-
公开(公告)号:US20230125872A1
公开(公告)日:2023-04-27
申请号:US17881417
申请日:2022-08-04
Inventor: Seon-Ho HAN , Young-Su KWON
Abstract: Disclosed are a digital CDR circuit and a feedback loop circuit including the same. The digital CDR circuit includes a phase detector that receives an input signal and outputs a phase detection result signal corresponding to a determination result for a sampling time based on the input signal, a charge pump that receives the phase detection result signal and outputs an amplified signal obtained by multiplying the phase detection result signal by a gain, a loop filter that receives the amplified signal and filters the amplified signal to output a filtered signal, and a phase shift control code generator that generates a control signal for controlling a phase of a signal based on the filtered signal, and the input signal includes plural data signals and plural error signals, and the data signals and the error signals are digital signals which are quantized based on a signal magnitude.
-
公开(公告)号:US20220164192A1
公开(公告)日:2022-05-26
申请号:US17533788
申请日:2021-11-23
Inventor: Chun-Gi LYUH , Hyun Mi KIM , Young-Su KWON , Jin Ho HAN
Abstract: Disclosed is a parallel processor. The parallel processor includes a processing element array including a plurality of processing elements arranged in rows and columns, a row memory group including row memories corresponding to rows of the processing elements, a column memory group including column memories corresponding to columns of the processing elements, and a controller to generate a first address and a second address, to send the first address to the row memory group, and to send the second address to the column memory group. The controller supports convolution operations having mutually different forms, by changing a scheme of generating the first address.
-
公开(公告)号:US20200167245A1
公开(公告)日:2020-05-28
申请号:US16694913
申请日:2019-11-25
Inventor: Jin Ho HAN , Young-Su KWON , Min-Seok CHOI
Abstract: Provided is an image recognition processor. The image recognition processor includes a plurality of nano cores each configured to perform a pattern recognition operation and arranged in rows and columns, an instruction memory configured to provide instructions to the plurality of nano cores in a row unit, a feature memory configured to provide input features to the plurality of nano cores in a row unit, a kernel memory configured to provide a kernel coefficient to the plurality of nano cores in a column unit, and a difference checker configured to receive a result of the pattern recognition operation of each of the plurality of nano cores, detect whether there is an error by referring to the received result, and provide a fault tolerance function that allows an error below a predefined level.
-
公开(公告)号:US20190108105A1
公开(公告)日:2019-04-11
申请号:US16117403
申请日:2018-08-30
Inventor: Jin Ho HAN , Min-Seok CHOI , Young-Su KWON
IPC: G06F11/14
Abstract: Provided is a semiconductor system including: a fault detector configured to obtain fault information related to a fault occurring in a first intellectual property (IP); a fault manager configured to store recovery information providing one or more recovery methods related to the fault information and determine a recovery method for recovering the fault occurring in the first IP among the one or more recovery methods based on the recovery information; and a fault recovery module configured to control the first IP based on the determined recovery method. The determined recovery method involves communication between the first IP and a second IP and the fault occurring in the first IP is recovered based on data delivered according to the communication between the first IP and the second IP.
-
公开(公告)号:US20190079801A1
公开(公告)日:2019-03-14
申请号:US16038243
申请日:2018-07-18
Inventor: Chun-Gi LYUH , Young-Su KWON , Chan KIM , Hyun Mi KIM , Jeongmin YANG , Jaehoon CHUNG , Yong Cheol Peter CHO
Abstract: Provided is a neural network accelerator which performs a calculation of a neural network provided with layers, the neural network accelerator including a kernel memory configured to store kernel data related to a filter, a feature map memory configured to store feature map data which are outputs of the layers, and a Processing Element (PE) array including PEs arranged along first and second directions, wherein each of the PEs performs a calculation using the feature map data transmitted in the first direction from the feature map memory and the kernel data transmitted in the second direction from the kernel memory, and transmits a calculation result to the feature map memory in a third direction opposite to the first direction.
