-
公开(公告)号：US20230214445A1
公开(公告)日：2023-07-06
申请号：US18120241
申请日：2023-03-10
Applicant: SAMSUNG ELECTRONICS CO., LTD.
Inventor： Chulsoo PARK , Jongho KIM , Youngrae CHO , Kiseok KWON , Jonghun LEE
CPC classification number: G06F17/153 , G06F17/16
Abstract: A memory of an electronic device stores three-dimensional input data comprising (i) input values, (ii) first kernel information, and (iii) second kernel information. The processor includes multiplication modules corresponding to the channels and performs a convolution operation based on the input values and the weights through the multiplication modules. Based on a depthwise convolution operation, a processor of the electronic device controls an input selection module to (a) configure the input values to correspond to a first channel among the channels and (b) input the input values to two or more multiplication modules among the multiplication modules. The processor inputs weights, obtains intermediate values, and obtains output values based on each of a summed result by summing intermediate values respectively corresponding to locations of the kernels from among the intermediate values through a first intermediate value accumulation module.
-
公开(公告)号：US20210019625A1
公开(公告)日：2021-01-21
申请号：US17044104
申请日：2019-01-04
Applicant: SAMSUNG ELECTRONICS CO., LTD.
Inventor： Dongsoo LEE , Byeoungwook LIM , Kwanghoon SON , Jonghun LEE
Abstract: Disclosed is an electronic device. The present electronic device includes: a memory; and a processor which quantizes a neural network, trained on the basis of deep learning, to generate a quantized neural network, and stores the quantized neural network in the memory, wherein the processor quantizes, in preset first bit units, trained connection strengths between neurons of the trained neural network, inverse-quantizes the quantized connection strengths in preset second bit units, retrains the inverse-quantized connection strengths, and quantizes the retrained connection strengths in the preset first bit units.
-
公开(公告)号：US20250089181A1
公开(公告)日：2025-03-13
申请号：US18956774
申请日：2024-11-22
Applicant: Samsung Electronics Co., Ltd.
Inventor： Jonghun LEE , Yunjin NAM , Kwondeuk YOON , Hyunsuk CHOI
IPC: H05K5/00
Abstract: According to various embodiments of the present disclosure, provided is an electronic device, comprising: a first housing including a first side and a second side opposite the first side; a second housing that covers at least a portion of the second side of the first housing; a printed circuit board disposed between the first housing and the second housing; and a display disposed on the first side of the first housing and electrically connected to the printed circuit board, wherein: the first housing comprises a first member including a first edge that forms an outer edge of the first housing, a second member including a second edge disposed within the edge formed by the first member and provides the second side of the first housing for having the printed circuit board disposed thereon, a mold member comprising an injectable mold material configured to combine the first member and the second member and at least a portion of which overlaps a first metal member comprising a first metal and a second metal member comprising a second metal, and a fastening hole for connecting to the second housing through a fastening member comprising a fastener; the second member includes a fastening area disposed inside the mold member, and at least a portion of which is formed by folding to extend in a direction that the second side faces; and the fastening hole is formed through at least a portion of the fastening area of the second member disposed inside the mold member.
-
公开(公告)号：US20230129845A1
公开(公告)日：2023-04-27
申请号：US18078199
申请日：2022-12-09
Applicant: SAMSUNG ELECTRONICS CO., LTD.
Inventor： Cheongyo BAHK , Eunji JEONG , Jonghun LEE
Abstract: A method by which an electronic device performs a DNN operation includes performing a first modified operation including extending a channel of an output of a first layer with respect to a feature map input into the first layer, the first layer being one of a plurality of layers of a DNN, wherein the first modified operation includes a space-to-depth transformation operation, performing a neural network operation corresponding to layers between the first layer and a second layer as a channel-extended neural network operation, wherein the second layer is a layer of the plurality of layers of the DNN after the first layer, performing a second modified operation including reducing a channel of an output of the second layer with respect to a channel-extended feature map input into the second layer, wherein the second modified operation includes a depth-to-space transformation operation, and outputting a result of the DNN operation.
-
公开(公告)号：US20220147806A1
公开(公告)日：2022-05-12
申请号：US17527305
申请日：2021-11-16
Applicant: SAMSUNG ELECTRONICS CO., LTD.
Inventor： Jeongin YUN , Sungpill CHOI , Jonghun LEE
Abstract: An electronic device and a method for controlling are provided. The electronic device may include a memory storing first input data and first weight data used in operations of a neural network model and a processor configured to input the first input data and the first weight data into a first module, and acquire second input data and second weight data, where a part of the first input data is truncated, and where a part of the first weight data is truncated, input the second input data and the second weight data into a second module that performs multiplication operations, and acquire first output data, and based on scaling factors of the first input data and first weight data identified through the first module, convert the acquired first output data into a floating point form expressing a first bit as a unit scale and acquire second output data.
