PROCESSOR

BACKGROUND

1. Field of the Disclosure

[0001] The present disclosure relates a circuit for implementing a neural network and more specifically to managing for neural network tasks.

2. Description of the Related Arts

[0001] An artificial neural network (ANN) is a computing system or model that uses a collection of connected nodes to process input data. The ANN is typically organized into layers where different layers perform different types of transformation on their input. Extensions or variants of ANN such as convolution neural network (CNN), deep neural networks (DNN), recurrent neural networks (RNN) and deep belief networks (DBN) have come to receive much attention. These computing systems or models often involve extensive computing operations including multiplication and accumulation. For example, CNN is a class of machine learning technique that primarily uses convolution between input data and kernel data, which can be decomposed into multiplication and accumulation operations.

[0002] Depending on the types of input data and operations to be performed, these machine learning systems or models can be configured differently. Such varying configuration would include, for example, pre-processing operations, number of channels in input data, kernel data to be used, non-linear function to be applied to convolution result, and applying of various post processing operations. Using a central processing unit (CPU) and its main memory to instantiate and execute machine learning systems or models of various configuration is relatively easy because such systems or models can be instantiated with mere updates to code. However, relying solely on the CPU for various operations of these machine learning systems or models would consume significant bandwidth of a central processing unit (CPET) as well as increase the overall power consumption.

SUMMARY

[0003] Embodiments relate to managing tasks that when executed by a neural processor circuit instantiates a neural network. The neural processor circuit includes one or more neural engine circuits and a neural task manager circuit. The neural task manager circuit includes multiple task queue circuits and a task arbiter circuit. Each task queue circuit stores a reference to a task list of tasks for a machine learning operation. Each task list and task queue circuit may be associated with a priority parameter. Based on the priority of the task queue circuits, the task arbiter circuit retrieves configuration data for a task from an external memory, and provides the configuration data to components of the neural processor circuit including the one or more neural engines. The configuration data programs the neural processor circuit to execute the task.

[0004] Some embodiments include a method of managing tasks in a neural processor circuit. A reference to a task list of tasks that instantiates a neural network is stored in a task queue circuit of the neural processor circuit. The task list includes configuration data of a task stored in a location of an external memory. The task list is retrieved from the task queue circuit, such as based on task priority. Responsive to retrieving the task list, the configuration data is retrieved from the location of the external memory. At least a portion of the configuration data is provided to a neural engine of the neural processor circuit to program the neural engine to execute the task.

BRIEF DESCRIPTION OF THE DRAWINGS

[0005] Figure (FIG.) l is a high-level diagram of an electronic device, according to one embodiment

[0006] FIG. 2 is a block diagram illustrating components in the electronic device, according to one embodiment.

[0007] FIG. 3 is a block diagram illustrating a neural processor circuit, according to one embodiment.

[0008] FIG. 4 is a block diagram of a neural engine in the neural processor circuit, according to one embodiment. [0009] FIG. 5 is a conceptual diagram illustrating loops for processing input data at the neural processor circuit, according to one embodiment.

[0010] FIG. 6 is a conceptual diagram illustrating segmenting the input data into slices, tiles and work units, according to one embodiment.

[0011] FIG. 7 is a diagram illustrating programming of rasterizers in components of the neural processor circuit, according to one embodiment.

[0012] FIG. 8 is a flowchart illustrating a method of processing input data in a neural processor circuit, according to one embodiment.

[0013] FIG. 9 is a schematic block diagram illustrating a neural network represented by a list of tasks, according to one embodiment.

[0014] FIG. 10 is a block diagram illustrating the neural task manager, according to one embodiment.

[0015] FIG. 11 is a diagram illustrating retrieval of task descriptors using a task queue, according to one embodiment.

[0016] FIG. 12 is a diagram illustrating a task descriptor, according to one embodiment.

[0017] FIG. 13 is a block diagram illustrating a fetch queue and a configuration queue of a neural task manager, according to one embodiment.

[0018] FIG. 14 is a flowchart illustrating a method of managing tasks in a neural processor circuit, according to one embodiment.

[0019] The figures depict, and the detail description describes, various non-limiting embodiments for purposes of illustration only.

DETAILED DESCRIPTION

[0020] Reference will now be made in detail to embodiments, examples of which are illustrated in the accompanying drawings. In the following detailed description, numerous specific details are set forth in order to provide a thorough understanding of the various described embodiments. However, the described embodiments may be practiced without these specific details. In other instances, well-known methods, procedures, components, circuits, and networks have not been described in detail so as not to unnecessarily obscure aspects of the embodiments.

[0021] Embodiments of the present disclosure relate to managing tasks that when executed by a neural processor circuit instantiates a neural network. A machine learning operation, such as an inferencing operation or a training operation, is defined by a task list of tasks. The neural processor circuit includes one or more neural engines and a neural task manager. The neural task manager includes multiple task queues and a task arbiter. Each task queue stores a task list of tasks for a machine learning operation. Each task list or task queue may be associated with a priority parameter. The task arbiter retrieves configuration data for a task from an external memory based on the priority parameters, and provides the configuration data to components of the neural processor circuit including the one or more neural engines. In some embodiments, the neural task manager includes a configuration queue that stores configuration data of committed tasks selected by the task arbiter, and provides the configuration data to other components of the neural processor circuit. The configuration data programs the neural processor circuit to execute the task. For example, the configuration data may include input data and kernel data processed by a neural engine to execute the task. The configuration data may further include instructions for retrieving and handling the configuration data, and instructions for storing output data of the neural engine. Among other things, the neural task manager allows the neural processor circuit to efficiently handle multiple machine learning operations. Furthermore, the neural task manager may facilitate task switching when a higher priority task is stored in a task queue while a lower priority task is being executed.

[0022] A“task” described herein refers to a processing operation of the neural processor circuit that instantiates a network layer of a neural network, multiple network layers of a neural network, or a portion of a network layer of a neural network. A“task list” described herein refers to a sequence of tasks, such as a sequence of tasks that executed by the neural processor circuit instantiates multiple network layers of a neural network.

Exemplary Electronic Device

[0023] Embodiments of electronic devices, user interfaces for such devices, and associated processes for using such devices are described. In some embodiments, the device is a portable communications device, such as a mobile telephone, that also contains other functions, such as personal digital assistant (PDA) and/or music player functions. Exemplary embodiments of portable multifunction devices include, without limitation, the iPhone®, iPod Touch®, Apple Watch®, and iPad® devices from Apple Inc. of Cupertino, California. Other portable electronic devices, such as wearables, laptops or tablet computers, are optionally used. In some

embodiments, the device is not a portable communications device, but is a desktop computer or other computing device that is not designed for portable use. In some embodiments, the disclosed electronic device may include a touch sensitive surface (e.g., a touch screen display and/or a touch pad). An example electronic device described below in conjunction with FIG. 1 (e.g., device 100) may include a touch-sensitive surface for receiving user input. The electronic device may also include one or more other physical user-interface devices, such as a physical keyboard, a mouse and/or a joystick.

