Canaan Creative Co., Ltd.

Canaan Bright Sight Co., Ltd.

Hangzhou Canaan Intelligence Information Technology Co, Ltd

Canaan Creative (sh) Co., Ltd.

Nangeng Zhang

Ning Zhang

Bing Xu

Shaohua Zhang

Jingjie Wu

Jieyao Liu

Shenghou Ma

Min Chen

Huan Yang

Hongbin Mao

Yingtao Xu

Minli Liu

Nan Li

Jian Zhang

Xiaogang Sun

Guanchao Zhao

Yanbin Zhu

Huanlai Zhu

Suncheng Gu

Xingang Zhai

Chenghai Huo

Jinhua Bao

Xuguang Liu

Jiakun Ma

Xiaoye Zhao

Shishuang He

He Ying

Zonghai Chen

Circuit board

Circuit board with heat sink

Methods and devices for generating a training sample, training a model and recognizing a character are provided. The method for generating a training sample comprises: acquiring an image of characters, and determining respective characters contained in the image; and using a projection method to determine weights of the respective characters contained in the image, tagging the image with labels according to the weights of the respective characters contained in the image, and forming a training sample. The method for training a model comprises: using the training sample to train a character recognition model. The method for recognizing a character comprises: using the character recognition model to perform character recognition. The above methods and devices realize accurate recognition of characters, such as double-half characters, contained in an image of a wheel-type meter, and can provide a highly accurate biased recognition result.

Methods and devices for generating training sample, training model and recognizing character

A character segmentation method and apparatus, and a computer-readable storage medium are provided. The method includes converting a character area image into a grayscale image; converting the grayscale image into an edge binary image by edge detection; acquiring character box segmentation blocks from the edge binary image by projection; and determining a target character area from the character box segmentation blocks by contour detection, and performing character segmentation on the character area image according to the target character area; or comprises: converting a character area image into a grayscale image; performing clustering analysis on the grayscale image by fuzzy C-means clustering, and binarizing the grayscale image according to the analysis result; acquiring character positioning blocks from a binary image by projection; and performing character segmentation on the character area image according to position information of the character positioning blocks. Character segmentation can be performed on a relatively low quality image.

Character segmentation method and device based on edge detection and contour detection

Cooling container for computing devices

Disclosed is an immersed liquid cooling heat dissipation system, which comprises a liquid cooling module, an oil path circulation device, and a plurality of computing devices to undergo heat dissipation. Each computing device comprises a frame, a control module, a power module, and computing modules. The liquid cooling module comprise a first device slot tank, a second device slot tank, a return flow slot tank, and a flow-equalizing plate. The return flow slot tank is located between the first device slot tank and the second device slot tank, wherein the flow-equalizing plate is disposed in the first device slot tank and the second device slot tank, and the computing devices are disposed on the flow-equalizing plate. The frame is internally provided with a power module accommodating region used for accommodating the power module and a computing module accommodating region used for accommodating the computing module.

Immersed liquid cooling heat dissipation system

A computational heat dissipation structure includes a circuit board including a plurality of heating components; and a radiator provided corresponding to the circuit board; wherein a space between the adjacent heating components is negatively correlated with heat dissipation efficiency of a region where the adjacent heating components are located. Since the space between the adjacent heating components of the disclosure is negatively correlated with the heat dissipation efficiency of the region where the adjacent heating components are located, i.e., the higher the heat dissipation efficiency of the region where the adjacent heating components are located is, the smaller the space between the adjacent heating components in the region will be, the heat dissipation efficiencies corresponding to the heating components are balanced, and load of a fan is reduced.

Computational heat dissipation structure, computing device comprising same, mine

A chip frequency modulation method and apparatus of a computing device, a hash board, a computing device, and a storage medium are disclosed. The computing device is provided with at least one operational chip, and the operational chip is provided with a plurality of cores. The chip frequency modulation method includes: operating each of the plurality of cores in the operational chip configured with a plurality of frequencies to run at a working frequency, the working frequency being one of the plurality of frequencies; analyzing a computing performance indicator of each of the plurality of cores at the working frequency; and modulating a working frequency of at least one core up or down according to the computing performance indicator, a modulated working frequency being one of the plurality of frequencies.

Chip frequency modulation method and apparatus of computing device, hash board, computing device and storage medium

Provided are a cloud computing power allocation method, a user terminal, a cloud computing power platform, and a system. The method includes: generating a computing power request including a computing power demand and account information of a computing power scheduling center; sending the computing power request to a cloud computing power platform, so that the cloud computing power platform sends a configuration instruction to a computing device cluster according to the computing power request, where the configuration instruction is to allocate to the user terminal a target computing device meeting the computing power demand from the computing device cluster and configure based on the account information the target computing device to execute a computing task issued by the computing power scheduling center; and acquiring from the computing power scheduling center computing power information determined according to a computing result from the target computing device, by using the account information.

Cloud computing power allocation method, user terminal, cloud computing power platform and system

A waste heat utilization system of an immersed liquid cooling heat dissipation device. The immersed liquid cooling heat dissipation device () comprises a liquid cooling tank (). The liquid cooling tank () comprises an oil tank inlet () and an oil tank outlet (). The system further comprises a waste heat utilization device (). The waste heat utilization device () comprises a waste heat utilization body (), a cold oil outlet () and a hot oil inlet (), the cold oil outlet () and the hot oil inlet () being connected to the waste heat utilization body (); the cold oil outlet () is connected with the oil tank inlet (); the hot oil inlet () is connected to the oil tank outlet (); the waste heat utilization body () is connected to a heat utilization end ().

Waste heat utilization system of immersed liquid cooling heat dissipation device

Calculating equipment

A series circuit and a computing device include a power supply terminal for providing voltage for a plurality of chips disposed on the computing device; a ground terminal disposed at one end of each of the plurality of chips relative to the power supply terminal; and a first connection line for separately connecting a first predetermined number of chips of the plurality of chips in series, wherein a communication line is connected between adjacent chips of the first predetermined number of chips, a portion of the communication line is connected to a target connection point, which is disposed on the first connection line and adapted to the adjacent chips, via a third connection line, and the voltage at the target connection point is greater than or equal to the minimum voltage required for communication between the adjacent chips.

Series circuit and computing device

A dilated convolution acceleration calculation method and apparatus. The method comprises: decomposing a dilated convolution computation of R×S into S sub-dilated convolution computations of R×1 (); for each sub-dilated convolution computation, caching a plurality of weight values in parallel to a plurality of calculation units in a calculation unit array (); determining, from input image data, a plurality of input data streams respectively corresponding to the plurality of weight values, and inputting the plurality of input data streams in parallel into the plurality of calculation units (); within the plurality of calculation units, executing a sliding window operation and a multiplication operation on the basis of the cached weight values and the input data streams, and executing an accumulation operation between the plurality of calculation units, so as to output an intermediate result of the sub-dilated convolution computation (); and superimposing intermediate results of the S sub-dilated convolution computations of R×1, so as to obtain a convolution result of the dilated convolution computation (). By using the method, a dilated convolution operation is accelerated with a relatively low complexity, and the function of Im2col does not need to be separately realized, thereby reducing the complexity.

Dilated convolution acceleration calculation method and apparatus

Growing community of inventors

Beijing, China

Nangeng Zhang