Abstract: Disclosed are a multi-scale photoacoustic detection method of geological structure around a borehole and related devices. The method includes: obtaining depth information and direction information of the borehole; generating trajectory data of the borehole according to the depth information and direction information; obtaining an optical image of the geological structure around the borehole; generating a first velocity model according to the optical image and the trajectory data; obtaining low-frequency acoustic wave data and high-frequency acoustic wave data of the geological structure around the borehole; performing a full waveform inversion on the first velocity model according to the low-frequency acoustic wave data and the high-frequency acoustic wave data to obtain a second velocity model; and determining the geological structure around the borehole according to the second velocity model.

Abstract: Disclosed is a system and a method of transient electromagnetic advanced detection. The system includes a detection host, an electromagnetic signal transmitter, a probe, and a communication device. The system and method of transient electromagnetic advance detection disclosed by examples of the present disclosure can realize an early-warning through a transient electromagnetic advance detection in a borehole while drilling the borehole in a tunnel or a roadway.

Abstract: A system for pruning weights during training of a neural network includes a configurable pruning hardware unit that is configured to: receive, from a neural network training engine, inputs including the weights, gradients associated with the weights, and a prune indicator per weight; select unpruned weights for pruning; prune the unpruned weights selected for pruning; update the prune indicator per weight for the weights that are selected and pruned; and provide the updated prune indicator to the training engine for the next iteration or epoch. The pruning hardware unit can be configured to perform incremental pruning or non-incremental pruning.

Abstract: Disclosed are a multi-scale photoacoustic detection method of geological structure around a borehole and related devices. The method includes: obtaining depth information and direction information of the borehole; generating trajectory data of the borehole according to the depth information and direction information; obtaining an optical image of the geological structure around the borehole; generating a first velocity model according to the optical image and the trajectory data; obtaining low-frequency acoustic wave data and high-frequency acoustic wave data of the geological structure around the borehole; performing a full waveform inversion on the first velocity model according to the low-frequency acoustic wave data and the high-frequency acoustic wave data to obtain a second velocity model; and determining the geological structure around the borehole according to the second velocity model.

Abstract: Vector-vector multiplication or matrix-matrix multiplication computation on computing systems can include computing a first portion of a vector-vector multiplication product based on a most-significant-bit set of a first vector and a most-significant-bit set of a second vector, and determining if the first portion of the vector-vector multiplication product is less than a threshold. If the first partial vector-vector multiplication product is not less than the threshold, a remaining portion of the vector-vector multiplication product can be computed, and a rectified linear vector-vector multiplication product can be determined for the sum of the first portion of the vector-vector multiplication product and the remaining portion of the vector-vector multiplication product.

Abstract: Disclosed are systems and methods that determine whether instances of data (e.g., forward activations, backward derivatives of activations) that are used to train deep neural networks are to be stored on-chip or off-chip. The disclosed systems and methods are also used to prune the data (discard or delete selected instances of data). A system includes a hierarchical arrangement of on-chip and off-chip memories, and also includes a hierarchical arrangement of data selector devices that are used to decide whether to discard data and where in the system the data is to be discarded.

Abstract: The present disclosure relates to an accelerator for systolic array-friendly data placement. The accelerator may include: a systolic array comprising a plurality of operation units, wherein the systolic array is configured to receive staged input data and perform operations using the staged input to generate staged output data, the staged output data comprising a number of segments; a controller configured to execute one or more instructions to generate a pattern generation signal; a data mask generator; and a memory configured to store the staged output data using the generated masks. The data mask generator may include circuitry configured to: receive the pattern generation signal from the controller, and, based on the received signal, generate a mask corresponding to each segment of the staged output data.

Abstract: The present disclosure provides methods and systems for executing a neural network. The method includes: receiving a plurality of input vectors of input data; generating, among the plurality of input vectors, an estimation value associated with a subset of an input vector based on a weight vector of the activation function; determining whether the estimation value associated with the subset of the input vector satisfies a threshold condition; and determining an output of the activation function based on the estimation value.

Abstract: Disclosed is a system and a method of transient electromagnetic advanced detection. The system includes a detection host, an electromagnetic signal transmitter, a probe, and a communication device. The system and method of transient electromagnetic advance detection disclosed by examples of the present disclosure can realize an early-warning through a transient electromagnetic advance detection in a borehole while drilling the borehole in a tunnel or a roadway.

Abstract: A method and an apparatus for scheduling commands in a virtual computing environment includes picking a command. It is determined whether the command is a synchronization command or a conditional command. A synchronization command is an independent command. A conditional command is a dependent command that depends on a synchronization command. In response to the command being determined as the synchronization command, a waiting queue is enabled for the command, the waiting queue storing conditional commands dependent on a running synchronization command. The command is dispatched to a processing engine.

Abstract: Disclosed is a system and a method for monitoring water inrush, including: a host terminal, a field host, at least one controller and at least one electrode array. Each electrode array is placed in one borehole and to detect electric field signals of surrounding rocks around the borehole; each controller connects with one electrode array and is to control the electrode array to carry out a high-density induced polarization measurement on the surrounding rocks; the field host connects with the controller and is to send control signals to the controller, receive and process the electric field signals output by the electrode array; and the host terminal connects with the field host and is to receive the electric field signals processed by the field host, determine changes on apparent resistivity and apparent chargeability of the surrounding rocks, and determine whether there exists water inrush according to the changes.