[0024] Figure (FIG.) 1 is a high-level diagram of an electronic device 100, according to one embodiment. Device 100 may include one or more physical buttons, such as a“home” or menu button 104. Menu button 104 is, for example, used to navigate to any application in a set of applications that are executed on device 100. In some embodiments, menu button 104 includes a fingerprint sensor that identifies a fingerprint on menu button 104. The fingerprint sensor may be used to determine whether a finger on menu button 104 has a fingerprint that matches a fingerprint stored for unlocking device 100. Alternatively, in some embodiments, menu button 104 is implemented as a soft key in a graphical user interface (GUI) displayed on a touch screen.

[0025] In some embodiments, device 100 includes touch screen 150, menu button 104, push button 106 for powering the device on/off and locking the device, volume adjustment buttons 108, Subscriber Identity Module (SIM) card slot 110, head set jack 112, and docking/charging external port 124. Push button 106 may be used to turn the power on/off on the device by depressing the button and holding the button in the depressed state for a predefined time interval; to lock the device by depressing the button and releasing the button before the predefined time interval has elapsed; and/or to unlock the device or initiate an unlock process. In an alternative embodiment, device 100 also accepts verbal input for activation or deactivation of some functions through microphone 113. The device 100 includes various components including, but not limited to, a memory (which may include one or more computer readable storage mediums), a memory controller, one or more central processing units (CPUs), a peripherals interface, an RF circuitry, an audio circuitry, speaker 111, microphone 113, input/output (I/O) subsystem, and other input or control devices. Device 100 may include one or more image sensors 164, one or more proximity sensors 166, and one or more accelerometers 168. The device 100 may include components not shown in FIG. 1.

[0026] Device 100 is only one example of an electronic device, and device 100 may have more or fewer components than listed above, some of which may be combined into a components or have a different configuration or arrangement. The various components of device 100 listed above are embodied in hardware, software, firmware or a combination thereof, including one or more signal processing and/or application specific integrated circuits (ASICs).

[0027] FIG. 2 is a block diagram illustrating components in device 100, according to one embodiment. Device 100 may perform various operations including image processing. For this and other purposes, the device 100 may include, among other components, image sensor 202, system-on-a chip (SOC) component 204, system memory 230, persistent storage (e.g., flash memory) 228, motion (orientation) sensor 234, and display 216. The components as illustrated in FIG. 2 are merely illustrative. For example, device 100 may include other components (such as speaker or microphone) that are not illustrated in FIG. 2. Further, some components (such as motion sensor 234) may be omitted from device 100.

[0028] Image sensor 202 is a component for capturing image data and may be embodied, for example, as a complementary metal-oxide-semiconductor (CMOS) active-pixel sensor a camera, video camera, or other devices. Image sensor 202 generates raw image data that is sent to SOC component 204 for further processing. In some embodiments, the image data processed by SOC component 204 is displayed on display 216, stored in system memory 230, persistent storage 228 or sent to a remote computing device via network connection. The raw image data generated by image sensor 202 may be in a Bayer color kernel array (CFA) pattern (hereinafter also referred to as“Bayer pattern”).

[0029] Motion sensor 234 is a component or a set of components for sensing motion of device 100. Motion sensor 234 may generate sensor signals indicative of orientation and/or acceleration of device 100. The sensor signals are sent to SOC component 204 for various operations such as turning on device 100 or rotating images displayed on display 216.

[0030] Display 216 is a component for displaying images as generated by SOC component 204. Display 216 may include, for example, liquid crystal display (LCD) device or an organic light emitting diode (OLED) device. Based on data received from SOC component 204, display 216 may display various images, such as menus, selected operating parameters, images captured by image sensor 202 and processed by SOC component 204, and/or other information received from a user interface of device 100 (not shown).

[0031] System memory 230 is a component for storing instructions for execution by SOC component 204 and for storing data processed by SOC component 204. System memory 230 may be embodied as any type of memory including, for example, dynamic random access memory (DRAM), synchronous DRAM (SDRAM), double data rate (DDR, DDR2, DDR3, etc.) RAMBUS DRAM (RDRAM), static RAM (SRAM) or a combination thereof. In some embodiments, system memory 230 may store pixel data or other image data or statistics in various formats.

[0032] Persistent storage 228 is a component for storing data in a non-volatile manner.

Persistent storage 228 retains data even when power is not available. Persistent storage 228 may be embodied as read-only memory (ROM), flash memory or other non-volatile random access memory devices.

[0033] SOC component 204 is embodied as one or more integrated circuit (IC) chip and performs various data processing processes. SOC component 204 may include, among other subcomponents, image signal processor (ISP) 206, a central processor unit (CPU) 208, a network interface 210, sensor interface 212, display controller 214, neural processor circuit 218, graphics processor (GPU) 220, memory controller 222, video encoder 224, storage controller 226, and bus 232 connecting these subcomponents. SOC component 204 may include more or fewer subcomponents than those shown in FIG. 2.

[0034] ISP 206 is hardware that performs various stages of an image processing pipeline. In some embodiments, ISP 206 may receive raw image data from image sensor 202, and process the raw image data into a form that is usable by other subcomponents of SOC component 204 or components of device 100. ISP 206 may perform various image-manipulation operations such as image translation operations, horizontal and vertical scaling, color space conversion and/or image stabilization transformations, as described below in detail with reference to FIG. 3.

[0035] CPU 208 may be embodied using any suitable instruction set architecture, and may be configured to execute instructions defined in that instruction set architecture. CPU 208 may be general-purpose or embedded processors using any of a variety of instruction set architectures (ISAs), such as the x86, PowerPC, SPARC, RISC, ARM or MIPS ISAs, or any other suitable ISA. Although a single CPU is illustrated in FIG. 2, SOC component 204 may include multiple CPUs. In multiprocessor systems, each of the CPUs may commonly, but not necessarily, implement the same ISA.

[0036] Graphics processing unit (GPU) 220 is graphics processing circuitry for performing graphical data. For example, GPU 220 may render objects to be displayed into a frame buffer (e.g., one that includes pixel data for an entire frame). GPU 220 may include one or more graphics processors that may execute graphics software to perform a part or all of the graphics operation, or hardware acceleration of certain graphics operations.

