Abstract: A processing apparatus is provided comprising a multiprocessor having a multithreaded architecture. The multiprocessor can execute at least one single instruction to perform parallel mixed precision matrix operations. In one embodiment the apparatus includes a memory interface and an array of multiprocessors coupled to the memory interface. At least one multiprocessor in the array of multiprocessors is configured to execute a fused multiply-add instruction in parallel across multiple threads.

Abstract: One embodiment provides for a graphics processing unit to accelerate machine-learning operations, the graphics processing unit comprising a multiprocessor having a single instruction, multiple thread (SIMT) architecture, the multiprocessor to execute at least one single instruction; and a first compute unit included within the multiprocessor, the at least one single instruction to cause the first compute unit to perform a two-dimensional matrix multiply and accumulate operation, wherein to perform the two-dimensional matrix multiply and accumulate operation includes to compute a 32-bit intermediate product of 16-bit operands and to compute a 32-bit sum based on the 32-bit intermediate product.

Abstract: A mechanism is described for facilitating person tracking and data security in machine learning at autonomous machines. A method of embodiments, as described herein, includes detecting, by a camera associated with one or more trackers, a person within a physical vicinity, where detecting includes capturing one or more images the person. The method may further include tracking, by the one or more trackers, the person based on the one or more images of the person, where tracking includes collect tracking data relating to the person. The method may further include selecting a tracker of the one or more trackers as a preferred tracker based on the tracking data.

Abstract: A disaggregated processor package can be configured to accept interchangeable chiplets. Interchangeability is enabled by specifying a standard physical interconnect for chiplets that can enable the chiplet to interface with a fabric or bridge interconnect. Chiplets from different IP designers can conform to the common interconnect, enabling such chiplets to be interchangeable during assembly. The fabric and bridge interconnects logic on the chiplet can then be configured to confirm with the actual interconnect layout of the on-board logic of the chiplet. Additionally, data from chiplets can be transmitted across an inter-chiplet fabric using encapsulation, such that the actual data being transferred is opaque to the fabric, further enable interchangeability of the individual chiplets. With such an interchangeable design, higher or lower density memory can be inserted into memory chiplet slots, while compute or graphics chiplets with a higher or lower core count can be inserted into logic chiplet slots.

Abstract: A scheme is provided for dynamically adjusting an amount of power drawn from individual power sources to optimize the power usage without violating power limits. Coarse adjustment is provided through dynamic phase reallocation while a fine adjustment is provided through dynamic current steering. By adding a control loop around current steering techniques in digital voltage regulator controllers, power drawn from multiple input rails is balanced. The apparatus allows users to maximize the power delivered to discrete graphics cards without violating PCIe specifications. This allows maximum performance with minimal bill-of-material (BOM) cost.

Abstract: A mechanism is described for facilitating intelligent dispatching and vectorizing at autonomous machines. A method of embodiments, as described herein, includes detecting a plurality of threads corresponding to a plurality of workloads associated with tasks relating to a graphics processor. The method may further include determining a first set of threads of the plurality of threads that are similar to each other or have adjacent surfaces, and physically clustering the first set of threads close together using a first set of adjacent compute blocks.

Abstract: A processing apparatus is provided comprising a multiprocessor having a multithreaded architecture. The multiprocessor can execute at least one single instruction to perform parallel mixed precision matrix operations. In one embodiment the apparatus includes a memory interface and an array of multiprocessors coupled to the memory interface. At least one multiprocessor in the array of multiprocessors is configured to execute a fused multiply-add instruction in parallel across multiple threads.

Abstract: A mechanism is described for facilitating inference coordination and processing utilization for machine learning at autonomous machines. A method of embodiments, as described herein, includes detecting, at training time, information relating to one or more tasks to be performed according to a training dataset relating to a processor including a graphics processor. The method may further include analyzing the information to determine one or more portions of hardware relating to the processor capable of supporting the one or more tasks, and configuring the hardware to pre-select the one or more portions to perform the one or more tasks, while other portions of the hardware remain available for other tasks.

Abstract: A processing apparatus is provided comprising a multiprocessor having a multithreaded architecture. The multiprocessor can execute at least one single instruction to perform parallel mixed precision matrix operations. In one embodiment the apparatus includes a memory interface and an array of multiprocessors coupled to the memory interface. At least one multiprocessor in the array of multiprocessors is configured to execute a fused multiply-add instruction in parallel across multiple threads.

Abstract: Embodiments described herein are generally directed to improvements relating to power, latency, bandwidth and/or performance issues relating to GPU processing/caching. According to one embodiment, a system includes a producer intellectual property (IP) (e.g., a media IP), a compute core (e.g., a GPU or an AI-specific core of the GPU), a streaming buffer logically interposed between the producer IP and the compute core. The producer IP is operable to consume data from memory and output results to the streaming buffer. The compute core is operable to perform AI inference processing based on data consumed from the streaming buffer and output AI inference processing results to the memory.

