Apple Inc.

Advanced Micro Devices Corporation

Ati Technologies Inc.

Ati International Srl

Ralph Clay Taylor

Michael Mantor

Rex Eldon McCrary

Philip J Rogers

Mark M Leather

Robert Scott Hartog

Sebastien Nussbaum

Thomas R Woller

Kevin John McGrath

Michael Andrew Mang

Nuwan S Jayasena

Justin A Hensley

Karl D Mann

Terence M Potter

Jeffrey T Brady

Jason David Carroll

Yoong Chert Foo

Jonathan M Redshaw

Ali Rabbani Rankouhi

Frank W Liljeros

Christopher L Spencer

Randall R Rauwendaal

Jingfei Kong

Steven John Fishwick

Aaftab A Munshi

Andrew M Havlir

Richard Warren Schreyer

David A Gotwalt

Mladen Wilder

Benjamin Bowman

Brian A Buchner

Dimitri Tan

Fergus W MacGarry

Rex Eldon Mccrary

Melissa L Velez

Dinesh D Kuwar

Mark D Earl

Thomas B Pringle

Michael J Manter

Fergus W Macgarry

Steven Fishwick

Techniques are disclosed relating to distributing geometry work in a graphics processor. In some embodiments, multiple geometry pipelines are configured to process, in parallel, multiple segments of a set of geometry work. In some embodiments, control circuitry is configured to split operations from the set of graphics work into the multiple segments. This may include parsing a control stream for the set of graphics work to determine an initial work estimate based on numbers of primitives in different draw calls and adjusting the initial work estimate based on a complexity determination for the draw call.

Parse techniques for graphics workload distribution

Techniques are disclosed relating to dynamically allocating and mapping private memory for requesting circuitry. Disclosed circuitry may receive a private address and translate the private address to a virtual address (which an MMU may then translate to physical address to actually access a storage element). In some embodiments, private memory allocation circuitry is configured to generate page table information and map private memory pages for requests if the page table information is not already setup. In various embodiments, this may advantageously allow dynamic private memory allocation, e.g., to efficiently allocate memory for graphics shaders with different types of workloads. Disclosed caching techniques for page table information may improve performance relative to traditional techniques. Further, disclosed embodiments may facilitate memory consolidation across a device such as a graphics processor.

On-demand memory allocation

Techniques are disclosed relating to using an initial version of an object shader to determine a child count and distribute geometry work based on the child count. In some embodiments, graphics shader circuitry is configured to execute shader programs including object shaders and mesh shaders. Vertex control circuitry is configured to, for a given object shader: launch an initial version of the given object shader to determine a number of meshlets to be generated by the given object shader (e.g., where the initial version of the given object shader does not commit side effects to architectural state of the apparatus) and select shader circuitry to execute a complete version of the given object shader based on the determined number of meshlets.

Initial object shader run for graphics workload distribution

Techniques are disclosed relating to object and mesh shaders executed by a graphics processor. In some embodiments, a device includes buffer circuitry and shader circuitry configured to execute graphics programs. Control circuitry may: generate object shader work and mesh shader work for the shader circuitry, receive output information generated by a mesh shader that indicates a number of vertices and primitives to be output by the mesh shader, allocate, based on the output information and after execution of at least a portion of the mesh shader, a region of the buffer circuitry for storage of the vertices to be output by the mesh shader, and store the vertices output by the mesh shader in the allocated region. Disclosed techniques may advantageously provide efficient use of limited buffer resources.

Resource allocation for mesh shader outputs

Techniques are disclosed relating to object and mesh shaders executed by a graphics processor. In some embodiments, a device includes buffer circuitry, shader circuitry configured to execute graphics programs, including mesh shaders that store output data in the buffer circuitry, and primitive processing circuitry configured to read data from buffer circuitry and process the data, including to cull primitives that are not visible in a graphics frame. Vertex control circuitry may receive: first signaling from the primitive processing circuitry that indicates whether the primitive processing circuitry is waiting for data from the buffer circuitry and second signaling from the shader circuitry that indicates whether the shader circuitry is blocked waiting for allocation in the buffer circuitry. The vertex control circuitry may adjust distribution of mesh shader work to the shader circuitry based on the first signaling and the second signaling.