[0037] Neural processor circuit 218 is a circuit that performs various machine learning operations based on computations including multiplication, addition and accumulation. Such computations may be arranged to perform, for example, convolution of input data and kernel data. Neural processor circuit 218 is a configurable circuit that performs these operations in a fast and power-efficient manner while relieving CPU 208 of resource-intensive operations associated with neural network operations. Neural processor circuit 218 may receive the input data from sensor interface 212, the image signal processor 206, system memory 230 or other sources such as network interface 210 or GPU 220. The output of neural processor circuit 218 may be provided to various components of device 100 such as the image signal processor 206, system memory 230 or CPU 208 for various operations. The structure and operation of neural processor circuit 218 is described below in detail with reference to FIG. 3.

[0038] Network interface 210 is a subcomponent that enables data to be exchanged between devices 100 and other devices via one or more networks (e.g., carrier or agent devices). For example, video or other image data may be received from other devices via network interface 210 and be stored in system memory 230 for subsequent processing (e.g., via a back-end interface to image signal processor 206, such as discussed below in FIG. 3) and display. The networks may include, but are not limited to, Local Area Networks (LANs) (e.g., an Ethernet or corporate network) and Wide Area Networks (WANs). The image data received via network interface 210 may undergo image processing processes by ISP 206.

[0039] Sensor interface 212 is circuitry for interfacing with motion sensor 234. Sensor interface 212 receives sensor information from motion sensor 234 and processes the sensor information to determine the orientation or movement of the device 100.

[0040] Display controller 214 is circuitry for sending image data to be displayed on display 216. Display controller 214 receives the image data from ISP 206, CPU 208, graphic processor or system memory 230 and processes the image data into a format suitable for display on display 216.

[0041] Memory controller 222 is circuitry for communicating with system memory 230. Memory controller 222 may read data from system memory 230 for processing by ISP 206, CPU 208, GPU 220 or other subcomponents of SOC component 204. Memory controller 222 may also write data to system memory 230 received from various subcomponents of SOC component 204.

[0042] Video encoder 224 is hardware, software, firmware or a combination thereof for encoding video data into a format suitable for storing in persistent storage 128 or for passing the data to network interface 210 for transmission over a network to another device.

[0043] In some embodiments, one or more subcomponents of SOC component 204 or some functionality of these subcomponents may be performed by software components executed on ISP 206, CPU 208 or GPU 220. Such software components may be stored in system memory 230, persistent storage 228 or another device communicating with device 100 via network interface 210.

[0044] Image data or video data may flow through various data paths within SOC component 204. In one example, raw image data may be generated from the image sensor 202 and processed by ISP 206, and then sent to system memory 230 via bus 232 and memory controller 222. After the image data is stored in system memory 230, it may be accessed by video encoder 224 for encoding or by display 116 for displaying via bus 232.

Example Neural processor circuit

[0045] Neural processor circuit 218 is a configurable circuit that performs neural network operations on the input data based at least on kernel data 340. For this purpose, neural processor circuit 218 may include, among other components, neural task manager 310, a plurality of neural engines 314A through 314N (hereinafter collectively referred as“neural engines 314” and individually also referred to as“neural engine 314”), kernel direct memory access (DMA) 324, data buffer 318 and buffer DMA 320. Neural processor circuit 218 may include other components not illustrated in FIG. 3.

[0046] Each of neural engines 314 performs computing operations for neural network operations in parallel. Depending on the load of operation, entire set of neural engines 314 may be operated or only a subset of the neural engines 314 may be operated while the remaining neural engines 314 are placed in a power save mode to conserve power. Each of neural engines 314 includes components for storing one or more kernels, for performing multiply-accumulate operations, and for post-processing to generate an output data 328, as described below in detail with reference to FIG. 4. One example of a neural network operation is a convolution operation. [0047] Neural task manager 310 manages the overall operation of neural processor circuit 218. Neural task manager 310 may receive a task list from a compiler executed by CPU 208, store tasks in its task queues, choose a task to perform, and send instructions to other components of the neural processor circuit 218 for performing the chosen task. Neural task manager 310 may also perform switching of tasks on detection of events such as receiving instructions from CPU 208. In one or more embodiments, the neural task manager 310 sends rasterizer information to the components of the neural processor circuit 218 to enable each of the components to track, retrieve or process appropriate portions of the input data and kernel data, as described below in detail with reference to FIGS. 5 through 7. Although neural task manager 310 is illustrated in FIG. 3 as part of neural processor circuit 218, neural task manager 310 may be a component outside the neural processor circuit 218.

[0048] Kernel DMA 324 is a read circuit that fetches kernel data from a source (e.g., system memory 230) and sends kernel data 326 A through 326N to each of the neural engines 314.

Kernel data represents information from which kernel elements can be extracted. In one embodiment, the kernel data may be in a compressed format which is decompressed at each of neural engines 314. Although kernel data provided to each of neural engines 314 may be the same in some instances, the kernel data provided to each of neural engines 314 is different in most instances.

[0049] Data buffer 318 is a temporary storage for storing data associated with the neural network operations. In one embodiment, data buffer 318 is embodied as a memory that can be accessed by all of the neural engines 314. Data buffer 318 may store input data 322A through 322N for feeding to corresponding neural engines 314A through 314N, as well as output from each of neural engines 314A through 314N for feeding back into neural engines 314 or sending to a target circuit (e.g., system memory 230). The operations of data buffer 318 and other components of the neural processor circuit 218 are coordinated so that the input data and intermediate data stored in the data buffer 318 is reused across multiple operations at the neural engines 314, and thereby reduce data transfer to and from system memory 230. Data buffer 318 may be operated in a broadcast mode where input data of all input channels are fed to all neural engines 314 or in a unicast mode where input data of a subset of input channels are fed to each neural engine 314. [0050] The input data 322 stored in data buffer 318 can be part of, among others, image data, histogram of oriented gradients (HOG) data, audio data, meta data, output data 328 of a previous cycle of the neural engine 314, and other processed data received from other components of the SOC component 204.

[0051] Buffer DMA 320 includes a read circuit that receives a portion (e.g., tile) of the input data from a source (e.g., system memory 230) for storing in data buffer 318, and a write circuit that forwards data from data buffer 318 to a target (e.g., system memory).

Example Neural Engine Architecture

[0052] FIG. 4 is a block diagram of the neural engine 314, according to one embodiment.

The neural engine 314 performs various operations to facilitate neural network operations such as convolution, spatial pooling and local response normalization. The neural engine 314 receives the input data 322, performs multiply-accumulate operations (e.g., convolution operations) on the input data 322 based on stored kernel data, performs further post-processing operations on the result of the multiply-accumulate operations, and generates the output data 328. The input data 322 and/or the output data 328 of the neural engine 314 may be of a single channel or multiple channels.

