System and method for efficient multi-GPU rendering of geometry by pretesting against screen regions using configurable shaders

Patent metadata

Field	Value
Publication number	US-12112394-B2
Application number	US-202016780745-A
Country	US
Kind code	B2
Filing date	Feb 3, 2020
Priority date	Feb 3, 2020
Publication date	Oct 8, 2024
Grant date	Oct 8, 2024

Abstract

A method including rendering graphics for an application using graphics processing units (GPUs). Responsibility for rendering of geometry is divided between GPUs based on screen regions, each GPU having a corresponding division of the responsibility which is known. A plurality of pieces of geometry of an image frame is assigned to the GPUs for geometry testing. A first GPU state configuring one or more shaders to perform the geometry testing is set. Geometry testing is performed at GPUs on the plurality of pieces of geometry to generate information regarding each piece of geometry and its relation to each of the plurality of screen regions. A second GPU state configuring the one or more shaders to perform rendering is set. The information generated for each of the plurality of pieces of geometry is used when rendering the plurality of pieces of geometry at the GPUs.

First claim

What is claimed is:

1. A method for graphics processing, comprising:
rendering graphics for an application using a plurality of graphics processing units (GPUs);
dividing responsibility for the rendering of geometry of the graphics for a plurality of image frames between the plurality of GPUs based on a plurality of screen regions, each GPU having a corresponding division of the responsibility which is known to the plurality of GPUs;
assigning in a frame period a plurality of pieces of geometry of an image frame to the plurality of GPUs for geometry testing;
setting in the frame period a first value in random access memory defining a first GPU state for the plurality of GPUs to perform the geometry testing, wherein the first value is set externally from execution of a command buffer;
calling in the frame period a subroutine a first time to perform the geometry testing based on the first value defining the first GPU state, wherein the subroutine includes a plurality of commands in the command buffer;
in a first pass of the command buffer in the frame period by each of the plurality of GPUs to process the subroutine called for the first time, executing a first group of commands of the plurality of commands to configure each of one or more shaders executed by the plurality of GPUs to perform the geometry testing on the plurality of pieces of geometry of the image frame, wherein execution of a command in the first group of commands is skipped based on the first value as not being required for configuring the each of the one or more shaders to perform the geometry testing;
in the first pass of the command buffer in the frame period, executing remaining commands in the plurality of commands to perform the geometry testing by the one or more shaders executed by the plurality of GPUs on the plurality of pieces of geometry of the image frame to generate information regarding each piece of geometry and its relation to each of the plurality of screen regions for the image frame, wherein the executing of the remaining commands in the plurality of commands to perform the geometry testing includes interpreting a subset of the remaining commands using a first interpretation based on the first value defining the first GPU state;
setting in the frame period a second value to the random access memory defining a second GPU state for the plurality of GPUs to perform rendering of the image frame using the plurality of commands in the command buffer, wherein the second configuration value is set externally from execution of the command buffer;
calling in the frame period the subroutine for a second time to perform the rendering of the image frame based on the second value defining the second GPU state;
in a second pass of the command buffer in the frame period by each of the plurality of GPUs to process the subroutine called for the second time, executing the first group of commands to configure the each of the one or more shaders executed by the plurality of GPUs to perform the rendering of the image frame, wherein execution of the command in the first group of commands is performed based on the second value as being required for configuring the each of the one or more shaders to perform the rendering of the image frame; and
in the second pass of the command buffer in the frame period, using the information generated for each of the plurality of pieces of geometry of the image frame when executing the remaining commands in the plurality of commands to perform the rendering of the plurality of pieces of geometry of the image frame by the plurality of GPUs to render the image frame,
wherein a corresponding shader as executed by a corresponding GPU is configured to perform the geometry testing on a piece of geometry of a corresponding object of the image frame in the first pass of the command buffer in the frame period, and the corresponding shader as executed by the corresponding GPU is configured to perform rendering of the piece of geometry of the corresponding object of the image frame in the second pass of the command buffer in the frame period, and
wherein the executing the remaining commands in the plurality of commands to perform the rendering includes interpreting the subset of the remaining commands using a second interpretation based on the second value defining the second GPU state.

2. The method of claim 1 , wherein the using the information includes:
skipping rendering a piece of geometry at a rendering GPU when the information indicates that the piece of geometry does not overlap any screen region assigned to the rendering GPU for object rendering, wherein the rendering GPU is one of the plurality of GPUs.

3. The method of claim 1 , further comprising:
providing the information as a hint to a rendering GPU, wherein the rendering GPU is one of the plurality of GPUs,
wherein the information is considered by the rendering GPU if received before rendering of the piece of geometry,
wherein the piece of geometry is fully rendered at the rendering GPU when the information is received after rendering of the piece of geometry begins.

4. The method of claim 1 ,
wherein pieces of geometry in the plurality of pieces of geometry are assigned evenly or unevenly throughout the plurality of GPUs,
wherein the plurality of pieces of geometry is assigned such that successive pieces of geometry are processed by different GPUs.

5. The method of claim 1 ,
wherein a first GPU performs geometry pretesting on more pieces of geometry than a second GPU, or the first GPU performs geometry pretesting while the second GPU performs no geometry pretesting at all.

6. The method of claim 1 , wherein depending on whether the plurality of GPUs is configured in the first GPU state or the second GPU state, when commands in the corresponding command buffer are executed, the commands cause the output of the information regarding the piece of geometry or cause the output of vertex position and parameter information for use by later one or more rendering stages,
wherein the corresponding command buffer includes the commands of a subroutine that when executed performs the geometry testing or performs the rendering of the image frame depending on which of the first GPU state or the second GPU state is configured for the one or more shaders.

7. The method of claim 6 , wherein depending on whether the plurality of GPUs is configured in the first GPU state or the second GPU state, the commands that affect the GPU configuration are interpreted in one of two ways.

8. The method of claim 1 , further comprising:
interleaving in a command buffer generation of first information for a first piece of geometry and its relation to the plurality of screen regions and rendering of the first piece of geometry with generation of second information for a second piece of geometry and its relation to the plurality of screen regions and rendering of the second piece of geometry.

9. The method of claim 1 , wherein hardware contexts are preserved, or saved and restored.

10. The method of claim 1 ,
wherein one or more of the plurality of GPUs are portions of a larger GPU that is configured as a plurality of virtual GPUs.

11. A method for graphics processing, comprising:
rendering graphics for an application using a plurality of graphics processing units (GPUs);
dividing responsibility for rendering of geometry of the graphics for a plurality of image frames between the plurality of GPUs based on a plurality of screen regions, each GPU having a corresponding division of the responsibility which is known to the plurality of GPUs;
assigning in a frame period a plurality of pieces of geometry of an image frame to the plurality of GPUs for geometry tes

Assignees

• Sony Interactive Entertainment Inc

Inventors

• Cerny Mark E
• Strauss Florian
• Berghoff Tobias

Classifications

• G06T17/10

  Constructive solid geometry [CSG] using solid primitives, e.g. cylinders, cubes · CPC title

• G06F9/5061

  Partitioning or combining of resources · CPC title

• G06F9/3877

  using a secondary processor, e.g. coprocessor (peripheral processor G06F13/12) · CPC title

• G06F9/461

  Saving or restoring of program or task context · CPC title

• G06T15/10

  Geometric effects · CPC title