Cambricon Technologies Corporation Limited

Shanghai Cambricon Information Technology Co., Ltd

International Business Machines Corporation

Beijing Xiaomi Mobile Software Co., Ltd.

Lenovo (beijing) Limited

Chinese Academy of Sciences

Cambricon (xi'an) Semiconductor Co., Ltd.

Analog Devices International Unlimited Company

Peking University

Beijing Lenovo Software Ltd.

Mitac International Corporation

Beijing Institute of Technology

Mitac Research (shanghai) Ltd.

Beijing Taijieweiye Technology Co., Ltd

Qi Guo

Tianshi Chen

Yunji Chen

Shaoli Liu

Zidong Du

Tian Zhi

Xunyu Chen

Yangyang Cai

Yimin Zhuang

Yuzhe Luo

Yan Li

Zhenhua Liu

Dong Han

Shengyuan Zhou

Guan Cheng Chen

Jianwei Lu

Tao Liu

Jun Liang

Antonio J Montalvo

Wen Wei Tang

Shijin Zhang

Linyang Wu

Huiying Lan

Xiao Ming Bao

Michael Cobb

Hao Jing

Kangyu Wang

Xin Li

Bin Yang

Zhen Wang

Rui Wang

Yong Yu

Jian Wu

Wei Xu

Jian Li

Zhen Li

Hongwei Li

Jin Wang

Jia Tian Zhong

Xing Xing Shen

Ning Zhang

Zai Wang

Qian Xu

Gang Chen

Shuang Hong Wang

Zhigang Li

Yifan Hao

Jie Wei

Jie Tan

Qingtao Sun

Weixian Guo

Jiantao Zhang

Wei Li

Chunli Song

Hongjiang Bi

Zhu Chen

Can Liu

Yingsheng Xu

Lijun Ma

Yongsong Xu

Hang Tang

Zhimei Liao

Ting Jiang

Shangying Li

Dakai Zhou

Jun Qiu

Aspects for modifying data in a multi-layer neural network (MNN) acceleration processor for neural networks are described herein. As an example, the aspects may include receiving a predetermined weight value array and connection data. Further, the aspects may include modifying the weight values included in the predetermined weight value array based on the connection data. Further still, the aspects may include calculating one or more groups of output data based on the modified weight values.

Calculating device and method for a sparsely connected artificial neural network

A neural network processing method, comprising the following steps: obtaining a model dataset and model structure parameters of an original network (S); obtaining an operational attribute of each compute node in the original network; operating the original network according to the model dataset and the model structure parameters of the original network and the operational attribute of each compute node, to obtain an instruction corresponding to each compute node in the original network (S); and if the operational attribute of the current compute node is a first operational attribute, storing a network weight and the instruction corresponding to the current compute node into a first non-volatile memory, so as to obtain a first offline model corresponding to the original network (S). Further provided are a computer system and a storage medium. The neural network processing method, the computer system, and the storage medium shorten the time for a processor to operate the same network, and improve the processing speed and efficiency of the processor.

Neural network processing method, computer system and storage medium

Aspects for forward propagation of a convolutional artificial neural network are described herein. The aspects may include a direct memory access unit configured to receive input data from a storage device and a master computation module configured to select one or more portions of the input data based on a predetermined convolution window. Further, the aspects may include one or more slave computation modules respectively configured to convolute a convolution kernel with one of the one or more portions of the input data to generate a slave output value. Further still, the aspects may include an interconnection unit configured to combine the one or more slave output values into one or more intermediate result vectors, wherein the master computation module is further configured to merge the one or more intermediate result vectors into a merged intermediate vector.

Apparatus and methods for forward propagation in convolutional neural networks

Aspects for neural network operations with fixed-point number of short bit length are described herein. The aspects may include a fixed-point number converter configured to convert one or more first floating-point numbers to one or more first fixed-point numbers in accordance with at least one format. Further, the aspects may include a neural network processor configured to process the first fixed-point numbers to generate one or more process results.

Apparatus and methods for neural network operations supporting fixed point numbers of short bit length

Aspects for backpropagation of a convolutional neural network are described herein. The aspects may include a direct memory access unit configured to receive input data from a storage device and a master computation module configured to select one or more portions of the input data based on a predetermined convolution window. Further, the aspects may include one or more slave computation modules respectively configured to convolute one of the one or more portions of the input data with one of one or more previously calculated first data gradients to generate a kernel gradient, wherein the master computation module is further configured to update a prestored convolution kernel based on the kernel gradient.

