利索-全球专利检索

  1. 登录
  2. 注册

搜索结果

22 条记录 (耗时 0.012 秒)

    METHOD AND APPARATUS FOR DATA CACHING
    1.
    发明公开
    METHOD AND APPARATUS FOR DATA CACHING 审中-公开

    公开(公告)号：EP4012546A1

    公开(公告)日：2022-06-15

    申请号：EP19944564.4

    申请日：2019-11-28

    申请人： Inspur Electronic Information Industry Co., Ltd

    发明人： LIU, Haiwei ,  DONG, Gang ,  YANG, Hongbin ,  ZHAO, Yaqian ,  LI, Rengang ,  SHI, Hongzhi

    IPC分类号： G06F3/06 ,  G06N3/04

    摘要： The present invention provides a method and apparatus for data caching. The method comprises: output matrixes are acquired one by one, a plurality of acquired output matrixes are written alternately into two queue sets of a first cache unit according to a sequence in which the output matrixes are acquired, and the output matrixes stored line by line in a first cache unit are written into a second cache unit one by one, according to the sequence in which the output matrixes are written into the second cache unit, valid data of each output matrix of the second cache unit is determined one by one according to preset parameters, and the valid data of each output matrix is written into a third cache unit, and the valid data of the output matrixes stored in the third cache unit are configured to be sequentially written into a memory according to a sequence in which the valid data are written into the third cache unit. In the present solution, the output matrixes are cached by using cache units with the writing speed matching with the computing speed of a processor, and the output matrixes are completely written into a memory one by one according to a sequence of generation time. Therefore, the present invention may solve the problem that the computing speed of the processor does not match with the writing speed of the memory.

    METHOD FOR IMPLEMENTING PACKET SEARCHING BY USING TWO SLIDING WINDOWS
    2.
    发明授权
    METHOD FOR IMPLEMENTING PACKET SEARCHING BY USING TWO SLIDING WINDOWS 有权
    通过使用两个滑动窗口来实现分组搜索的方法

    公开(公告)号：EP2899926B1

    公开(公告)日：2017-11-29

    申请号：EP14754402.7

    申请日：2014-01-23

    申请人： Inspur Electronic Information Industry Co., Ltd.

    发明人： WANG, Endong ,  HU, Leijun ,  LI, Rengang

    IPC分类号： H04L12/70 ,  H04L12/935 ,  H04L12/931 ,  H04L12/46 ,  H04L12/26

    CPC分类号： H04L43/50 ,  H04L12/4625 ,  H04L49/30 ,  H04L49/40

    摘要： A method of implementing packet search by double sliding windows is provided. The method adopts a three-level barrel shift register to store input packet data, and a position of a sliding window 1 is determined at 32 positions by primary testing of a link, so as to ensure that the packet data is located at the center of the sliding window 1, thereby ensuring that the position of the sliding window 1 meets a transmission characteristic of a specific link to the maximum extent. After the position of the sliding window 1 is determined, 32-bit packet data can be effectively searched in the sliding window 1 by dynamically adjusting a sliding window 2, and 32-bit transmission offset is allowed for the packet data. The method of implementing packet search by double sliding windows meets a transmission characteristic of a specific link to the maximum extent.

    NEURAL NETWORK MODEL FOR IMAGE SEGMENTATION AND IMAGE SEGMENTATION METHOD THEREFOR
    3.
    发明公开
    NEURAL NETWORK MODEL FOR IMAGE SEGMENTATION AND IMAGE SEGMENTATION METHOD THEREFOR 审中-公开

    公开(公告)号：EP4053739A1

    公开(公告)日：2022-09-07

    申请号：EP20907894.8

    申请日：2020-08-25

    申请人： Inspur Electronic Information Industry Co., Ltd

    发明人： WANG, Li ,  GUO, Zhenhua ,  ZHAO, Yaqian

    IPC分类号： G06K9/34 ,  G06K9/46 ,  G06N3/04

    摘要： A neural network model for image segmentation, an image segmentation method therefor and device thereof, and a readable storage medium. The model comprises an intelligent selection module, and the intelligent selection module further comprises a feature extraction unit and an intelligent selection unit. Since the feature extraction unit uses multi-scale dilated convolution to obtain information of different scales of an input feature map, a large quantity of diverse feature information is provided for later feature screening. In addition, the intelligent selection unit trains a weight value, and performs intelligent screening on an input feature map channel according to the size of the weight value. Therefore, the intelligent selection module can ensure segmentation accuracy while reducing the number of parameters and the amount of calculation. Therefore, by using the described intelligent selection module, the neural network model of the present application can quickly extract an effective feature of an image; moreover, the amount of calculation is small and model parameters are few, and the model is applicable to a mobile terminal.

    EXTENDED CACHE COHERENCE PROTOCOL-BASED MULTI-LEVEL COHERENCE DOMAIN SIMULATION VERIFICATION AND TEST METHOD
    6.
    发明公开
    EXTENDED CACHE COHERENCE PROTOCOL-BASED MULTI-LEVEL COHERENCE DOMAIN SIMULATION VERIFICATION AND TEST METHOD 审中-公开
    测试和测试PROCEDURE ON与多级一致性EXTENDED缓存的一致性基于域仿真甲LOG

    公开(公告)号：EP2869194A1

    公开(公告)日：2015-05-06

    申请号：EP13872015.6

    申请日：2013-10-11

    申请人： Inspur Electronic Information Industry Co., Ltd.

