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.

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.

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.

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.

A parallel processor system for machine learning includes an arithmetic unit (ALU) array including several ALUs and a controller to provide instructions for the ALUs. The system includes a direct-access memory (DMA) block containing multiple DMA engines to access an external memory to retrieve data. An input-stream buffer decouples the DMA block from the ALU array and provides aligning and reordering of the retrieved data. The DMA engines operate in parallel and include rasterization logic capable of performing a three-dimensional (3-D) rasterization.

Technologies for allocating resources of managed nodes to workloads to balance multiple resource allocation objectives include an orchestrator server to receive resource allocation objective data indicative of multiple resource allocation objectives to be satisfied. The orchestrator server is additionally to determine an initial assignment of a set of workloads among the managed nodes and receive telemetry data from the managed nodes. The orchestrator server is further to determine, as a function of the telemetry data and the resource allocation objective data, an adjustment to the assignment of the workloads to increase an achievement of at least one of the resource allocation objectives without decreasing an achievement of another of the resource allocation objectives, and apply the adjustments to the assignments of the workloads among the managed nodes as the workloads are performed. Other embodiments are also described and claimed.

A core of neural processing units is configured to efficiently process a depthwise convolution by maximizing spatial feature-map locality using adder trees. Data paths of activations and weights are inverted, and 2-to-1 multiplexers are every 2/9 multipliers along a row of multipliers. During a depthwise convolution operation, the core is operated using a RS×HW dataflow to maximize the locality of feature maps. For a normal convolution operation, the data paths of activations and weights may be configured for a normal convolution configuration and in which multiplexers are idle.

Technologies for switching network traffic include a network switch. The network switch includes one or more processors and communication circuitry coupled to the one or more processors. The communication circuitry is capable of switching network traffic of multiple link layer protocols. Additionally, the network switch includes one or more memory devices storing instructions that, when executed, cause the network switch to receive, with the communication circuitry through an optical connection, network traffic to be forwarded, and determine a link layer protocol of the received network traffic. The instructions additionally cause the network switch to forward the network traffic as a function of the determined link layer protocol. Other embodiments are also described and claimed.

A parallel processor system for machine learning includes an arithmetic unit (ALU) array including several ALUs and a controller to provide instructions for the ALUs. The system further includes a direct-access memory (DMA) block containing multiple DMA engines to access an external memory to retrieve data. An input-stream buffer decouples the DMA block from the ALU array and provides aligning and reordering of the retrieved data. The DMA engines operate in parallel and include rasterization logic capable of performing a three-dimensional (3-D) rasterization.

Technologies for switching network traffic include a network switch. The network switch includes one or more processors and communication circuitry coupled to the one or more processors. The communication circuity is capable of switching network traffic of multiple link layer protocols. Additionally, the network switch includes one or more memory devices storing instructions that, when executed, cause the network switch to receive, with the communication circuitry through an optical connection, network traffic to be forwarded, and determine a link layer protocol of the received network traffic. The instructions additionally cause the network switch to forward the network traffic as a function of the determined link layer protocol. Other embodiments are also described and claimed.