Apple Inc.

Benjamin Bowman

Andrew M Havlir

Fergus W MacGarry

Kutty Banerjee

Steven John Fishwick

Ali Rabbani Rankouhi

David A Gotwalt

Pratik Chandresh Shah

Terence M Potter

Gokhan Avkarogullari

Ralph Clay Taylor

Jeffrey T Brady

Tatsuya Iwamoto

Jonathan M Redshaw

Jason David Carroll

Mladen Wilder

Arjun Thottappilly

Ajay Simha Modugala

Melissa L Velez

Yunjun Zhang

Techniques are disclosed relating to geometry processing in graphics processors. In some embodiments, geometry processing circuitry is configured to execute multiple segments of a segmented set of geometry work at least partially in parallel using multiple geometry processing pipelines, execute a single-segment set of geometry work using a single geometry processing pipeline, and identify a group of multiple sets of geometry work that includes the segmented set and the single-segment set. Fragment processing circuitry may be configured to initiate execution of a single set of fragment processing work that consumes output data from the multiple sets of geometry work in the group. The geometry processing circuitry may utilize disclosed techniques to process the segmented and single-segment sets in a manner than mitigates overhead associated with segmented processing.

Graphics geometry processing with segmented and non-segmented sets of work

Techniques are disclosed relating to affinity-based scheduling of graphics work. In disclosed embodiments, first and second groups of graphics processor sub-units may share respective first and second caches. Distribution circuitry may receive a software-specified set of graphics work and a software-indicated mapping of portions of the set of graphics work to groups of graphics processor sub-units. The distribution circuitry may assign subsets of the set of graphics work based on the mapping. This may improve cache efficiency, in some embodiments, by allowing graphics work that accesses the same memory areas to be assigned to the same group of sub-units that share a cache.

Affinity-based graphics scheduling

Disclosed techniques relate to circuitry configured to aggregate and report usage information in a distributed processor (e.g., a GPU). In some embodiments, graphics processor circuitry that includes at least first and second portions that are respectively configured to execute sets of graphics work. First utilization circuitry may track execution time for sets of graphics work on the first portion of the graphics processor circuitry and second utilization circuitry may track execution time for sets of graphics work on the second portion of the graphics processor circuitry. Command queue circuitry may store multiple different command queues. Control circuitry may access the first and second utilization circuitry and aggregate utilization data on a per-command-queue basis, where for a given command queue, the aggregated utilization data indicates respective utilization of the first and second portions of the graphics processor circuitry. The control circuitry may provide the aggregated per-command-queue utilization data in software-accessible registers.

Quality of service techniques in distributed graphics processor

Disclosed embodiments relate to controlling sets of graphics work (e.g., kicks) assigned to graphics processor circuitry. In some embodiments, tracking slot circuitry implements entries for multiple tracking slots. Slot manager circuitry may store, using an entry of the tracking slot circuitry, software-specified information for a set of graphics work, where the information includes: type of work, dependencies on other sets of graphics work, and location of data for the set of graphics work. The slot manager circuitry may prefetch, from the location and prior to allocating shader core resources for the set of graphics work, configuration register data for the set of graphics work. Control circuitry may program configuration registers for the set of graphics work using the prefetched data and initiate processing of the set of graphics work by the graphics processor circuitry according to the dependencies. Disclosed techniques may reduce kick-to-kick transition time, in some embodiments.

Kickslot manager circuitry for graphics processors

Disclosed techniques relate to work distribution in graphics processors. In some embodiments, an apparatus includes circuitry that implements a plurality of logical slots and a set of graphics processor sub-units that each implement multiple distributed hardware slots. The circuitry may determine different distribution rules for first and second sets of graphics work and map logical slots to distributed hardware slots based on the distribution rules. In various embodiments, disclosed techniques may advantageously distribute work efficiently across distributed shader processors for graphics kicks of various sizes.

Logical slot to hardware slot mapping for graphics processors

Disclosed techniques relate to distributing graphics work based on priority. In some embodiments, circuitry implements a plurality of tracking slots for sets of graphics work. A set of graphics processor sub-units may each implement multiple distributed hardware slots. Control circuitry may attempt to assign a first set of graphics work having a first priority to a graphics processor sub-unit that is currently executing graphics work having an equal or higher priority than the first priority, where the first set of graphics work is from a first tracking slot. The control circuitry may, in response to a failure of the attempt, generate a signal to graphics software that indicates the failure, wherein the signal indicates the first tracking slot. Disclosed techniques may reduce or avoid problems relating to higher priority work being scheduled behind lower priority work.

Priority inversion mitigation techniques

Techniques are disclosed relating to dynamically adjusting buffering for distributing compute work in a graphics processor. In some embodiments, the graphics processor includes shader circuitry configured to process compute work from a compute kernel, multiple distributed workload parser circuits configured to send compute work to the shader circuitry, primary workload parser circuitry configured to send, via a communications fabric, compute work from the compute kernel to the distributed workload parser circuits, and buffer circuitry configured to buffer compute work received by one or more of the distributed workload parser circuits from the primary workload parser circuitry. In some embodiments, the graphics processor is configured to dynamically adjust a limit on the number of entries used in the buffer circuitry based on information indicating complexity of the compute kernel. This may advantageously maintain launch rates while reducing or avoiding workload imbalances, in some embodiments.

Dynamic buffering control for compute work distribution

In general, embodiments are disclosed for tracking and allocating graphics processor hardware resources. More particularly, a graphics hardware resource allocation system is able to generate a priority list for a plurality of data masters for graphics processor based on a comparison between a current utilizations for the data masters and a target utilizations for the data masters. The graphics hardware resource allocation system designate, based on the priority list, a first data master with a higher priority to submit work to the graphics processor compared to a second data master. The graphics hardware resource allocation system determines a stall counter value for the data master and generates a notification to pause work for the second data master based on the stall counter value.

Graphics hardware driven pause for quality of service adjustment

Techniques are disclosed relating to distributing work from compute kernels using a distributed hierarchical parser architecture. In some embodiments, an apparatus includes a plurality of shader units configured to perform operations for compute workgroups included in compute kernels processed by the apparatus, a plurality of distributed workload parser circuits, and a communications fabric connected to the plurality of distributed workload parser circuits and a master workload parser circuit. In some embodiments, the master workload parser circuit is configured to iteratively determine a next position in multiple dimensions for a next batch of workgroups from the kernel and send batch information to the distributed workload parser circuits via the communications fabric to assign the batch of workgroups. In some embodiments, the distributed parsers maintain coordinate information for the kernel and update the coordinate information in response to the batch information, even when the distributed parsers are not assigned to execute the batch.

Growing community of inventors

London, United Kingdom

Benjamin Bowman