Abstract: Apparatuses including a graphics processing unit, graphics multiprocessor, or graphics processor having an error detection correction logic for cache memory or shared memory are disclosed. In one embodiment, a graphics multiprocessor includes cache or local memory for storing data and error detection correction circuitry integrated with or coupled to the cache or local memory. The error detection correction circuitry is configured to perform a tag read for data of the cache or local memory to check error detection correction information.

Abstract: Embodiments described herein provide techniques to disaggregate an architecture of a system on a chip integrated circuit into multiple distinct chiplets that can be packaged onto a common chassis. In one embodiment, a graphics processing unit or parallel processor is composed from diverse silicon chiplets that are separately manufactured. A chiplet is an at least partially packaged integrated circuit that includes distinct units of logic that can be assembled with other chiplets into a larger package. A diverse set of chiplets with different IP core logic can be assembled into a single device.

Abstract: One embodiment provides for a graphics processing unit to accelerate machine-learning operations, the graphics processing unit comprising a multiprocessor having a single instruction, multiple thread (SIMT) architecture, the multiprocessor to execute at least one single instruction; and a first compute unit included within the multiprocessor, the at least one single instruction to cause the first compute unit to perform a two-dimensional matrix multiply and accumulate operation, wherein to perform the two-dimensional matrix multiply and accumulate operation includes to compute a 32-bit intermediate product of 16-bit operands and to compute a 32-bit sum based on the 32-bit intermediate product.

Abstract: Methods and apparatus relating to techniques for resource load balancing based on usage and/or power limits are described. In an embodiment, resource load balancing logic causes a first resource of a processor to operate at a first frequency and a second resource of the processor to operate at a second frequency. Memory stores a plurality of frequency values. The resource load balancing logic also selects the first frequency and the second frequency based on the stored plurality of frequency values. Operation of the first resource at the first frequency and the second resource at the second frequency in turn causes the processor to operate under a power budget. The resource load balancing logic causes change to the first frequency and the second frequency in response to a determination that operation of the processor is different than the power budget. Other embodiments are also disclosed and claimed.

Abstract: Dynamic power budget allocation in a multi-processor system is described. In an example, an apparatus includes a plurality of processor units; and a power control component, the power control component to monitor power utilization of each of the plurality of processor units, wherein power consumed by the plurality of processor units is limited by a global power budget. The apparatus is to assign a workload to each of the processor units and is to establish an initial power budget for operation of each of the processor units, and, upon the apparatus determining that one or more processor units require an increased power budget based on one or more criteria, the apparatus is to dynamically reallocate an amount of the global power budget to the one or more processor units.

Abstract: A mechanism is described for facilitating person tracking and data security in machine learning at autonomous machines. A method of embodiments, as described herein, includes detecting, by a camera associated with one or more trackers, a person within a physical vicinity, where detecting includes capturing one or more images the person. The method may further include tracking, by the one or more trackers, the person based on the one or more images of the person, where tracking includes collect tracking data relating to the person. The method may further include selecting a tracker of the one or more trackers as a preferred tracker based on the tracking data.

Abstract: Embodiments are generally directed to providing power savings for a neural network architecture with zero activations during inference. An embodiment of an apparatus includes one or more processors including one or more processor cores; and a memory to store data for processing including neural network processing, wherein the apparatus to perform a fast clear operation to initialize activation buffers for a neural network by updating metadata to indicate zero values, the neural network including a plurality of layers, wherein the apparatus is to compare outputs for the neural network to the metadata values and to write an output to memory only if the output is non-zero.

Abstract: A disaggregated processor package can be configured to accept interchangeable chiplets. Interchangeability is enabled by specifying a standard physical interconnect for chiplets that can enable the chiplet to interface with a fabric or bridge interconnect. Chiplets from different IP designers can conform to the common interconnect, enabling such chiplets to be interchangeable during assembly. The fabric and bridge interconnects logic on the chiplet can then be configured to confirm with the actual interconnect layout of the on-board logic of the chiplet. Additionally, data from chiplets can be transmitted across an inter-chiplet fabric using encapsulation, such that the actual data being transferred is opaque to the fabric, further enable interchangeability of the individual chiplets. With such an interchangeable design, higher or lower density memory can be inserted into memory chiplet slots, while compute or graphics chiplets with a higher or lower core count can be inserted into logic chiplet slots.

Abstract: A mechanism is described for facilitating inference coordination and processing utilization for machine learning at autonomous machines. A method of embodiments, as described herein, includes detecting, at training time, information relating to one or more tasks to be performed according to a training dataset relating to a processor including a graphics processor. The method may further include analyzing the information to determine one or more portions of hardware relating to the processor capable of supporting the one or more tasks, and configuring the hardware to pre-select the one or more portions to perform the one or more tasks, while other portions of the hardware remain available for other tasks.

Abstract: A mechanism is described for facilitating intelligent dispatching and vectorizing at autonomous machines. A method of embodiments, as described herein, includes detecting a plurality of threads corresponding to a plurality of workloads associated with tasks relating to a graphics processor. The method may further include determining a first set of threads of the plurality of threads that are similar to each other or have adjacent surfaces, and physically clustering the first set of threads close together using a first set of adjacent compute blocks.