-
Patent Agency Ranking
公开(公告)号：US20250068894A1
公开(公告)日：2025-02-27
申请号：US18762315
申请日：2024-07-02
Applicant: SamSung Electronics Co., Ltd.
Inventor： Jonghun LEE , Chulsoo PARK
IPC: G06N3/048
Abstract: A method performed by at least one processor configured to implement an accelerator for processing a softmax function includes: obtaining input data comprising a plurality of quantized input values; generating input data distribution information indicating a plurality of frequencies corresponding to the plurality of quantized input values included in the input data; identifying a largest value from among the plurality of quantized input values as a first maximum value, based on the input data distribution information; determining an offset value based on a difference between the first maximum value and a second maximum value, wherein the second maximum value indicates a maximum quantization value that is representable by an input value of the input data; determining a plurality of index values by applying the offset value to each quantized input value of the plurality of quantized input values; and outputting a value of the softmax function corresponding to each of the plurality of quantized input values based on the plurality of index values.
-
公开(公告)号：US20240412052A1
公开(公告)日：2024-12-12
申请号：US18811302
申请日：2024-08-21
Applicant: SAMSUNG ELECTRONICS CO., LTD.
Inventor： Meejeong PARK , Jihun OH , Youngrae CHO , Jonghun LEE
IPC: G06N3/0495 , G06N3/0464
Abstract: A data processing method for neural network quantization, includes: obtaining a quantized weight by quantizing a weight of a neural network; obtaining a quantization error that is a difference between the weight and the quantized weight; obtaining input data with respect to the neural network; obtaining a first convolution result by performing convolution on the quantized weight and the input data; obtaining a second convolution result by performing convolution on the quantization error and the input data; obtaining a scaled second convolution result by scaling the second convolution result based on bit shifting; and obtaining output data by using the first convolution result and the scaled second convolution result.
-
公开(公告)号：US20220004858A1
公开(公告)日：2022-01-06
申请号：US17478246
申请日：2021-09-17
Inventor： Jonghun LEE , Youngsok KIM , Jangwoo KIM , Daehyun KIM , Myungsun KIM
Abstract: A method for processing an artificial network by an electronic device includes obtaining, by using a first processor and a second processor, a neural network computation plan for performing computation of a first neural network layer of the artificial neural network, performing a first portion of a computation of the first neural network layer by using the first processor, and performing a second portion of the computation of the first neural network layer by using the second processor based on the obtained neural network computation plan, obtaining a first output value based on a performance result of the first processor and a second output value based on a performance result of the second processor, and using the obtained first output value and the second output value as an input value of a second neural network layer of the artificial neural network.
-
公开(公告)号：US20240232282A1
公开(公告)日：2024-07-11
申请号：US18482524
申请日：2023-10-06
Applicant: Samsung Electronics Co., Ltd.
Inventor： Cheolgyu JIN , Jinhwan BAIK , Jaeun PARK , Jonghun LEE
IPC: G06F17/15
CPC classification number: G06F17/15
Abstract: An electronic apparatus is provided. The electronic device includes acquiring padding data corresponding to the input data in case of acquiring a convolution calculation instruction for the input data, identifying a calculation processing unit based on a size of the buffer and a size of the padding data, classifying the input data and the padding data into a plurality of target regions based on the calculation processing unit and the sizes of the buffers, storing one target region among the plurality of target regions in the first buffer, the second buffer or the third buffer, acquiring target data for the convolution calculation convolution calculation based on the calculation processing unit and the plurality of target regions, and controlling the convolution calculation module to perform the convolution calculation convolution calculation based on the target data and kernel data.
-
公开(公告)号：US20210263741A1
公开(公告)日：2021-08-26
申请号：US17098683
申请日：2020-11-16
Applicant: SAMSUNG ELECTRONICS CO., LTD.
Inventor： Dongkwan SUH , Jonghun LEE
Abstract: An electronic device and a control method thereof are disclosed. The electronic device includes: a memory storing input data, and a processor including a first register file and a second register file storing index data corresponding to kernel data, wherein the processor is configured to: based on a first command being input, obtain offset information of valid data included in a part of the index data stored in the first register file, based on the number of pieces of the offset information being greater than or equal to a predetermined number, store data packed with the offset information in a unit of the predetermined number in the second register file, and obtain output data by performing an operation regarding the input data based on the packed data.
-
-
-
-
-
-
-
-
-
Search Results
12
records
(took 0.012 seconds)
