公开(公告)号：US20210157734A1

公开(公告)日：2021-05-27

申请号：US16953242

申请日：2020-11-19

Applicant: ELECTRONICS AND TELECOMMUNICATIONS RESEARCH INSTITUTE

Inventor： Byung Jo KIM ,  Joo Hyun LEE ,  Seong Min KIM ,  Ju-Yeob KIM ,  Jin Kyu KIM ,  Mi Young LEE

IPC: G06F12/0862 ,  G06F12/02 ,  G06F13/16 ,  G06N3/063

Abstract: A method for controlling a memory from which data is transferred to a neural network processor and an apparatus thereof are provided, the method including: generating prefetch information of data by using a blob descriptor and a reference prediction table after history information is input; reading the data in the memory based on the pre-fetch information and temporarily archiving read data in a prefetch buffer; and accessing next data in the memory based on the prefetch information and temporarily archiving the next data in the prefetch buffer after the data is transferred to the neural network from the prefetch buffer.

公开(公告)号：US20200151568A1

公开(公告)日：2020-05-14

申请号：US16541275

申请日：2019-08-15

Applicant: ELECTRONICS AND TELECOMMUNICATIONS RESEARCH INSTITUTE

Inventor： Mi Young LEE ,  Byung Jo KIM ,  Seong Min KIM ,  Ju-Yeob KIM ,  Jin Kyu KIM ,  Joo Hyun LEE

IPC: G06N3/08 ,  G06F17/18

Abstract: An embodiment of the present invention provides a quantization method for weights of a plurality of batch normalization layers, including: receiving a plurality of previously learned first weights of the plurality of batch normalization layers; obtaining first distribution information of the plurality of first weights; performing a first quantization on the plurality of first weights using the first distribution information to obtain a plurality of second weights; obtaining second distribution information of the plurality of second weights; and performing a second quantization on the plurality of second weights using the second distribution information to obtain a plurality of final weights, and thereby reducing an error that may occur when quantizing the weight of the batch normalization layer.

公开(公告)号：US20220180152A1

公开(公告)日：2022-06-09

申请号：US17545472

申请日：2021-12-08

Applicant: Electronics and Telecommunications Research Institute

Inventor： Mi Young LEE

IPC: G06N3/04 ,  G06N3/08

Abstract: Disclosed is an operating method of a neural network processing device that communicates with an external memory device and executes a plurality of layers including obtaining layer information of the plurality of layers by analyzing a connection structure of the plurality of layers, generating an input address and an output address for a target layer based on the layer information, receiving expected input data and expected output data for the target layer, storing the expected input data at an input address area of the external memory device corresponding the input address, storing output result data at an output address area of the external memory device corresponding the output address by executing the target layer, comparing the output result data with the expected output data, and determining whether an error for the target layer occurs.

公开(公告)号：US20180157729A1

公开(公告)日：2018-06-07

申请号：US15631774

申请日：2017-06-23

Applicant: ELECTRONICS AND TELECOMMUNICATIONS RESEARCH INSTITUTE

Inventor： Mi Young LEE

IPC: G06F17/30 ,  G06F9/50 ,  H04L29/08

CPC classification number: G06F16/27 ,  G06F9/505 ,  H04L29/08018 ,  H04L67/1029

Abstract: Disclosed herein is a distributed in-memory database system for partitioning a database and allocating the partitioned database to a plurality of distributed nodes, wherein at least one of the plurality of nodes includes a plurality of central processing unit (CPU) sockets in which a plurality of CPU cores are installed, respectively; a plurality of memories respectively connected to the plurality of CPU sockets; and a plurality of database server instances managing allocated database partitions, wherein each database server instance is installed in units of CPU socket groups including a single CPU socket or at least two CPU sockets.

公开(公告)号：US20150212741A1

公开(公告)日：2015-07-30

申请号：US14606916

申请日：2015-01-27

Applicant: Electronics and Telecommunications Research Institute

Inventor： Hun Soon LEE ,  Mi Young LEE ,  Chang Soo KIM ,  Kyoung Hyun PARK ,  Mai Hai THANH

IPC: G06F3/06

CPC classification number: G06F3/0613 ,  G06F3/0647 ,  G06F3/068 ,  G06F3/0685

Abstract: Disclosed is an apparatus for in-memory data management includes a hybrid memory including a plurality of types of memories with different characteristics and a storage engine. The storage engine rearranges data among the plurality of memories by the unit of a page by monitoring workloads for data stored in the memories and reconfigures a page layout by the unit of the page based on a data access characteristic of an application for each of pages constituting each of the plurality of memories.