[0053] Neural engine 314 may include, among other components, input buffer circuit 402, computation core 416, neural engine (NE) control 418, kernel extract circuit 432, accumulators 414 and output circuit 424. Neural engine 314 may include further components not illustrated in FIG. 4.

[0054] Input buffer circuit 402 is a circuit that stores a portion of the input data 322 as it is received from the data buffer 318 and sends an appropriate portion 408 of input data for a current task or process loop to computation core 416 for processing. Input buffer circuit 402 includes a shifter 410 that shifts read locations of input buffer circuit 402 to change the portion 408 of input data sent to computation core 416. By changing portions of input data provided to the computation core 416 via shifting, neural engine 314 can perform multiply-accumulate for different portions of input data based on fewer number of read operations. In one or more embodiments, the input data 322 includes data of difference convolution groups and/or input channels.

[0055] Kernel extract circuit 432 is a circuit that receives kernel data 326 from kernel DMA 324 and extracts kernel coefficients 422. In one embodiment, the kernel extract circuit 432 references a look up table (LUT) and uses a mask to reconstruct a kernel from compressed kernel data 326. The mask indicates locations in the reconstructed kernel to be padded with zero and remaining locations to be filled with numbers. The kernel coefficients 422 of the reconstructed kernel are sent to computation core 416 to populate register in multiply-add (MAD) circuits of computation core 416. In other embodiments, the kernel extract circuit 432 receives kernel data in an uncompressed format and the kernel coefficients are determined without referencing a LUT or using a mask.

[0056] Computation core 416 is a programmable circuit that performs computation operations. For this purpose, the computation core 416 may include MAD circuits MADO through MADN and a post-processor 428. Each of MAD circuits MADO through MADN may store an input value in the portion 408 of the input data and a corresponding kernel coefficient in the kernel coefficients 422. The input value and the corresponding kernel coefficient are multiplied in each of MAD circuits to generate a processed value 412.

[0057] Accumulator 414 is a memory circuit that receives and stores processed values 412 from MAD circuits. The processed values stored in accumulator 414 may be sent back as feedback information 419 for further multiply and add operations at MAD circuits or sent to post-processor 428 for post-processing. Accumulator 414 in combination with MAD circuits form a multiply-accumulator (MAC) 404. In one or more embodiments, accumulator 414 may have subunits where each subunit sends data to different components of neural engine 314. For example, during a processing cycle, data stored in a first subunit of accumulator 414 is sent to MAC circuit while data stored in a second subunit of accumulator 414 is sent to post-processor 428.

[0058] Post-processor 428 is a circuit that performs further processing of values 412 received from accumulator 414. The post-processor 428 may perform operations including, but not limited to, applying nonlinear functions (e.g., Rectified Linear Unit (ReLU)), normalized cross- correlation (NCC), merging the results of performing neural operations on 8-bit data into 16-bit data, and local response normalization (LRN). The result of such operations is output from the post-processor 428 as processed values 417 to output circuit 424.

[0059] NE control 418 controls operations of other components of the neural engine 314 based on the operation modes and parameters of neural processor circuit 218. Depending on different modes of operation (e.g., group convolution mode or non-group convolution mode) or parameters (e.g., the number of input channels and the number of output channels), neural engine 314 may operate on different input data in different sequences, return different values from accumulator 414 to MAC circuits, and perform different types of post-processing operations at post processor 428. To configure components of the neural engine 314 to operate in a desired manner, the NE control 418 sends a control signal to components of the neural engine. NE control 418 may also include rasterizer 430 that tracks the current task or process loop being processed at neural engine 314, as described below in detail with reference to FIG. 5 through 7.

[0060] Output circuit 424 receives processed values 417 from the post-processor 428 and interfaces with data buffer 318 to store processed values 417 in data buffer 318. For this purpose, output circuit 424 may send out as output data 328 in a sequence or a format that is different from the sequence or format in which the processed values 417 are processed in post- processor 428.

[0061] The components in the neural engine 314 may be configured during a configuration period by the NE control 418 and the neural task manager 310. For this purpose, the neural task manager 310 sends configuration information to the neural engine 314 during the configuration period. The configurable parameters and modes may include, but are not limited to, mapping between input data elements and kernel elements, the number of input channels, the number of output channels, performing of output strides, and enabling /selection of post-processing operations at the post processor 428.

Operation of Segmenting of Data for Processing at Neural Processor Circuit

[0062] Input data is typically split into smaller pieces of data for parallel processing at multiple neural engines 314. Often multiple cycles of operations are performed to generate output for a task associated with a neural network. A compiler executed by CPET 208 analyzes the hierarchy and nodes of the neural network and determines how the input data is to be segmented based on the hardware constraints of the neural processor circuit 218. One of functions of the compiler is to determine how input data is to be split into smaller data units for processing at the neural engines 314, and how the processing is to be iterated in loops to produce the result for tasks.

[0063] FIG. 5 is a conceptual diagram illustrating loops for processing the input data at neural processor circuit 218, according to one embodiment. The outermost loop represents processing for a convolution group, if group convolution involving multiple convolution group is used. Group convolutions are convolutions where input data of the input channels in each group are used only for generating output data of output channels of each group but are not used for generating output data for output channels of other groups. Hence, each group of the group convolution can be treated as a separate convolution operation.

[0064] In the loop for each convolution group is a processing loop for a slice of the input data. The entire input data for a convolution operation is segmented into multiple strips of slices in an overlapping manner, as shown in FIG. 6. The overlapping portions 602, 604, 606 are parts of the input data that are overfetched in two adjacent slices to provide spatial support for a corresponding kernel. The second outermost loop performs convolution operation for each slice in the input data. Within the loop for a slice is a processing loop for a tile of the slice. Each slice is segmented into a plurality of tiles, as shown in FIG. 6. The overlapping portions 608, 610,

612, 614 are parts of the input data in slice 4 that are overfetched in two adjacent tiles to provide spatial support for a corresponding kernel. The rightmost tile will typically have a width smaller than other tiles of the slice. In one embodiment, input data for each tile is loaded onto data buffer 318 in a read cycle and reused for operations in processing loops for the tile. In the processing loop for the tile is a processing loop for a work unit. Each tile is segmented into multiple work units as shown in FIG. 6. A work unit is a portion of the input data having a size that produces output values that fit into accumulator 414 of neural engine 314 during a single cycle of the computation core 416. Although the shape of each work unit is shown as a horizontal strip in FIG. 6, the shape of the work unit can be different depending on the shape and size of the tile. The work units also have overlapping parts that represent overfetched data to provide support for a corresponding kernel. Especially, work units for the last tile of a slice may have a shape of a vertical strip if the tile is tall. In one or more embodiments, the size of each work unit is 256 bytes. In such embodiments, for example, work units be shaped to one of 16 x 16, 32 x 8, 64 x 4, 128 x 2 or 256 x 1 dimension.