Apparatus and methods for training in convolutional neural networks

A double-balloon thrombolysis and thrombectomy device, comprising a catheter, an access sheath, and a hemostasis valve. The catheter comprises a base, a body, a first balloon, and a second balloon, the first balloon and the second balloon are sleeved over the body, the first balloon is located at a distal end of the body, and the second balloon is located at the proximal end of the first balloon; the base is connected to the proximal end of the body, and comprises a main connection port, a first connection port and a second connection port; the first balloon is connected to the first connection port, and the second balloon is connected to the second connection port; the surface of the second balloon is provided with a plurality of micropores; and the access sheath is sleeved over the catheter body, and connected to the catheter body through the hemostasis valve.

Double-balloon thrombolysis and thrombectomy device

Aspects for backpropagation in a fully connect layer of a convolutional neural network are described herein. The aspects may include a direct memory access unit configured to receive input data and one or more first data gradients from a storage device. The aspects may further include a master computation module configured to transmit the input data and the one or more first data gradients to one or more slave computation modules. The slave computation modules are respectively configured to multiply one of the one or more first data gradients with the input data to generate a default weight gradient vector.

Apparatus and methods for training in fully connected layers of convolutional networks

Embodiments of the present disclosure relate to multi-band FDD transceivers. An example transceiver includes a LO, configured to generate a LO signal to be shared between a receiver and a transmitter of the transceiver. Both the receiver and the transmitter use quadrature signal processing and are configured to multi-band operation. Sharing a single LO to perform frequency conversion of different frequency bands of received and transmitted signals advantageously allows reducing the number of LOs used in a multi-band FDD transceiver.

Single local oscillator in a multi-band frequency division duplex transceiver

Aspects for backpropagation of a multilayer neural network (MNN) in a neural network processor are described herein. The aspects may include a computation module configured to receive one or more groups of MNN data. The computation module may further include a master computation module configured to calculate an input gradient vector based on a first output gradient vector from an adjacent layer and based on a data type of each of the one or more groups of MNN data. Further still, the computation module may include one or more slave computation modules configured to parallelly calculate portions of a second output vector based on the input gradient vector calculated by the master computation module and based on the data type of each of the one or more groups of MNN data.

Apparatus and methods for backward propagation in neural networks supporting discrete data

A model conversion method is disclosed. The model conversion method includes obtaining model attribute information of an initial offline model and hardware attribute information of a computer equipment, determining whether the model attribute information of the initial offline model matches the hardware attribute information of the computer equipment according to the initial offline model and the hardware attribute information of the computer equipment and in the case when the model attribute information of the initial offline model does not match the hardware attribute information of the computer equipment, converting the initial offline model to a target offline model that matches the hardware attribute information of the computer equipment according to the hardware attribute information of the computer equipment and a preset model conversion rule.

Model conversion method, device, computer equipment, and storage medium

Mobile phone

Aspects for neural network operations with floating-point number of short bit length are described herein. The aspects may include a neural network processor configured to process one or more floating-point numbers to generate one or more process results. Further, the aspects may include a floating-point number converter configured to convert the one or more process results in accordance with at least one format of shortened floating-point numbers. The floating-point number converter may include a pruning processor configured to adjust a length of a mantissa field of the process results and an exponent modifier configured to adjust a length of an exponent field of the process results in accordance with the at least one format.

Apparatus and methods for neural network operations supporting floating point numbers of short bit length

Aspects of data modification for neural networks are described herein. The aspects may include a data modifier configured to receive input data and weight values of a neural network. The data modifier may include an input data configured to modify the received input data and a weight modifier configured to modify the received weight values. The aspects may further include a computing unit configured to calculate one or more groups of output data based on the modified input data and the modifier weight values.

Device and method for artificial neural network operation

Aspects for Long Short-Term Memory (LSTM) blocks in a recurrent neural network (RNN) are described herein. As an example, the aspects may include one or more slave computation modules, an interconnection unit, and a master computation module collectively configured to calculate an activated input gate value, an activated forget gate value, a current cell status of the current computation period, an activated output gate value, and a forward pass result.

Apparatus and method for executing recurrent neural network and LSTM computations

Aspects of a neural network operation device are described herein. The aspects may include a matrix element storage module configured to receive a first matrix that includes one or more first values, each of the first values being represented in a sequence that includes one or more bits. The matrix element storage module may be further configured to respectively store the one or more bits in one or more storage spaces in accordance with positions of the bits in the sequence. The aspects may further include a numeric operation module configured to calculate an intermediate result for each storage space based on one or more second values in a second matrix and an accumulation module configured to sum the intermediate results to generate an output value.

Multiplication and addition device for matrices, neural network computing device, and method

Growing community of inventors

Beijing, China

Qi Guo