13results about How to "Improve acceleration" patented technology

The invention relates to the technical field of integrated circuits, in particular to a processing system of a single-precision configurable matrix operation unit based on RISC-V instruction expansion, which comprises an RISC-V processor used for executing a software program and identifying a user-defined expansion instruction, and the RISC-V processor is connected with the single-precision configurable matrix operation unit through a user-defined coprocessor interface; the single-precision configurable matrix operation unit comprises an instruction response and decoding module and an instruction retirement and write-back module, and the instruction response and decoding module is used for receiving an instruction function code, source operation data and a destination register number and decoding the instruction function code; after decoding is completed, different execution units are selected according to a decoding result to send corresponding control signals; and after the execution unit completes the corresponding task, the instruction retirement and write-back module returns an execution result and a completion signal to the RISC-V processor, so that execution of a self-defined extension instruction is completed. According to the system, the matrix operation speed can be accelerated through bottom hardware under limited hardware resource consumption.

The invention discloses a block storage caching method, device and equipment and a storage medium, relates to the technical field of data processing, is applied to a server, and comprises the steps of monitoring real-time performance data of a block storage caching system through a user mode program constructed based on an extended Berkley packet filter when the block storage caching system is used for processing input and output requests; the block storage cache system is a system which is constructed in advance based on a machine learning algorithm and contains configuration parameters; according to the real-time performance data, parameters representing the processing sequence of the input and output requests are calculated, and parameter values are obtained; and based on the parameter value, adjusting a configuration parameter of the block storage cache system through the file system interface so as to process the input / output request by utilizing the adjusted block storage cache system. Thus, in the process that the block storage cache system processes the input and output requests, the related configuration parameters can be dynamically adjusted according to the real-time performance data, and the requirements of multiple service scenes can be met.

Provided are an optimization method and an optimization system for non-multiplication-addition type computing operations in an FPGA hardware accelerator. The optimization method comprises: generating a first computing graph by decomposing the non-multiplication-addition type computing operations into a plurality of basic operations; obtaining at least one merged operation by merging consecutive basic operations in the first computing graph; generating a second computing graph based on the first computing graph and the at least one merged operation; and executing the operation in the FPGA hardware accelerator according to the second computing graph; wherein the executing the operation in the FPGA hardware accelerator according to the second computing graph comprises: executing each merged operation of the at least one merged operation by pipeline optimization.

The application relates to the technical field of powder production and processing equipment, in particular to a flow guide device for airflow mill pulverization and a pulverizer adopting the flow guide device, wherein the flow guide device for airflow mill pulverization comprises a device body, a flow guide area, a collection area and a drainage area are formed in the device body, and the collection area and the drainage area are distributed around the flow guide area; when the device body is matched with a pulverization cavity and the pulverization cavity is in a working state, airflow sprayed from a nozzle flows to the lower part of the collision area after collision in the middle of a pulverization area and then flows into the flow guide area and the drainage area in sequence; the collection area collects coarse particles sinking downwards, the collection area is formed with an opening, coarse particles in the collection area can flow out through the opening, the flowed-out coarse particles can flow into the drainage area together with airflow and flow out of the drainage area; the flowing direction of airflow flowing out of the drainage area is towards the top of the pulverization cavity. The flow guide device for airflow mill pulverization can improve powder preparation efficiency and improve the vortex state at the bottom of the pulverization cavity.

The invention discloses an integrated motor structure with a bidirectional power recovery function, which belongs to the field of permanent magnet synchronous motors and comprises a casing suite, and a motor casing integrated electric control junction box assembly is mounted on the casing suite. A motor through type output shaft is arranged at the position of a central shaft of the machine shell suite, and a plurality of sets of stator assemblies and rotor assemblies are arranged in the machine shell suite. The stator assembly is fixedly connected with the inner wall of the machine shell suite, a radial gap is formed between the stator assembly and the circumference of the motor through type output shaft, the rotor assemblies are fixedly arranged on the motor through type output shaft, and the wet brakes are arranged between the different rotor assemblies. The motor is compact in structure, small in size, light in weight, capable of providing larger torque output, capable of reducing loss energy, capable of improving efficiency and capable of achieving accurate speed and torque control. The device is fast in dynamic response, does not generate electromagnetic noise and vibration in the operation process, is relatively simple in structure, and is relatively long in service life and relatively high in reliability.

The invention discloses a multi-module parallel image acceleration processing method, system and platform suitable for a GPU (Graphics Processing Unit) and a storage medium, and the method comprises the steps: obtaining first information data corresponding to the GPU, detecting the operation performance of the GPU in real time based on the first information data, and generating corresponding second information data; according to the second information data, performing real-time optimization processing on operation performance corresponding to the GPU, and generating corresponding third information data; on the basis of the GPU, in combination with the third information data, multi-module parallel acceleration processing and rendering of to-be-processed image data, a corresponding system, a platform and a storage medium, synchronous work of multiple acceleration modules is achieved, and the GPU performance is remarkably improved; the detection and optimization capability is improved, the acceleration effect is detected and further optimization is carried out through a performance analysis optimization module, and the effectiveness and stability of GPU acceleration are ensured; the actual application effect is enhanced, and the effect and performance in actual application are improved.

The diversion tunnel flood discharge structure comprises a diversion tunnel, a flood discharge pipe, a pressurizing pipe and an air compressor, the flood discharge pipe is fixedly arranged, one end of the flood discharge pipe is connected with the water outlet end of the diversion tunnel, the other end of the flood discharge pipe is connected with a river channel, one end of the pressurizing pipe is connected with the flood discharge pipe, and the other end of the pressurizing pipe is connected with the air compressor. The inner diameter of the pressurizing pipe is smaller than that of the flood discharge pipe. The diversion tunnel flood discharge structure is simple in overall structure, high in flood discharge efficiency and low in construction difficulty and cost.

