Samsung Electronics Co., Ltd.

Lg Electronics Inc.

Seoul National University

Skc Co., Ltd.

Amorepacific Corporation

Hyundai Motor Company

Kia Motors Corporation

Samsung Sds Co., Ltd.

Sk Microworks Co., Ltd.

Ahnlab, Inc.

Taiwan Semiconductor Manufacturing Comp. Ltd.

Seungwon Lee

Jinwook Choi

Jungsu Lee

Seungcheon Baek

Junhaeng Lee

Kiyoung Choi

Hyunsoo Kim

Ji Min Lee

Young Min Heo

Jung Kyu Lee

Shinhaeng Kang

Joonho Song

Seungwoo Seo

Namhyung Kim

Hyeonjin Shin

Daeseok Byeon

Sehwan Lee

Dongwoo Lee

Youngmin Jo

Sangwon Ha

Hyunsun Park

Taehyo Kim

Youngjin Cho

Jeongho Lee

Jeehoon Han

Changseok Lee

Keunwook Shin

Minsu Seol

Changhyun Kim

Dongwook Lee

Hyunjae Song

Eunkyu Lee

Younghwan Park

Donghyun Kang

Hyein Yang

Seheon Choi

Wonjo Lee

Sechul Lee

Hosang Yoon

Hanmin Park

Sanghyo Lee

Jongpil Cho

Kyeongdo Kim

Minkyoung Cho

Woonkyung Seo

Songyi Han

Byeonghoon Lee

Sungbum Kang

Younghwan Son

Hyuncheol Park

Kyounghoon Kim

Dawoo Jeong

Seogoo Kang

Hyunjun Jung

Mincheol Kim

Kyungeun Byun

Juyoung Lim

Yuhwan Ro

Hansu Cho

Hanwoong Jung

Gunhee Lee

Seongwook Jo

Sukjin Kim

Kyu-Hun Kim

Donghyuk Kwon

Jung Won Yu

Soohan Yoo

Kyungsun Kang

Won-jo Lee

Hwanyeol Choi

Gyeongmin Lee

Ki-Yuel Kwak

Soyeun Park

Changyeon Hwang

Bongjun Kim

Haneol Lee

Yongkook Park

Joonsang Yu

Hyoujoo Kwon

Chang Soo Shin

Yonggoo Kim

Seol Woo Joo

Hojin Yu

Jae Hyeok Jang

A system configured to perform an operation includes: a hardware device comprising a plurality of computing modules and a plurality of memory modules arranged in a lattice form, each of the computing modules comprising a coarse-grained reconfigurable array and each of the memory modules comprising a static random-access memory and a plurality of functional units connected to the static random-access memory; and a compiler configured to divide a target operation and assign the divided target operation to the computing modules and the memory modules such that the computing modules and the memory modules of the hardware device perform the target operation.

Near-memory operator and method with accelerator performance improvement

Cosmetic container

Lipstick case

A processor-implemented method of processing neural network data includes: setting first limit data by performing a first operation based on first input data and weight data generated from weights included in a filter; comparing the first limit data with an intermediate result of a second operation performed based on second input data and the weight data; and determining whether to perform a subsequent second operation based on a result of the comparing.

Method and apparatus with data processing

A method and apparatus for scheduling a neural network operation. The method includes receiving data on a layer of a neural network, generating partitions to be assigned to cores by dividing the data, generating tiles by dividing the partitions, and scheduling an operation order of the tiles based on whether the data are shared between the cores.

Method and apparatus for partitioning neural network data

According to a method and apparatus for neural network quantization, a quantized neural network is generated by performing learning of a neural network, obtaining weight differences between an initial weight and an updated weight determined by the learning of each cycle for each of layers in the first neural network, analyzing a statistic of the weight differences for each of the layers, determining one or more layers, from among the layers, to be quantized with a lower-bit precision based on the analyzed statistic, and generating a second neural network by quantizing the determined one or more layers with the lower-bit precision.

Method and apparatus for neural network quantization

A method of generating a fixed-point quantized neural network includes analyzing a statistical distribution for each channel of floating-point parameter values of feature maps and a kernel for each channel from data of a pre-trained floating-point neural network, determining a fixed-point expression of each of the parameters for each channel statistically covering a distribution range of the floating-point parameter values based on the statistical distribution for each channel, determining fractional lengths of a bias and a weight for each channel among the parameters of the fixed-point expression for each channel based on a result of performing a convolution operation, and generating a fixed-point quantized neural network in which the bias and the weight for each channel have the determined fractional lengths.

