can be read or written bit by bit by using the PCM, which improves the operation efficiency, increases the life of the system, and accelerates the ability of data access.