Mesh shader work distribution

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

Techniques are disclosed relating to rendering graphics objects. In some embodiments, a graphics unit is configured to transform graphics objects from a virtual space into a second space according to different transformation parameters for different portions of the second space. This may result in sampling different portions of the virtual space at different sample rates, which may reduce the number of samples required in various stages of the rendering process. In the disclosed techniques, transformation may occur prior to rasterization and shading, which may further reduce computation and power consumption in a graphics unit, improve image quality as displayed to a user, and/or reduce bandwidth usage or latency of video content on a network. In some embodiments, a transformed image may be viewed through a distortion-compensating lens or resampled prior to display.

Multi-space rendering with configurable transformation parameters

In various embodiments, a resource allocation management circuit may allocate a plurality of different types of hardware resources (e.g., different types of registers) to a plurality of threads. The different types of hardware resources may correspond to a plurality of hardware resource allocation circuits. The resource allocation management circuit may track allocation of the hardware resources to the threads using state identification values of the threads. In response to determining that fewer than a respective requested number of one or more types of the hardware resources are available, the resource allocation management circuit may identify one or more threads for deallocation. As a result, the hardware resource allocation system may allocate hardware resources to threads more efficiently (e.g., may deallocate hardware resources allocated to fewer threads), as compared to a hardware resource allocation system that does not track allocation of hardware resources to threads using state identification values.

Hardware resource allocation system for allocating resources to threads

A method for use in a processor for arbitrating between multiple processes to select wavefronts for execution on a shader core is provided. The processor includes a compute pipeline configured to issue wavefronts to the shader core for execution, a hardware queue descriptor associated with the compute pipeline, and the shader core. The shader core is configured to execute work for the compute pipeline corresponding to a first memory queue descriptor executed using data for the first memory queue descriptor that is loaded into a first hardware queue descriptor. The processor is configured to detect a context switch condition, and, responsive to the context switch condition, perform a context switch operation including loading data for a second memory queue descriptor into the first hardware queue descriptor. The shader core is configured to execute work corresponding to the second memory queue descriptor that is loaded into the first hardware queue descriptor.

Policies for shader resource allocation in a shader core

Techniques are disclosed relating to performing mid-render auxiliary compute tasks for graphics processing. In some embodiments, auxiliary compute tasks are performed during a render pass, using at least a portion of a memory context of the render pass, without accessing a shared memory during the render pass. Relative to flushing render data to shared memory to perform compute tasks, this may reduce memory accesses and/or cache thrashing, which may in turn increase performance and/or reduce power consumption.

Mid-render compute for graphics processing

Techniques are disclosed relating to rendering graphics objects that require shader operations to determine visibility. In some embodiments, a graphics unit is configured to process feedback objects, which may require shading to determine whether they are visible relative to previously-processed objects, out of draw order. For example, in embodiments where a buffer is used to store fragment data for deferred rendering, the graphics unit may bypass the buffer and shade feedback objects ahead of earlier non-feedback objects whose fragment data is stored in the buffer. This may allow a determination of whether to remove occluded non-feedback fragment data from the buffer, which may reduce graphics overdraw. In disclosed two-pass techniques, data for feedback objects is first allowed to bypass the buffer for visibility shading, but is then stored in the buffer for a second pass to perform fragment shading to actually determine pixel attributes, which may further reduce overdraw.

Punch-through techniques for graphics processing

Methods, systems, and computer readable media for preemptive context-switching of processes on an accelerated processing device are based upon a comparison of the running time of the process and a threshold time quanta. A method includes preempting a process running on an accelerated processing device based upon a running time of the process and a threshold time quanta.

Preemptive context switching of processes on ac accelerated processing device (APD) based on time quanta

A method resumes an accelerated processing device (APD) wavefront in which a subset of elements have faulted. A restore command for a job including a wavefront is received. A list of context states for the wavefront is read from a memory associated with a APD. An empty shell wavefront is created for restoring the list of context states. A portion of not acknowledged data is masked over a portion of acknowledged data within the restored wavefronts.

Method for resuming an APD wavefront in which a subset of elements have faulted

Growing community of inventors

Deland, FL, United States of America

Ralph Clay Taylor