Method and apparatus for generating fixed-point quantized neural network

A data processing apparatus and method are disclosed. The data processing apparatus may include a host core configured to generate a control message to control a direct memory access (DMA), the DMA configured to generate a memory request based on the control message, a memory controller configured to generate a memory command based on the memory request, and a processor configured to perform an operation in a memory based on the memory command.

Data processing apparatus and method using direct memory access

A processor-in-memory (PIM) and a method of outputting an instruction using a PIM. The PIM includes an internal processor, a memory, and a register configured to store instruction meta information. The memory is configured to store a lookup table generated by predicting a future instruction. The external processor is configured to retrieve an instruction corresponding to the instruction meta information from the lookup table in response to a PIM application programming interface (API) call instruction including the instruction meta information. The instruction corresponding to the instruction meta information is output based on a retrieval result.

Processor-in-memory storing a lookup table for providing an instruction in response to an API call instruction

A ring-type wearable electronic device is provided. The ring-type wearable electronic device includes a ring-shape housing including a first housing portion and a second housing portion coupled to the first housing portion, wherein the second housing portion is configured to contact a finger of a user wearing the ring-type wearable electronic device, a first light source including a plurality of light emitters disposed in the ring-shape housing, and configured to emit light of red wavelength and infrared wavelength bands towards the finger, a first sensor disposed in the ring-shape housing, and configured to receive light emitted by the first light source, a second sensor disposed in the ring-shape housing, and configured to receive light emitted by the first light source, wherein the first sensor is positioned farther from the first light source than the second sensor is positioned from the first light source, a flexible printed circuit board (FPCB) disposed in the ring-shape housing, wherein the first light source, the first sensor, and the second sensor are disposed on the FPCB at least partially bent corresponding to the ring-shape housing, at least one processor including a processing circuitry, and memory storing instructions, wherein the instructions, when executed by the at least one processor individually or collectively, cause the ring-type wearable electronic device to: emit, by the first light source at a first time, first light of red wavelength and infrared wavelength bands, the first light being configured to reach the first sensor through the finger, based on the first light emitted by the first light source at the first time, obtain first information corresponding to part of the first light that is received by the first sensor via the finger, emit, by the first light source at a second time different from the first time, second light of red wavelength and infrared wavelength bands, the second light being configured to reach the second sensor through the finger, based on the second light emitted by the first light source at the second time, obtain second information corresponding to part of the second light that is received by the second sensor via the finger, and acquire an oxygen saturation information of the user based on at least one of the first information or the second information.

Wearable electronic device for obtaining biometric information

A method of accelerating a training process of a neural network includes acquiring activations used in the training process and a bit-vector corresponding to the activations, selecting activations requiring an operation from among the acquired activations by using the bit-vector, and performing backward propagation using the selected activations and filters corresponding to the selected activations.

Method of accelerating training process of neural network and neural network device thereof

A security device includes a physical unclonable function (PUF) cell array that includes a plurality of PUF cells connected with a first word line, a controller that selects a target PUF cell of the plurality of PUF cells and outputs a control signal based on the target PUF cell, a decoder that applies a first voltage to the first word line in response to the control signal, a bit line selection circuit that outputs a target current across a bit line connected with the target PUF cell and a sum current corresponding to a sum of currents across the remaining bit lines connected with other PUF cells, and a bit determiner that outputs a target bit of the target PUF cell based on the target current and the sum current, and the security device generates a security key based on the target bit for responding to an authentication requests.

Security device including physical unclonable function cells, operation method of security device, and operation method of physical unclonable function cell device

A method and apparatus for the pruning of a neural network is provided. The method sets a weight threshold value based on a weight distribution of layers included in a neural network, predicts a change of inference accuracy of a neural network by pruning of each layer based on the weight threshold value, determines a current subject layer to be pruned with a weight threshold value among the layers included in the neural network, and prunes a determined current subject layer.

Neural network method and apparatus

A processor-implemented method of performing a convolution operation is provided. The method includes obtaining input feature map data and kernel data, determine the kernel data based on a number of input channels of the input feature map, a number of output channels of an output feature map, and a number of groups of the input feature map data and a number of groups of the kernel data related to the convolution operation, and performing the convolution operation based on the input feature map data and the determined kernel data.

Growing community of inventors

Seoul, South Korea

Seungwon Lee