-
公开(公告)号:US20140344551A1
公开(公告)日:2014-11-20
申请号:US14257067
申请日:2014-04-21
Inventor: Young-Su KWON
IPC: G06F9/38
CPC classification number: G06F9/3806 , G06F9/30189 , G06F9/3802 , G06F9/382 , G06F9/3844 , G06F9/3873
Abstract: The dual-mode instruction fetching apparatus includes a mode register, a branch prediction unit, a Program Counter (PC) calculator, an Instruction Queue (IQ), and a fetch multiplexer. The mode register is set to one of normal mode and line mode. The PC calculator accesses a tag in which the address indices of instructions have been stored or a line in which the instructions have been grouped and then outputs an instruction, or accesses only the line and then outputs an instruction depending on the type of set mode. The IQ stores instructions selected by an instruction selector from among the instructions grouped in the line. The fetch multiplexer fetches the instructions stored in the IQ if the normal mode has been set, and fetches instructions read from the line of an instruction cache if the line mode has been set.
Abstract translation: 双模指令获取装置包括模式寄存器,分支预测单元,程序计数器(PC)计算器,指令队列(IQ)和提取多路复用器。 模式寄存器设置为正常模式和线路模式之一。 PC计算器访问已经存储了指令的地址索引的标签或指令已被分组的行,然后输出指令,或仅访问该行,然后根据设置模式的类型输出指令。 IQ从存储在行中的指令中存储由指令选择器选择的指令。 如果已经设置了正常模式,则提取复用器将获取存储在IQ中的指令,并且如果设置了线路模式,则从指令高速缓存行读取指令。
-
公开(公告)号:US20240194241A1
公开(公告)日:2024-06-13
申请号:US18223097
申请日:2023-07-18
Inventor: Young-Deuk JEON , Young-Su KWON , Yi-Gyeong KIM , Su-Jin PARK , Min-Hyung CHO , Jae-Woong CHOI
IPC: G11C11/4074 , G11C11/4076
CPC classification number: G11C11/4074 , G11C11/4076
Abstract: Disclosed herein is an apparatus for adjusting a reference voltage. The apparatus may include a gate signal generation unit for generating an RDQS gate signal, a reference voltage generation unit for setting a reference voltage based on the RDQS gate signal, and a reset counter for holding a voltage at the time at which the RDQS gate signal becomes low when the RDQS gate signal is not applied to the reference voltage generation unit for a specific time period.
-
8.
公开(公告)号:US20190164037A1
公开(公告)日:2019-05-30
申请号:US16204599
申请日:2018-11-29
Inventor: Chan KIM , Young-Su KWON , Hyun Mi KIM , Chun-Gi LYUH , Yong Cheol Peter CHO , Min-Seok CHOI , Jeongmin YANG , Jaehoon CHUNG
Abstract: In the present invention, by providing an apparatus for processing a convolutional neural network (CNN), including a weight memory configured to store a first weight group of a first layer, a feature map memory configured to store an input feature map where the first weight group is to be applied, an address generator configured to determine a second position spaced from a first position of a first input pixel of the input feature map based on a size of the first weight group, and determine a plurality of adjacent pixels adjacent to the second position; and a processor configured to apply the first weight group to the plurality of adjacent pixels to obtain a first output pixel corresponding to the first position, a memory space may be efficiently used by saving the memory space.
-
公开(公告)号:US20190164035A1
公开(公告)日:2019-05-30
申请号:US16201871
申请日:2018-11-27
Inventor: Young-Su KWON , Chan KIM , Hyun Mi KIM , Jeongmin YANG , Chun-Gi LYUH , Jaehoon CHUNG , Yong Cheol Peter CHO
IPC: G06N3/04
Abstract: A reorganizable neural network computing device is provided. The computing device includes a data processing array unit including a plurality of operators disposed at locations corresponding to a row and a column. One or more chaining paths which transfer the first input data from the operator of the first row of the data processing array to the operator of the second row are optionally formed. The plurality of first data input processors of the computing device transfer the first input data for a layer of the neural network to the operators along rows of the data processing array unit, and the plurality of second data input processors of the computing device transfer the second input data to the operators along the columns of the data processing array.
-
10.
公开(公告)号:US20180189643A1
公开(公告)日:2018-07-05
申请号:US15847466
申请日:2017-12-19
Inventor: Chan KIM , Young-Su KWON , Jin Ho HAN
CPC classification number: G06N3/063 , G06K9/00993 , G06K9/4604 , G06K9/4628 , G06K9/6274 , G06K9/66 , G06N3/04 , G06N3/0454 , G06N3/08
Abstract: Provided is an operation method of a convolution circuit. The method includes receiving input feature maps, generating output feature maps corresponding to the respective input feature maps through convolution operations for performing parallel processing with a kernel unit, and outputting the output feature maps to an external memory.
-
-
-
-
-
-
-
-
-