[0065] For each work unit, an internal processing loop may be provided for an output channel group (OCG). The number of output channels produced for a given work unit by a single cycle of the computation core 416 is referred to as an OCG. Depending on operation modes, each neural engine 314 may process output data of different numbers of output channels (e.g., 8 channels, 32 channels) for a single load of input data into its input buffer circuit 402. [0066] For each output channel group, an internal processing loop may be provided for an input channel (Cin). If an input stride is implemented to skip certain input data, loops for sub- input channels (Sub-Cin) may be provided within the processing loop for the input channel (Cin).

[0067] For each input channel or each sub-input channel, internal loops are provided for processing horizontal spatial support for a kernel and the vertical support within each horizontal spatial support. The spatial support refers to the input data for convolution with the kernel, and includes overfetched input data for performing convolution at the edges of the input data.

[0068] Overfetch refers to fetching additional input data in current slice, tile or work unit so that proper dimension of input data can be provided for convolution with a kernel. In one or more embodiments, overfetch is performed vertically between slices to obtain additional rows of input data (shown as overlapping portions 602, 604, 606 in FIG. 6), horizontally between tiles to obtain additional columns of input data (shown as overlapping portions 608, 610, 612, 614 in FIG. 6), and vertically between work units within a tile to obtain additional rows of input data.

[0069] For each spatial support for the kernel, an internal processing loop for an output channel (OC) is provided to generate output data for each output channel (Cout). In cases where output stride implements a spatial upsampling, an additional inner loop for processing each sub- output channel is provided. Loading of kernel coefficients and MAC operations are performed within the loop for the output channel (OC) or sub-output channel if an output stride is implemented, to generate output data for the output channel (OC) or sub-output channel.

[0070] The nested loop structure of FIG. 5 is merely illustrative. Loops may be omitted, added or structured differently depending on various factors. For example, if only a single convolution group is used, the outermost loop may be removed. Further, the loop structure for the horizontal spatial support and the vertical spatial support may be reversed.

[0071] In one or more embodiments, the operations associated dividing the input space into smaller units and processing these smaller units as described above with reference to FIGS. 5 and 6 are performed by rasterizers 714, 718, 720, 722 in various components of neural processor circuit 218. A rasterizer is a circuit in various components of neural processor circuit 218 that keeps track of the segment of the input/output data (e.g., group, work unit, input channel, output channel) and instructs the components of neural processor circuit for proper handling of the segment of the input data. For example, rasterizer 720 in buffer DMA 320 tracks tiles and slices received from system memory 230 while rasterizer 718 in data buffer 318 broadcasts in sequence work units for processing by the neural engines 314. Rasterizer 722 in kernel DMA 322 determines which kernels are to be received and distributed to neural engines 314, while rasterizers 714 in neural engines 314 operate shifters 410 in input buffer circuits 402 to forward correct portions 408 of input data to MAC 404, and send the finished output data 328 to the data buffer 318.

[0072] FIG. 7 is a diagram illustrating programming of rasterizers 714, 718, 720, 722 in components 314, 318, 320, 322 of the neural processor circuit 218, according to one

embodiment. To perform their functions, each of rasterizers 714, 718, 720, 722 receives task information 710 (e.g., configuration data) indicating how the input data and/or kernel data are to be segmented and to be handled by each component of the neural processor circuit 218. The task information includes information about particulars of the current layer (e.g., dimensions of input and output data, dimension of an associated kernel, types of padding at the boundaries of input data). Rasterizers 714, 718, 720, and 722 may also receive constraints on their operations (e.g., whether to allow or disallow tile width over a threshold).

[0073] By providing rasterizers in different components of neural processor circuit 218, overhead in data transmitted between the components of the neural processor circuit 218 may be reduced. If a single central rasterizer is provided to control different components of the neural processor circuit 218, kernel data, input data, and output data transmitted between the components may be needed in these data to identify associated position in the loops of the task such as convolution group, tile, slice, work unit, input channel and output channel. By using distributed rasterizers, no separate metadata is needed to transmit the kernel data, input data and output data among components of the neural processor circuit 218.

Example Process at Neural engine Architecture

[0074] FIG. 8 is a flowchart illustrating a method of processing input data in neural processor circuit 218, according to one embodiment. The method may include different and/or additional steps, or the steps may be in different orders.

[0075] After neural task manager 310 programs rasterizers 714, 718, 720, 722, the process of operating buffer DMA 320 is initiated by rasterizer 720 instructing 804 buffer DMA 320 to cause buffer DMA 320 to receive a tile of input data from system memory 230. The tile received by buffer DMA 320 is stored 806 in data buffer 318. [0076] Rasterizer 718 in data buffer 318 then instructs 808 data buffer 318 to send a work unit to one or more neural engines 314. The work unit is then stored in input buffer circuits 402 of the one or more neural engines 314.

[0077] In one or more embodiments, input buffer circuit 402 selects 816 a portion of work unit to be sent to MAC 404 to perform multiply-accumulate operation. Then MAC 404 performs 820 multiply-accumulate operations on the selected portion of the work unit using a

corresponding kernel. Then it is determined 824 if the entire work unit is processed at one or more neural engines 314. If not, the selected portion of the work unit is shifted 828 by shifter 410 and returns to perform 820 another round of multiply-accumulate operations.

[0078] If it is determined 824 that the entire work unit was processed, then it proceeds to determine 832 if all work units in the tile was processed. If not, then the process proceeds 836 to the next work unit by having data buffer 318 send 808 a next work unit to one or more neural engines 314, and repeats the subsequent processes.

[0079] If it is determined 832 that all work units in the tile was processed by the neural engines 314, the process proceeds to determine 840 whether all tiles for the input data were processed. If not, the process proceeds 844 to a next tile by having rasterizer 720 instructs 804 buffer DMA 320 to receive a next tile from system memory 230 and repeats the subsequent processes.

[0080] If it is determined 840 that all tiles of the input data are processed, then the process ends for the current input data. Then, the process may be repeated to process the next input data or proceed to the next task.

[0081] Embodiments of the process as described above with reference to FIG. 8 are merely illustrative. Further loops may be embodied, as described above with reference to FIG. 5.