Abstract translation: 公开了一种用于存储器内数据管理的装置，包括具有不同特性的多种类型的存储器和存储引擎的混合存储器。 存储引擎通过监视存储在存储器中的数据的工作负载来重新排列多个存储器中的数据，并通过基于页面单元的页面单元重新配置页面布局，每个页面构成 每个多个存储器。

公开(公告)号：US20190079972A1

公开(公告)日：2019-03-14

申请号：US16129407

申请日：2018-09-12

Applicant: Electronics and Telecommunications Research Institute

Inventor： Mi Young LEE

IPC: G06F17/30 ,  G06F9/48 ,  G06F9/445

Abstract: The present invention relates to a system and method for parallel query processing by applying just-in-time (JIT) compilation-based query optimization when a query is processed. The system for parallel query processing based on JIT compilation according to the present invention includes a parallel processing scheduler configured to receive a database (DB) operation graph and operation dependency relation and distribute execution tasks and workers configured to execute a query executable code, wherein the workers include a worker for executing a JIT compiled executable code and a worker for executing the query executable code in an interpreter manner.

公开(公告)号：US20150199214A1

公开(公告)日：2015-07-16

申请号：US14249768

申请日：2014-04-10

Applicant: ELECTRONICS AND TELECOMMUNICATIONS RESEARCH INSTITUTE

Inventor： Myung Cheol LEE ,  Mi Young LEE ,  Sung Jin HUR

IPC: G06F9/48

CPC classification number: G06F9/505

Abstract: Disclosed is a system for distributed processing of stream data, including: a service management device which selects an operation device optimal to perform an operation constituting a service and assigns the operation in a node including the selected operation device; and a task execution device which performs one or more tasks included in the operation through the selected operation device when the assigned operation is an operation registered in a preregistered performance acceleration operation library.

Abstract translation: 本发明公开了一种流数据分散处理系统，包括：服务管理装置，其选择最优化的操作装置，以执行构成服务的操作，并在包括所选择的操作装置的节点中分配操作; 以及任务执行装置，当分配的操作是在预登记的性能加速操作库中登记的操作时，通过所选择的操作装置执行包括在操作中的一个或多个任务。

公开(公告)号：US20150040129A1

公开(公告)日：2015-02-05

申请号：US14079234

申请日：2013-11-13

Applicant: Electronics and Telecommunications Research Institute

Inventor： Choon Seo PARK ,  Mi Young LEE ,  Sung Jin HUR

IPC: G06F9/455

CPC classification number: G06F9/5077 ,  G06F9/45558 ,  G06F9/4856 ,  G06F2009/4557 ,  G06F2209/505 ,  Y02D10/22 ,  Y02D10/24 ,  Y02D10/32 ,  Y02D10/36

Abstract: A system for virtual machine placement and management monitors information regarding states of physical machines and virtual machines operated in a subgroup, and relocates the virtual machines operated in the subgroup according to information regarding states of the physical machines operated in the subgroup and a placement policy of the virtual machines.

Abstract translation: 用于虚拟机放置和管理的系统监视关于在子组中操作的物理机器和虚拟机的状态的信息，并且根据关于在子组中操作的物理机器的状态的信息重新定位在子组中操作的虚拟机，以及 虚拟机。

公开(公告)号：US20210357753A1

公开(公告)日：2021-11-18

申请号：US17317607

申请日：2021-05-11

Applicant: ELECTRONICS AND TELECOMMUNICATIONS RESEARCH INSTITUTE

Inventor： Jin Kyu KIM ,  Byung Jo KIM ,  Seong Min KIM ,  Ju-Yeob KIM ,  Ki Hyuk PARK ,  Mi Young LEE ,  Joo Hyun LEE ,  Young-deuk JEON ,  Min-Hyung CHO

IPC: G06N3/08 ,  G06N3/04

Abstract: A method and apparatus for multi-level stepwise quantization for neural network are provided. The apparatus sets a reference level by selecting a value from among values of parameters of the neural network in a direction from a high value equal to or greater than a predetermined value to a lower value, and performs learning based on the reference level. The setting of a reference level and the performing of learning are iteratively performed until the result of the reference level learning satisfies a predetermined value and there is no variable parameter that is updated during learning among the parameters.

公开(公告)号：US20200327390A1

公开(公告)日：2020-10-15

申请号：US16808124

申请日：2020-03-03

Applicant: ELECTRONICS AND TELECOMMUNICATIONS RESEARCH INSTITUTE

Inventor： Mi Young LEE ,  Joo Hyun LEE

IPC: G06N3/04 ,  G06N3/063 ,  G06F12/02

Abstract: Provided is a method of operating a neural network computing device that is configured to communicate with an external memory device and execute a plurality of layers. The method includes computing a first input address, based on first layer information of a first layer among the plurality of layers and a first memory management table, and updating the first memory management table to generate a second memory management table, reading first input data to be input to the first layer from the external memory device, based on the computed first input address, computing a first output address, based on the first layer information and the second memory management table, and updating the second memory management table to generate a third memory management table, and storing first output data output from the first layer, based on the first output address, in the external memory device.