    发明人： WANG, Endong ,  HU, Leijun ,  CHEN, Jicheng ,  ZHANG, Feng ,  ZHOU, Hengzhao ,  FU, Yunyue ,  GAN, Xiaowei

    IPC分类号： G06F9/455

    CPC分类号： G06F11/3608 ,  G06F11/3664 ,  G06F12/0815 ,  G06F12/0837 ,  G06F17/5009 ,  G06F2212/2542 ,  G06F2212/621

    摘要： An extension Cache Coherence protocol-based multi-level coherency domain simulation verification and test method. An extension Cache Coherence protocol-based multi-level coherency domain CC-NUMA (Cache Coherent Non-Uniform Memory Access) system protocol simulation model is built, a protocol table inquiring and state converting executing mechanism in a key node of a system ensures that a Cache Coherence protocol is maintained in a single computing domain and is simultaneously maintained among a plurality of computing domains, and accuracy and stability of intra-domain and inter-domain transmission are ensured; a credible protocol inlet conversion coverage rate evaluation driven verification method is provided, transactions are processed by loading an optimized transaction generator push model, a coverage rate index is obtained after the operation is ended, and the verification efficiency is increased in comparison with a random transaction promoting mechanism. Through building a multi-processor multi-level coherency domain verification system model and performing relevant simulation verification, the applicability and the effectiveness of the method are further confirmed.

    COMPUTING METHOD AND APPARATUS FOR CONVOLUTIONAL NEURAL NETWORK MODEL
    7.
    发明公开
    COMPUTING METHOD AND APPARATUS FOR CONVOLUTIONAL NEURAL NETWORK MODEL 审中-公开

    公开(公告)号：EP4024286A1

    公开(公告)日：2022-07-06

    申请号：EP19949969.0

    申请日：2019-11-27

    申请人： Inspur Electronic Information Industry Co., Ltd

    发明人： GUO, Zhenhua ,  FAN, Baoyu ,  WANG, Li ,  GAO, Kai

    IPC分类号： G06N3/063 ,  G06N3/04 ,  G06N3/08

    摘要： A computing method and apparatus for a convolutional neural network model. The method comprises: acquiring a computing model of a training task of a convolutional neural network model (S101); then splitting multiply-accumulate operation in a computing model of a training task of the convolutional neural network model into a plurality of multiply-add operation tasks (SI02); confirming a computing device corresponding to each multiply-add operation task according to the correlation between a preset computing model and the computing device (S103); and finally, respectively computing each multiply-add operation task by utilizing the computing device corresponding to each multiply-add operation task (S104). The purposes of improving the flexibility of migration of a CNN model training task on different computing devices or cooperative computing of different processors and improving the computing speed are achieved.

    METHOD FOR PARTIAL CONSTRUCTION OF SHARE-F STATE IN MULTILEVEL CACHE COHERENCY DOMAIN SYSTEM
    10.
    发明公开
    METHOD FOR PARTIAL CONSTRUCTION OF SHARE-F STATE IN MULTILEVEL CACHE COHERENCY DOMAIN SYSTEM 审中-公开
    VERFAHREN ZUR PARTIELLEN KONSTRUKING EINES分享F-STATUS在EINEM MEHRSCHICHTIGENDOMÄNENSYSTEMMIT CACHE-KOHÄRENZ

    公开(公告)号：EP2871579A1

    公开(公告)日：2015-05-13

    申请号：EP13878773.4

    申请日：2013-10-11

    申请人： Inspur Electronic Information Industry Co., Ltd.

    发明人： WANG, Endong ,  CHEN, Jicheng ,  HU, Leijun ,  GAN, Xiaowei ,  GONG, Weifeng

    IPC分类号： G06F12/08

    CPC分类号： G06F12/0804 ,  G06F12/0815 ,  G06F12/0817 ,  G06F12/0828 ,  G06F12/0897 ,  G06F2212/1016 ,  G06F2212/604

    摘要： A method of constructing a Share-F state in a local domain of a multi-level cache coherency domain system, includes: 1) when it is requested to access S state remote data at the same address, determining an accessed data copy by inquiring a remote proxy directory RDIR, and determining whether the data copy is in an inter-node S state and an intra-node F state; 2) directly forwarding the data copy to a requester, and recording the data copy of the current requester as an inter-node Cache coherency domain S state and an intra-node Cache coherency domain F state; and 3) after data forwarding is completed, recording, in a remote data directory RDIR, an intra-node processor losing an F permission state as the inter-node Cache coherency domain S state and the intra-node Cache coherency domain F state.

    摘要翻译： 一种在多级高速缓存一致性域系统的本地域中构建Share-F状态的方法，包括：1）当请求访问同一地址的S状态远程数据时，通过查询 远程代理目录RDIR，并且确定数据副本是否处于节点间S状态和节点内状态; 2）直接将数据拷贝转发给请求者，并将当前请求者的数据副本记录为节点间高速缓存一致性域S状态和节点内高速缓存一致性域F状态; 3）在完成数据转发之后，在远程数据目录RDIR中记录节点间处理器作为节点间高速缓存一致性域S状态和节点间高速缓存一致性域F状态而丢失F允许状态。