Moreover, sequence of the process may be modified or omitted.

Example Neural Task Manager Architecture

[0082] A neural network may include network layers or sub-layers that are instantiated or implemented as a series of tasks executed by the neural processor circuit 218. FIG. 9 is a schematic block diagram illustrating a neural network 900 represented by a list 904 of tasks, according to one embodiment. The neural network 900 includes network layers (or sub-layers) including convolution layers Cl, C2, C3 (including sub-layers C3oo, C3io, C3n, C3 ₂ o, and C3 _2I ), C5, and C6, and pooling layers P2 and P4. The neural network 900 is an example of a neural network architecture that may be instantiated by the neural processor circuit 218. That is, when the neural network 900 is converted into the task list 904 to become executable by the neural processor c rcuit 218. Other types of neural network architectures with different types of network layers or orders of network layers may also be instantiated by the neural processor circuit 218.

[0083] The neural network 900 is converted, such as by the CPU 208, to the task list 904.

The task list 904 includes a linear linkdist defining a sequence of tasks including task Cl, task C2+P2, task C3oo+P4, task C3io, task C3n+P4, task C3 ₂ o, task C3 _2I +P4, task C5 _a , task C5b, and task C6. Each task is associated with a task descriptor that defines a configuration of the neural processor circuit 218 to execute the task. Each task may correspond with a single network layer of the neural network 900, a portion of a network layer of the neural network 900, or multiple network layers of the neural network 900. For example, the task Cl corresponds with a single network layer Cl, the task C2+P2 corresponds with multiple network layers C2 and P2, and the tasks C5 _a and C5b each correspond with a portion of the network layer C5. The neural processor circuit 218 instantiates the neural network 900 by executing the tasks of the task list 904 under the control of the neural task manager 310.

[0084] FIG. 10 is a block diagram illustrating the neural task manager 310, according to one embodiment. The neural task manager 310 manages the execution of tasks for one or more neural networks 900 by the neural processor circuit 218. The neural task manager 310 may include, among other components, a task arbiter 1002, task queues 1004 A through 1004N (hereinafter collectively referred as“task queues 1004” and individually also referred to as“task queue 1004”), a task manager direct memory access (DMA) 1006, a fetch queue 1008, and a configuration queue 1010. The neural task manager 310 may include other components not illustrated in FIG. 10.

[0085] The task arbiter 1002 is a circuit or a combination of circuit and firmware that selects tasks from the task queues 1004 for execution by the neural processor circuit 218. The task arbiter 1002 dequeues tasks from the task queues 1004, and places tasks in the configuration queue 1010. While a task is in a configuration queue, it is committed to execution and the neural processor circuit performs a prefetch for input data and kernel data before the task is executed by other components of the neural processor circuit 218. For example, the task arbiter 1002 may perform fixed-priority arbitration between multiple task queues 1004, and select the task from the task queues 1004 with the highest priority for retrieval of a task descriptor 1012 from the system memory 230 by the task manager DMA 1006.

[0086] The neural task manager 310 may include one or more task queues 1004. Each task queue 1004 is coupled to the CPU 208 and the task arbiter 1002. Each task queue 1004 receives from the CPU 208 a reference to a task list 904 of tasks that when executed by the neural processor circuit 218 instantiates a neural network 900. The reference stored in each task queue 1004 may include a set of pointers and counters pointing to the task list 904 of the task descriptors 1012 in the system memory 230. Each task queue 1004 may be further associated with a priority parameter that defines the relative priority of the task queues 1004. The task descriptor 1012 of a task specifies a configuration of the neural processor circuit 218 for executing the task.

[0087] The task manager DMA 1006 is coupled to task arbiter 1002, the system memory 230, and the fetch queue 1008. The task manager DMA 1006 includes a read circuit that receives task descriptors 1012 of tasks from a source (e.g., system memory 230) for storing in the fetch queue 1008. For example, the task arbiter 1002 selects a task queue 1004 according to the priorities of the task queues 1004, and uses the task list 904 referenced by the selected task queue 1004 to control the task manager DMA 1006 to select the task descriptor 1012 of a task.

[0088] The fetch queue 1008 is a single entry queue that stores a task descriptor 1012 of a task that is pending to commit for execution. The fetch queue 1008 is coupled to the task manager DMA 1006 to receive the task descriptor 1012 from the system memory 230, and provides the task descriptor 1012 to the configuration queue 1010, or the configuration data 1014 extracted from the task descriptor 1012 to the configuration queue 1010.

[0089] The configuration queue 1010 holds configuration data 1014 of multiple tasks that have been committed for execution. When a task is in the configuration queue 1010, the kernel DMA 324 may fetch kernel data from the system memory 230 to store in the kernel extract circuit 432 of neural engines 314, and the buffer DMA 320 may fetch input data from the system memory 230 to store in the data buffer 318. To execute the task, the kernel extract circuit 432 provides the prefetched kernel data to the MAC 404 of the neural engine 314, and the data buffer 318 provides the prefetched input data to the MAC 404 of the neural engine 314. In some embodiments, the configuration queue 1010 may include multiple queues that holds

configuration data 1014 extracted from the committed task descriptors 1012. As discussed in greater detail in connection with FIG. 13, the configuration queue 1010 is further coupled to other components of the neural processor circuit 218 to configure the neural processor circuit 218 according to the configuration data 1014.

[0090] FIG. 11 is a diagram illustrating retrieval of task descriptors 1012 using a task queue 1004, according to one embodiment. The task queue 1004 includes a reference, such as a set of pointers, to the task descriptors 1012A through 1012N stored in the system memory 230. To that end, the task queue 1004 may include a memory storing a head parameter 1102, a network identifier (ID) 1104, a base address index 1106, a tail parameter 1108, a count parameter 1110, and a priority parameter 1112. The head parameter 1102 is a pointer to a location of the system memory 230 storing the task descriptor 1012A at the head of the task queue 1004. The network ID 1104 identifies the neural network 900 of the task descriptor 1012 at the head of the task queue 1004, and the base address index 1106 is an index to a base-address table 1114 tagged with the network ID 1104 of the task descriptor 1012A at the head of the task queue 1004. The count parameter 1110 defines the number of task descriptors 1012 in the task queue 1004. The priority parameter 1112 defines the priority of the task queue 1004, which is used by the task arbiter 1002 to select between multiple task queues 1004.

