Devices and techniques for non-blocking external device calls are described herein. Specifically, when a processor receives an instruction with a no-return indication from a thread for a device, the processor can increase a counter corresponding to the thread based on the no-return indication. The processor can then continue execution of the thread without waiting for a return value from the device. When a return value is received for the instruction, the processor can decrement the counter. While the counter is not zero, the processor prevents the thread from completing (exiting).

In various examples, a VPU and associated components may be optimized to improve VPU performance and throughput. For example, the VPU may include a min/max collector, automatic store predication functionality, a SIMD data path organization that allows for inter-lane sharing, a transposed load/store with stride parameter functionality, a load with permute and zero insertion functionality, hardware, logic, and memory layout functionality to allow for two point and two by two point lookups, and per memory bank load caching capabilities. In addition, decoupled accelerators may be used to offload VPU processing tasks to increase throughput and performance, and a hardware sequencer may be included in a DMA system to reduce programming complexity of the VPU and the DMA system. The DMA and VPU may execute a VPU configuration mode that allows the VPU and DMA to operate without a processing controller for performing dynamic region based data movement operations.

Various examples are directed to systems and methods in which a first flow controller of a first synchronous flow may receive an instruction to execute a first loop using the first synchronous flow. The first flow controller may determine a first iteration index for a first iteration of the first loop. The first flow controller may send, to a first compute element of the first synchronous flow, a first synchronous message to initiate a first synchronous flow thread for executing the first iteration of the first loop. The first synchronous message may comprise the iteration index. The first compute element may execute an input/output operation at a first location of a first compute element memory indicated by the first iteration index.

Systems, apparatuses, and methods for dynamic graphics processing unit (GPU) register allocation are disclosed. A GPU includes at least a plurality of compute units (CUs), a control unit, and a plurality of registers for each CU. If a new wavefront requests more registers than are currently available on the CU, the control unit spills registers associated with stack frames at the bottom of a stack since they will not likely be used in the near future. The control unit has complete flexibility determining how many registers to spill based on dynamic demands and can prefetch the upcoming necessary fills without software involvement. Effectively, the control unit manages the physical register file as a cache. This allows younger workgroups to be dynamically descheduled so that older workgroups can allocate additional registers when needed to ensure improved fairness and better forward progress guarantees.

Methods and electronic circuits for executing instructions in a central processing unit (CPU) are provided. One of the methods includes forming an instruction block by sequentially fetching, from a current thread queue, one or more instructions including one jump instruction, wherein the jump instruction is the last instruction in the instruction block; transmitting the instruction block to a CPU execution unit for execution; replenishing the current thread queue with at least one instruction to form a thread queue to be executed; determining a target instruction of the jump instruction according to an execution result of the CPU execution unit; determining whether the target instruction is contained in the thread queue to be executed; and if not, flushing the thread queue to be executed, obtaining the target instruction and adding the target instruction to the thread queue to be executed.

An apparatus to facilitate neural network (NN) training is disclosed. The apparatus includes training logic to receive one or more network constraints and train the NN by automatically determining a best network layout and parameters based on the network constraints.

An apparatus is disclosed. The apparatus includes one or more processors comprising register sharing circuitry to receive meta-information indicating a number of threads that are to be disabled and provide an indication that an associated thread is disabled, a plurality of General Purpose Register Files (GRFs), wherein one or more of the plurality of GRFs is associated with one of the plurality of threads and a plurality of multiplexers coupled to the one or more GRFs to receive the indication from the register sharing circuitry and disable thread access to an associated GRF based on an indication that a thread is to be disabled.

Systems and methods are disclosed for scheduling threads on a processor that has at least two different core types, such as an asymmetric multiprocessing system. Each core type can run at a plurality of selectable voltage and frequency scaling (DVFS) states. Threads from a plurality of processes can be grouped into thread groups. Execution metrics are accumulated for threads of a thread group and fed into a plurality of tunable controllers for the thread group. A closed loop performance control (CLPC) system determines a control effort for the thread group and maps the control effort to a recommended core type and DVFS state. A closed loop thermal and power management system can limit the control effort determined by the CLPC for a thread group, and limit the power, core type, and DVFS states for the system. Deferred interrupts can be used to increase performance.

Representative apparatus, method, and system embodiments are disclosed for configurable computing. In a representative embodiment, a system includes an interconnection network, a processor, a host interface, and a configurable circuit cluster. The configurable circuit cluster may include a plurality of configurable circuits arranged in an array; an asynchronous packet network and a synchronous network coupled to each configurable circuit of the array; and a memory interface circuit and a dispatch interface circuit coupled to the asynchronous packet network and to the interconnection network. Each configurable circuit includes instruction or configuration memories for selection of a current data path configuration, a master synchronous network input, and a data path configuration for a next configurable circuit.

Representative apparatus, method, and system embodiments are disclosed for a self-scheduling processor which also provides additional functionality. Representative embodiments include a self-scheduling processor, comprising: a processor core adapted to execute instructions; and a core control circuit adapted to automatically schedule an instruction for execution by the processor core in response to a received work descriptor data packet. In a representative embodiment, the processor core is further adapted to execute a non-cached load instruction to designate a general purpose register rather than a data cache for storage of data received from a memory circuit. The core control circuit is also adapted to schedule a fiber create instruction for execution by the processor core, and to generate one or more work descriptor data packets to another circuit for execution of corresponding execution threads. Event processing, data path management, system calls, memory requests, and other new instructions are also disclosed.