Abstract: A method for testing a many-core processor comprises grouping a plurality of cores in the processor into a plurality of super cores, wherein each super core comprises one or more scan chains that propagate through a respective super core. Further, the method comprises grouping the plurality of super cores into a plurality of clusters. The method also comprises comparing one or more scan chain outputs of respective super cores in each cluster using a network of XOR and OR gates to generate a single bit fault signature for each scan chain in a respective cluster and compacting the single bit fault signatures for each scan chain using a hybrid of spatial and temporal compactors to generate a single bit fault signature for each cluster. The method also comprises method of using a cost function to obtain hierarchical parameters to achieve optimized ATPG effort, area overhead and test time.

Abstract: The present disclosure relates to an accelerator for systolic array-friendly data placement. The accelerator may include: a systolic array comprising a plurality of operation units, wherein the systolic array is configured to receive staged input data and perform operations using the staged input to generate staged output data, the staged output data comprising a number of segments; a controller configured to execute one or more instructions to generate a pattern generation signal; a data mask generator; and a memory configured to store the staged output data using the generated masks. The data mask generator may include circuitry configured to: receive the pattern generation signal from the controller, and, based on the received signal, generate a mask corresponding to each segment of the staged output data.

Abstract: Vector-vector multiplication or matrix-matrix multiplication computation on computing systems can include computing a first portion of a vector-vector multiplication product based on a most-significant-bit set of a first vector and a most-significant-bit set of a second vector, and determining if the first portion of the vector-vector multiplication product is less than a threshold. If the first partial vector-vector multiplication product is not less than the threshold, a remaining portion of the vector-vector multiplication product can be computed, and a rectified linear vector-vector multiplication product can be determined for the sum of the first portion of the vector-vector multiplication product and the remaining portion of the vector-vector multiplication product.

Abstract: A method for testing a many-core processor comprises grouping a plurality of cores in the processor into a plurality of super cores, wherein each super core comprises one or more scan chains that propagate through a respective super core. Further, the method comprises grouping the plurality of super cores into a plurality of clusters. The method also comprises comparing one or more scan chain outputs of respective super cores in each cluster using a network of XOR and OR gates to generate a single bit fault signature for each scan chain in a respective cluster and compacting the single bit fault signatures for each scan chain using a hybrid of spatial and temporal compactors to generate a single bit fault signature for each cluster. The method also comprises method of using a cost function to obtain hierarchical parameters to achieve optimized ATPG effort, area overhead and test time.

Abstract: Disclosed is a system and a method for monitoring water inrush, including: a host terminal, a field host, at least one controller and at least one electrode array. Each electrode array is placed in one borehole and to detect electric field signals of surrounding rocks around the borehole; each controller connects with one electrode array and is to control the electrode array to carry out a high-density induced polarization measurement on the surrounding rocks; the field host connects with the controller and is to send control signals to the controller, receive and process the electric field signals output by the electrode array; and the host terminal connects with the field host and is to receive the electric field signals processed by the field host, determine changes on apparent resistivity and apparent chargeability of the surrounding rocks, and determine whether there exists water inrush according to the changes.

Abstract: Embodiments of the present disclosure provide a processing unit, a processor core, a neural network training machine and a method for processing. The method can include: acquiring a compressed weight signal; and decompressing the compressed weight signal into a weight signal and a trimming signal, wherein the weight signal comprises a weight of each neural network node, the trimming signal indicates whether the weight of each neural network node is used in a weight gradient computation, the trimming signal is used for controlling an access to an operand memory storing operands used in the weight computation of one or more neural network nodes corresponding to the operand memory, and the trimming signal is further used for controlling a computing unit to perform weight gradient computation using the weight signal and the operands for the one or more neural network nodes.

Abstract: The present disclosure relates to a method and an apparatus for virtualization. In some embodiments, an exemplary method includes: picking a command; determining whether the command is a synchronization command or a conditional command; in response to the command being determined as the synchronization command, enabling a waiting queue for the command; and dispatching the command to a processing engine.

Abstract: The present disclosure provides methods and systems for executing a neural network. The method includes: receiving a plurality of input vectors of input data; generating, among the plurality of input vectors, an estimation value associated with a subset of an input vector based on a weight vector of the activation function; determining whether the estimation value associated with the subset of the input vector satisfies a threshold condition; and determining an output of the activation function based on the estimation value.

Abstract: Disclosed is a three-dimensional directional transient electromagnetic advanced detection device, wherein the CPU and the bus communication end of the transient electromagnetic transmitting module are both connected to the system bus, the signal output end of the transient electromagnetic transmitting module is connected to the transient electromagnetic transmitting coil outside the borehole to be detected, the signal input end of the electromagnetic signal receiving module is connected to the signal output ends of the three-dimensional magnetic field sensor and the one-dimensional Z-directional electric field sensor, the signal output end of the electromagnetic signal receiving module is connected to the electromagnetic signal input end of the SCM, the communication end of the first memory is connected to the data storage end of the SCM, the communication end of the three-dimensional electronic compass is connected to the compass signal communication end of the SCM, the host data communication of the SCM is