[0091] When a particular task queue 1004 is selected (e.g., according to the priority parameter 1112), the task arbiter 1002 references the head parameter 1102, the network ID 1104, the base address index 1106, and the base address table 1114 to retrieve a task descriptor 1012 from the system memory 230, and places the task descriptor 1012 into the fetch queue 1008 to initiate commitment of the task for execution. In each configuration period, the task arbiter 1002 may continue to place a task descriptor 1012 into the fetch queue 1008 according to the order of tasks defined by the task list 904 of the task queue 1004, such as by retrieving the next task descriptor 1012B, and so forth.

[0092] FIG. 12 is a diagram illustrating a task descriptor 1012, according to one

embodiment. The task arbiter 1002 places the task descriptor 1012 in the fetch queue 1008 from system memory 230, which is then transferred to the configuration queue 1010. The highest priority (e.g., first in) task descriptor 1012 in the configuration queue 1010 is used to configure the neural processor circuit 218 for execution during the configuration period. The task descriptor 1012 includes configuration data 1014 including a task descriptor header 1202 and address data 1204 A through 1204N (hereinafter referred as“address data 1204”). The task descriptor header 1202 includes configuration data 1014 that configures various operations of the neural task manager 310, including operations related to task selection and task switching. For example, the task descriptor header 1202 may be parsed by the task arbiter 1002 to extract configuration data 1014 that programs the neural task manager 310 and other components of the neural processing circuit 218. The task descriptor header 1202 may include a task identifier (ID) 1206 that identifies the task, a neural network identifier (ID) 1208 that identifies a neural network 900 instantiated by the task, a task switch parameter 1210 defining whether the neural task manager 310 should initiate a task switch (e.g., to execute a task of a different task queue 1004) after execution of the task, an input surface parameter 1212 defining whether the input data for the task should be retrieved from the system memory 230 or the data buffer 318, an output surface parameter 1214 defining whether the output data of the task should be stored in the system memory 230 or the data buffer 318, various (e.g., base address) pointers 1216 to facilitate the programming of the neural processor circuit 218, and one or more debug/exception parameters 1218 that control event, exception, or debug logging.

[0093] Each instance of address data 1204 A through 1204N (collectively or individually referred to as“address data 1204”) defines an address and data payload pair used to program the components of the neural processor circuit 218. The data payload may include input data and kernel data used to execute the task. For example, each instance of address data 1204 includes register data defining the data payload, a register address defining a destination memory location of the neural processing circuit 218 for receiving the register data, and a register count defining a number of consecutive memory locations (e.g., registers) to be written with the register data. In some embodiments, the register address is combined with the base address stored in the header 1202 to define the full address of each memory location. If the task descriptor 1116 is generated at compile time, then the actual run time addresses may not be known. The base address table 1114 is used avoid duplicating or updating all task descriptors with dynamically assigned addresses.

[0094] FIG. 13 is a block diagram illustrating the fetch queue 1008 and configuration queue 1010, according to one embodiment. The configuration queue 1010 is coupled to the fetch queue 1008, which is coupled to the system memory 230 via the task manager DMA 1006. The configuration queue 1010 is further coupled to the rasterizer 714 of one or more neural engines 314, the rasterizer 718 of the data buffer 318, the rasterizer 720 of the buffer DMA 320, and the rasterizer 722 of the kernel DMA 322. The fetch queue 1008 stores a task descriptor 1012 (e.g., including the task descriptor header 1202 and the address data 1204 A through 1204N) for a task that is pending and not committed to execution. The fetch queue 1008 reduces the latency of reading the next task descriptor 1012 into the configuration queue 1010 from the system memory 230. The fetch queue 1008 stores the highest priority task descriptor 1012 as determined by the task arbiter 1002. The task arbiter 1002 may replace the task descriptor 1012 stored in the fetch queue 1008 if a higher priority task descriptor 1012 has been has been enqueued (e.g., from a higher priority task queue 1004). The task descriptor 1012 in the fetch queue 1008 does not initiate an input data or kernel prefetch, and does not affect task queue priorities, pointers, or counters. As such, a task descriptor 1012 in the fetch queue 1008 may be readily replaced by a higher priority task descriptor 1012 by writing the higher priority task descriptor 1012 into the fetch queue 1008. When a task descriptor 1012 stored in the configuration queue 1010 is executed by the neural processor circuit 218, the task descriptor 1012 stored in the fetch queue 1008 is transferred to the configuration queue 1010, and another task descriptor 1012 of a subsequent task may be stored in the fetch queue 1008.

[0095] The configuration queue 1010 stores task descriptors 1012 of tasks committed for execution by the neural processor circuit 218. In some embodiments, the configuration queue 1010 includes multiple separate queues 1310 that each store a portion of the configuration data 1014 (including configuration data 1014A through 1014E) extracted from the task descriptor 1012. Furthermore, the queues 1310 are each coupled to a respective component of the neural processor circuit 218 for programming the component with the configuration data 1014.

Through operation of the configuration queue 1010, the neural task manager 310 programs the rasterizers 714, 718, 720, 722 to perform the functionality discussed above in FIGS. 7 and 8. For example, a queue 1310A is coupled to the rasterizers 714 of the neural engines 314 to provide configuration data 1014A that controls the operations of the shifters 410 in the input buffer circuits 402 to forward correct portions 408 of input data to MAC 404, and send the finished output data 328 to the data buffer 318. The queue 1310B is coupled to the rasterizer 718 of the data buffer 318 to provide configuration data 1014B that controls the broadcasting of input data (e.g., work units) by the data buffer 318 for processing by the neural engines 314. The queue 1310C is a read queue that is coupled to the rasterizer 720 of the buffer DMA 320 to provide configuration data 1014C that controls the buffer DMA 320 to retrieve input data (e.g., a tile) from system memory 230 and store the input data in the data buffer 318. The queue 1310D is a write queue that is coupled to the rasterizer 720 of the buffer DMA 320 to provide configuration data 1014D that controls the buffer DMA 320 to store output data in the system memory 230. The queue 1310E is coupled to the rasterizer 722 of the kernel DMA 322 to provide

configuration data 1014E that controls which kernels are to be received and distributed to neural engines 314. In some embodiments, a task descriptor 1012 or configuration data 1014 stored in the configuration queue 1010 cannot be replaced and will be executed in a first in, first out (FIFO) order.

Example Process at Neural Task Manager Architecture

[0096] FIG. 14 is a flowchart illustrating a method of managing tasks in the neural processor circuit 218, according to one embodiment. The method may include different and/or additional steps, or the steps may be in different orders.

