Abstract: Certain aspects provide an apparatus for signal processing in a neural network. The apparatus generally includes computation circuitry configured to perform a convolution operation, the computation circuitry having multiple input rows, and an activation buffer having multiple buffer segments coupled to the multiple the multiple input rows of the computation circuitry, respectively. In some aspects, each of the multiple buffer segments comprises a first multiplexer having a plurality of multiplexer inputs, and each of the plurality of multiplexer inputs of one of the first multiplexers on one of the multiple buffer segments is coupled to a data output of the activation buffer on another one of the multiple buffer segments.

Abstract: Certain aspects of the present disclosure provide a method for processing input data by a set of configurable nonlinear activation function circuits, including generating an exponent output by processing input data using one or more first configurable nonlinear activation function circuits configured to perform an exponential function, summing the exponent output of the one or more first configurable nonlinear activation function circuits, and generating an approximated log softmax output by processing the summed exponent output using a second configurable nonlinear activation function circuit configured to perform a natural logarithm function.

Abstract: Certain aspects of the present disclosure provide techniques for efficient depthwise convolution. A convolution is performed with a compute-in-memory (CIM) array to generate CIM output, and at least a portion of the CIM output corresponding to a first output data channel, of a plurality of output data channels in the CIM output, is written to a digital multiply-accumulate (DMAC) activation buffer. A patch of the CIM output is read from the DMAC activation buffer, and weight data is read from a DMAC weight buffer. Multiply-accumulate (MAC) operations are performed with the patch of CIM output and the weight data to generate a DMAC output.

Abstract: Certain aspects of the present disclosure provide a method for processing input data by a configurable nonlinear activation function circuit, including determining a nonlinear activation function for application to input data; determining, based on the determined nonlinear activation function, a set of parameters for a configurable nonlinear activation function circuit; and processing input data with the configurable nonlinear activation function circuit based on the set of parameters to generate output data.

Abstract: Certain aspects of the present disclosure provide a processor, comprising: a configurable nonlinear activation function circuit configured to: determine, based on a selected nonlinear activation function, a set of parameters for the nonlinear activation function; and generate output data based on application of the set of parameters for the nonlinear activation function, wherein: the configurable nonlinear activation function circuit comprises at least one nonlinear approximator comprising at least two successive linear approximators, and each linear approximator of the at least two successive linear approximators is configured to approximate a linear function using one or more function parameters of the set of parameters.

Abstract: A compute-in-memory bitcell is provided that includes a pair of cross-coupled inverters for storing a stored bit. The compute-in-memory bitcell includes a logic gate formed by a pair of switches for multiplying the stored bit with an input vector bit. A controller controls the pair of switches responsive to a sign bit during a computation phase of operation and controls the pair of switches responsive to a magnitude bit during an execution phase of operation.

Abstract: A compute-in-memory bitcell is provided that includes a pair of cross-coupled inverters for storing a stored bit. The compute-in-memory bitcell includes a logic gate formed by a pair of switches for multiplying the stored bit with an input vector bit. A controller controls the pair of switches responsive to a sign bit during a computation phase of operation and controls the pair of switches responsive to a magnitude bit during an execution phase of operation.

Abstract: A compute-in-memory bitcell is provided that includes a pair of cross-coupled inverters for storing a stored bit. The compute-in-memory bitcell includes a logic gate formed by a pair of switches for multiplying the stored bit with an input vector bit. A controller controls the pair of switches responsive to a sign bit during a computation phase of operation and controls the pair of switches responsive to a magnitude bit during an execution phase of operation.

Abstract: A method includes sending data from a first serial low-power inter-chip media bus (SLIMbus) component to a second SLIMbus component. The method further includes sending the data via at least a first SLIMbus data line of a plurality of SLIMbus data lines.

Abstract: Methods, systems, apparatuses, and computer-readable media for controlling components connected to and/or otherwise associated with a data bus are presented. According to one or more aspects of the disclosure, a plurality of processing devices having data bus management capability and at least one data bus associated with the plurality of processing devices may be identified. Subsequently, an inter-processor communication (IPC) layer for communication between the plurality of processing devices and the at least one data bus may be established over a messaging layer utilized by the at least one data bus. At least one component associated with the at least one data bus may then be controlled via the IPC layer using at least one of the plurality of processing devices.

Abstract: Arrangements for restarting data transmission on a serial low-power inter-chip media bus (SLIMbus) are presented. A clock signal may be provided in an active mode to a component communicatively coupled with the SLIMbus. Immediately prior to the clock signal in the active mode being provided, the clock signal may have been in a paused mode. While the clock signal was in the paused mode at least until the clock signal is provided in the active mode, the data line may have been inactive (e.g., a toggle on the data line may not have been present). Frame synchronization data for a frame may be transmitted. The frame synchronization data for the frame, as received by the component, may not match expected frame synchronization data. Payload data may be transmitted as part of the frame to the component, wherein the payload data is expected to be read properly by the component.

Abstract: Techniques are disclosed for utilizing a non-ported generic device (NGD) or other non-ported hardware to couple processing device(s) to access components on a serial data bus without the need for integrated manager hardware. Using the NGD, a processing device(s) can utilize available unused bandwidth on the serial data bus to communicate with components coupled with the serial data bus, including a processing device having the manager hardware. Various alterations and embodiments are disclosed.

Abstract: Methods, systems, apparatuses, and computer-readable media for controlling components connected to and/or otherwise associated with a data bus are presented. According to one or more aspects of the disclosure, a plurality of processing devices having data bus management capability and at least one data bus associated with the plurality of processing devices may be identified. Subsequently, an inter-processor communication (IPC) layer for communication between the plurality of processing devices and the at least one data bus may be established over a messaging layer utilized by the at least one data bus. At least one component associated with the at least one data bus may then be controlled via the IPC layer using at least one of the plurality of processing devices.

Abstract: Techniques are disclosed for utilizing a non-ported generic device (NGD) or other non-ported hardware to couple processing device(s) to access components on a serial data bus without the need for integrated manager hardware. Using the NGD, a processing device(s) can utilize available unused bandwidth on the serial data bus to communicate with components coupled with the serial data bus, including a processing device having the manager hardware. Various alterations and embodiments are disclosed.

Abstract: Arrangements for restarting data transmission on a serial low-power inter-chip media bus (SLIMbus) are presented. A clock signal may be provided in an active mode to a component communicatively coupled with the SLIMbus. Immediately prior to the clock signal in the active mode being provided, the clock signal may have been in a paused mode. While the clock signal was in the paused mode at least until the clock signal is provided in the active mode, the data line may have been inactive (e.g., a toggle on the data line may not have been present). Frame synchronization data for a frame may be transmitted. The frame synchronization data for the frame, as received by the component, may not match expected frame synchronization data. Payload data may be transmitted as part of the frame to the component, wherein the payload data is expected to be read properly by the component.