Abstract: Disclosed are a general machine learning model generation method and apparatus, and a computer device and a storage medium. The method comprises: acquiring task parameters of a machine learning task (S1201); performing classification processing on the task parameters to obtain task instructions and model parameters (S1202); aggregating the task instructions and the model parameters according to a data type to obtain stack data and heap data (S1203); and integrating the stack data and the heap data to obtain a general machine learning model (S1204). By means of the method, compiled results of a corresponding general model in the running of an algorithm can be directly executed, which avoids repetitive compilation, thus greatly improving the efficiency of machine learning algorithm implementation and shortening the time from compilation to obtaining execution results.

Abstract: The present disclosure provides a data screening device and method, which employ a storage unit and a register unit, and are capable of performing operations on data of different storage structures and different sizes efficiently.

Abstract: Aspect for storage device with fault tolerance capability for neural networks are described herein. The aspects may include a first storage unit of a storage device. The first storage unit is configured to store one or more first bits of data and the data includes floating point type data and fixed point type data. The first bits include one or more sign bits of the floating point type data and the fixed point type data. The aspect may further include a second storage unit of the storage device. The second storage unit may be configured to store one or more second bits of the data. In some examples, the first storage unit may include an ECC memory and the second storage unit may include a non-ECC memory. The ECC memory may include an ECC check Dynamic Random Access Memory and an ECC check Static Random Access Memory.

Abstract: A data sharing system may include a storage module and at least two processing modules. The at least two processing modules may share the storage module and the at least two processing modules communicate to implement data sharing. A data sharing method for the data sharing system is provided. According to the disclosure, a storage communication overhead may be reduced, and a data access delay may be effectively reduced.

Abstract: The present application provides an operation device and related products. The operation device is configured to execute operations of a network model, wherein the network model includes a neural network model and/or non-neural network model; the operation device comprises an operation unit, a controller unit and a storage unit, wherein the storage unit includes a data input unit, a storage medium and a scalar data storage unit. The technical solution provided by this application has advantages of a fast computation speed and energy-saving.

Abstract: The application provides a Dynamic Voltage Frequency Scaling device. The Dynamic Voltage Frequency Scaling device in a convolutional operation device acquires working state information of the convolutional operation device and its internal units/modules in real time and scales working voltage or working frequency of the convolutional operation device and its internal units/modules according to the working state information of the convolutional operation device and its internal units/modules, so as to reduce the overall running power consumption of the convolutional operation device during the convolutional operation.

Abstract: Aspect for storage device with fault tolerance capability for neural networks are described herein. The aspects may include a first storage unit of a storage device. The first storage unit is configured to store one or more first bits of data and the data includes floating point type data and fixed point type data. The first bits include one or more sign bits of the floating point type data and the fixed point type data. The aspect may further include a second storage unit of the storage device. The second storage unit may be configured to store one or more second bits of the data. In some examples, the first storage unit may include an ECC memory and the second storage unit may include a non-ECC memory. The ECC memory may include an ECC check Dynamic Random Access Memory and an ECC check Static Random Access Memory.

Abstract: The application provides a Dynamic Voltage Frequency Scaling device including an information acquisition unit configured to acquire working state information or application scenario information of a chip connected with the Dynamic Voltage Frequency Scaling device in real time and a voltage frequency scaling unit configured to send voltage frequency scaling information to the chip according to the working state information or the application scenario information of the chip. The embodiment of the application dynamically scales the voltage frequency on chip-level and reduces the power consumption of chips.

Abstract: The present disclosure provides a computation device and method, which are capable of using a single instruction to complete a transpose computation of a matrix of any size within constant time. Compared with conventional methods for performing a matrix transpose computation, the device and method may reduce the time complexity of a matrix transpose computation as well as making the usage of the computation simpler and more efficient.

Abstract: The application provides an operation method and device. Quantized data is looked up to realize an operation, which simplifies the structure and reduces the computation energy consumption of the data, meanwhile, a plurality of operations are realized.

Abstract: Aspects of a neural network convolution device are described herein. The aspects may include a matrix transformer and a matrix multiplication module. The matrix transformer may be configured to receive an input data matrix and a weight matrix, transform the input data matrix into a transformed input data matrix based on a first transformation matrix, and transform the weight matrix into a transformed weight matrix based on a second transformation matrix. The matrix multiplication module may be configured to multiply one or more input data elements in the transformed input data matrix with one or more weight elements in the transformed weight matrix to generate an intermediate output matrix. The matrix transformer may be further configured to transform the intermediate output matrix into an output matrix based on an inverse transformation matrix.

Abstract: Systems and methods for scheduling an instruction list for parallel processing tasks are provided. An exemplary method includes obtaining an instruction set in the instruction list to be scheduled and determining data dependencies among instructions in the instruction set by performing a data dependency analysis on the instruction set. The method also includes obtaining, based on the data dependencies, selection nodes for performing instruction selections during the scheduling of the instruction list. The method further includes determining, based on a preset rule, an order of instructions in a scheduled instruction list according to a corresponding order of the selection nodes.

Abstract: Computer systems, data processing methods, and computer-readable media are provided to run original networks. An exemplary computer system includes first and second processors a memory storing offline models and corresponding input data of a plurality of original networks, and a runtime system configured to run on the first processor. The runtime system, when runs on the first processor, causes the first processor to implement a plurality of virtual devices comprising a data processing device configured to obtain an offline model and corresponding input data of an original network from the memory, an equipment management device configured to control turning on or off of the second processor, and a task execution device configured to control the second processor to run the offline model of the original network.

Abstract: Computer systems, data processing methods, and computer-readable media are provided to run original networks. An exemplary computer system includes first and second processors and first and second memories. The first memory stores offline models and corresponding input data of a plurality of original networks, and a runtime system configured to run on the first processor. The second memory stores an operating system configured to run on the first processor or the second processor. When the runtime system runs on the first processor, the runtime system obtains an offline model and corresponding input data of an original network from the first memory and controls the second processor to run the offline model of the original network. The offline model of the original network includes model parameters, instructions, and interface data of respective computation nodes of the original network.

Abstract: The present application provides an operation device and related products. The operation device is configured to execute operations of a network model, wherein the network model includes a neural network model and/or non-neural network model; the operation device comprises an operation unit, a controller unit and a storage unit, wherein the storage unit includes a data input unit, a storage medium and a scalar data storage unit. The technical solution provided by this application has advantages of a fast calculation speed and energy-saving.

Abstract: An example image compression method may include acquiring an original image with a first resolution; compressing the original image on the basis of the target model to obtain a compressed image with a second resolution; recognizing the compressed image on the basis of a recognition neural network model to obtain reference tag information; acquiring a loss function according to the target tag information and the reference tag information; if the loss function is convergent to a first threshold value or the present number of training times of the compression neural network is more than or equal to a second threshold value, acquiring a target original image with the first resolution, and determining the target model as a corresponding compression neural network model if training of the compression neural network is completed; compressing the target original image on the basis of the compression neural network model.

Abstract: The present disclosure provides a task parallel processing method, a device, a system, a storage medium and computer equipment, which are capable of distributing and regulating tasks to be executed according to a task directed acyclic graph, and may thereby realize task parallelism of a multi-core processor and improve the efficiency of data processing.

Abstract: 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.

Abstract: Aspects of data compression/decompression for neural networks are described herein. The aspects may include a model data converter configured to convert neural network content values into pseudo video data. The neural network content values may refer to weight values and bias values of the neural network. The pseudo video data may include one or more pseudo frames. The aspects may further include a compression module configured to encode the pseudo video data into one or more neural network data packages.