[0097] The CPEi 208 generates 1402 a task list 904 of task descriptors 1012 of tasks that when executed by the neural processor circuit 218 instantiates a neural network 900. For example, the CPEI 208 may receive input data for a machine learning operation from the image sensor 202, the system memory 230, the persistent storage 228, the network interface 210, or some other components. The machine learning operation may include an inferencing operation, or a training operation. The neural network 900 may include kernel data and a neural network architecture including network layers. The input data is applied to the neural network 900 to perform the machine learning operation. The kernel data and network layers of the neural network 900 may be computed in a machine learning training process. The CPEI 208 performs a compile operation (offline or on-the-fiy) to turn a neural network description 900 into a linked list of task descriptors 1012 referred to herein as the task list 904. Each task is defined by a task descriptor 1012, and when executed by the neural processor circuit 218 instantiates a single network layer, multiple network layers, or a portion of a network layer. Each task descriptor 1012 of a task includes configuration data 1014, such as the task descriptor header 1202 and the address data 1204 defining neural processor circuit 218 address and data payload pairs. The data payload may include the kernel data of the neural network 900 and the input data. The configuration data 1014 further includes instructions that configure operation of the rasterizers 714, 718, 720, and 722 to execute the task. [0098] The CPU 208 stores 1404 the task list 904 of task descriptors 1012 in the system memory 230. In some embodiments, the CPU 208 or another CPU external to the electronic device 100 generates the task descriptors 1012 and stores the task descriptors 1012 in the persistent storage 228 or some other non-volatile memory. The task descriptors 1012 are loaded along with the kernel data and input data in the system memory 230 for use by the neural processor circuit 218. The CPU 208 may be coupled to the system memory 230 via the bus 232 and the memory controller 222.

[0099] The task list 904 may include a set of pointers that reference the locations of the task descriptors 1012 in the system memory 230. Furthermore, the CPU 208 may update parameters of the task descriptors 1102 such as memory addresses or the network ID. For example, the task list 904 may include a head parameter 1102 and a tail parameter 1108 respectively defining a beginning register address and an end register addresses of the system memory 230 where multiple tasks descriptors 1012 are stored. In some embodiments, the references to the register addresses in the task list 904 may be partial addresses, and a base address table 1114 is used to define the full reference addresses to the system memory 230. In some embodiments, the CPU

208 av tch absolute addresses as necessary The CPU 208 mav further configure the neural processor circuit 218 by setting its base address registers.

[00100] The task arbiter 1002 of the neural task manager 310 enqueues 1406 the task list 904 for execution by the neural processor circuit 218. For example, the neural task manager 310 includes multiple task queues 1004 A through 1004N. The task queues 1004 may each store a reference to a task list 904. Furthermore, the task queues 1004 are prioritized for execution according to the priority parameters 11 12 of the task lists 904 referenced by the task queues

1004.

[00101] In some embodiments, the CPU 208 performs a general configuration of the neural processor circuit 218 to execute the task. The CPU 208 may further start the neural processor circuit 218 if the neural processor circuit 218 is not already running.

[00102] The task arbiter 1002 selects 1408 a task queue with highest priority for execution. For example, the task arbiter 1002 in each programming period selects a task queue 1004 based a comparison of priority parameters of the task queues 1004 or the task lists 904 of the task queues 1004, and executes tasks from the task list 904 from the highest priority task queue 1004. [00103] The neural processor circuit 218 executes 1410 a task from the selected task queue. For example, the neural processor circuit 218 performs the method of FIG. 8 to execute the task. The configuration queue 1010 of the neural task manager 310 may provide the configuration data 1014 (or task information 710) to the rasterizers 714, 718, 720, 722 of the neural processor circuit 218 to program the neural processor circuit 218 to execute the task, as shown in FIG. 7. Furthermore, the execution of the task may include processing of the prefetched kernel data and input data while the task was in the configuration queue 1010. The execution of each task may include multiple processing loops for handling the input data as shown in FIGS. 5 and 6. The task arbiter 1002 may dequeue the task from the configuration queue 1010 after execution of the task.

[00104] In some embodiments, to execute the task, the task arbiter 1002 controls the task manager DMA 1006 to retrieve the task descriptor 1012 of the task of the task list 904, and store the task descriptor 1012 in the fetch queue 1008. After execution of a previously committed task, the task arbiter 1002 may dequeue the executed task by removing the task descriptor or configuration data of the task from the configuration queue 1010. The task descriptor 1012 or extracted configuration data 1014 of the current task is then placed in the configuration queue 1010 from the fetch queue 1008. When the task is in the configuration queue 1010, the neural processor circuit 218 may initiate a prefetch operation by the kernel DMA 324 for kernel data from the system memory 230 to the kernel extract circuit 432, and a prefetch operation by the buffer DMA 320 for input data from the system memory 230 to the data buffer 318.

[00105] If a different task queue has a higher priority, a task switch to the task queue with the higher priority may be performed when possible. To perform a task switch (e.g., to another machine learning operation), the task arbiter 1002 replaces a task in the fetch queue 1008 with another task descriptor referenced in a different task queue 1004. The task arbiter 1002 may resume the interrupted machine learning operation defined by the task list 904 after completion of the task switch by storing the replaced task into the fetch queue 1008.

[00106] In some embodiments, the CPU 208 determines 1408 a priority parameter 1110 of the task list 904 to select a task queue. The priority parameter 1110 defines the priority of the tasks of the machine learning operation relative to other tasks of other machine learning operations executed by the neural processor circuit 218. The CPU 208 may facilitate the execution of multiple machine learning operations. The machine learning operations may be different, such as by using different input data or different neural network architectures and kernel data. The CPU 208 may determine the priority parameter of the task list 904 programmatically, or based on user input. The task descriptor 1012 referenced by the task queue 1004 with the highest priority may be selected for execution. The priority of a task queue is either determined by the CPU 208, or dynamically by information from a previously executed task descriptor.

[00107] The task arbiter 1002 determines 1412 whether all tasks of the task list 904 have been executed. If a task of the task list 904 has not been executed, the process returns to step 1410, where the task arbiter 1002 executes the unexecuted task of the task list 904.

[00108] If each task of the task list 904 has been executed, the task arbiter 1002 removes 1414 the task list 904 from the task queue 1004. In another example, a reference to a task descriptor 1012 of a task may be removed from the task queue 1004 subsequent to execution of the task.

The process may end. In another example, the process may return to 1402, where the CPU 208 may continue to generate task descriptors of tasks for other machine learning operations. In that sense, the process may be repeated to place tasks into task queues for execution of machine learning operations by the neural processor circuit 218 according to specified task priorities.

[00109] While particular embodiments and applications have been illustrated and described, it is to be understood that the invention is not limited to the precise construction and components disclosed herein and that various modifications, changes and variations which will be apparent to those skilled in the art may be made in the arrangement, operation and details of the method and apparatus disclosed herein without departing from the spirit and scope of the present disclosure.