The utility model provides a kind of abrasive slurry continuous jet device.Device includes: high-pressure power component, including oil tank, hydraulic oil pump and reversing valve assembly;Diaphragm continuous slurry supply component, diaphragm pump can pump abrasive slurry in abrasive slurry tank into first diaphragm tank or second diaphragm tank;Wherein, the first cavity of first diaphragm tank and the first cavity of second diaphragm tank are respectively connected to oil tank by reversing valve assembly;The second cavity of first diaphragm tank is connected to abrasive slurry tank by first stop valve, and the second cavity of second diaphragm tank is connected to abrasive slurry tank by second stop valve;Processing execution component, including abrasive slurry jet nozzle, abrasive slurry jet nozzle is connected to the second cavity of first diaphragm tank by third stop valve, and is connected to the second cavity of second diaphragm tank by fourth stop valve.The utility model uses first diaphragm tank and second diaphragm tank to alternate and spray abrasive slurry, and abrasive slurry can be continuously and uninterruptedly sprayed.

The application discloses a big data table hash connection hardware acceleration method and device, equipment and medium, and relates to the database heterogeneous acceleration field. The method is applied to an acceleration board card and comprises the following steps: receiving a connection operation type instruction; applying a first memory space of a first to-be-connected data table and a second memory space of a second to-be-connected data table respectively by using a connection table size instruction, so as to store each first shard of the first to-be-connected data table into the first memory space and store each second shard of the second to-be-connected data table into the second memory space; and performing hash connection calculation on each first shard in the first memory space and each second shard in the second memory space according to the connection operation type instruction and a preset parallel processing row number, so as to obtain corresponding connection table row data. Through the above scheme, the hash connection hardware acceleration effect can be effectively improved.

The invention provides an abrasive slurry continuous jet device and a control method thereof. The high-pressure power assembly comprises an oil tank, a hydraulic oil pump and a reversing valve assembly; according to the diaphragm continuous slurry supply assembly, a diaphragm pump can pump abrasive slurry in an abrasive slurry box into a first diaphragm tank or a second diaphragm tank; a first cavity of the first diaphragm tank and a first cavity of the second diaphragm tank are connected to the oil tank through the reversing valve assembly. A second cavity of the first diaphragm tank is connected to the abrasive slurry box through a first stop valve, and a second cavity of the second diaphragm tank is connected to the abrasive slurry box through a second stop valve; and the processing execution assembly comprises an abrasive slurry jet flow nozzle, and the abrasive slurry jet flow nozzle is connected to the second cavity of the first diaphragm tank through a third stop valve and connected to the second cavity of the second diaphragm tank through a fourth stop valve. The first diaphragm tank and the second diaphragm tank are adopted for alternately injecting and spraying abrasive slurry, and the abrasive slurry can be continuously sprayed.

The present invention discloses a method for locating the sound source of equipment failure based on deep learning theory, comprising the following steps: collecting sound source signals through a microphone array MEMS sensor to obtain the original sound source signal of the fault point; performing Fourier transform and grayscale processing on the original sound source data to convert it into an STFT grayscale image, and inputting the STFT grayscale image into a trained sound source position neural network module to obtain the positioning information of the fault point; inputting the original sound source data into a trained sound pressure intensity neural network module to obtain the sound pressure intensity of the fault point; combining the positioning information of the fault point with the sound pressure intensity to judge the fault condition of the equipment. Based on the characteristics of the original sound source data, the present invention innovatively adopts a multi-task learning network architecture of a mutually independent sound source position neural network module + a sound pressure intensity neural network module, thereby simultaneously obtaining the positioning information and sound pressure intensity of the fault point, providing two-way support for the judgment of the equipment fault condition.

A driving sound control system for an electric vehicle may include: a sound output device for generating a driving sound of an electric vehicle; a torque measurement sensor for measuring the torque of a motor in the electric vehicle; an interior noise measurement sensor for detecting the noise of the electric vehicle; and a signal processing controller for controlling the sound output device to reach a target tone by receiving signals from the torque measurement sensor and the interior noise measurement sensor, and the signal processing controller uses the sound output device to extract and output only an acceleration sound component from an engine sound during acceleration by using driving information related to the electric vehicle input in real time from CAN communication and signals input from the interior noise measurement sensor and the torque measurement sensor, and masks high-frequency noise during deceleration.

The low-pressure injection increasing system comprises an air inlet pipe, an exhaust pipe, an air storage module, a turbocharging module and an air compressor, and the low-pressure injection increasing system is connected with an air inlet of the engine through the air inlet pipe, supplies air to the engine and is connected with an exhaust port of the engine through the exhaust pipe. The turbocharging module is connected with the air inlet pipe and the exhaust pipe, in the running process of the vehicle, supercharged air inlet of the engine can be achieved through tail gas of the engine, horsepower of the engine is increased, and the vehicle is assisted in running, when the vehicle is in the starting or accelerating working condition, under the normal working condition, the air compressor pumps air into the air storage module, and compressed air is stored through the air storage module; when the vehicle is in a starting or accelerating working condition, the air compressor introduces compressed air into the air inlet pipe to increase the air inlet amount and the air inlet pressure of the engine, so that the air inlet pressure of the engine is remarkably improved, vehicle starting is accelerated, and the acceleration capacity of the vehicle is improved.