# US 2021/0241410 A1 · System and method for efficient multi-gpu rendering of geometry by region testing while rendering

> Agents: this data is available as JSON from the patentgrep API and MCP server — see https://patentgrep.com/llms.txt

- Number: US 2021/0241410 A1
- Document ID: US20210241410A1
- Status: Published application
- Kind: A1
- Application number: 16/780,566
- Filed: 2020-02-03
- Published: 2021-08-05
- Earliest claimed: 2020-02-03
- Assignees: Sony Interactive Entertainment Inc.
- Applicants: Sony Interactive Entertainment Inc.
- Inventors: Mark E. Cerny; Florian Strauss; Tobias Berghoff
- Web page: https://patentgrep.com/patent/US20210241410A1

## CPC

- G06T 1/20: Processor architectures; Processor configuration, e.g. pipelining
- G06T 15/005: {General purpose rendering architectures}
- G06T: IMAGE DATA PROCESSING OR GENERATION, IN GENERAL

## Abstract

A method for graphics processing. The method including rendering graphics for an application using a plurality of graphics processing units (GPUs). The method including dividing responsibility for the rendering geometry of the graphics 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. The method including generating information regarding a piece of geometry with respect to a first screen region for which a first GPU has a first division of responsibility, while rendering the piece of geometry at a second GPU for an image. The method including rendering the piece of geometry at the first GPU using the information.

## Claims

### Claim 1

A method for graphics processing, comprising:
rendering graphics for an application using a plurality of graphics processing units (GPUs) for presentation on a display;
dividing responsibility for the rendering geometry of the graphics between the plurality of GPUs based on a plurality of screen regions corresponding to the display, each GPU having a corresponding division of the responsibility which is known to the plurality of GPUs;
while rendering a piece of geometry at a second GPU for an image, generating information regarding the piece of geometry with respect to a first screen region for which a first GPU has a first division of responsibility; and
rendering the piece of geometry for the image at the first GPU using the information.

### Claim 2 (depends on 1)

The method of claim 1, wherein the information indicates that the first GPU should skip rendering of the piece of geometry entirely.

### Claim 3 (depends on 1)

The method of claim 1,
wherein the information is a hint to the first GPU, such that the information is considered if received before rendering of the first piece of geometry begins at the first GPU,
wherein the first piece of geometry is fully rendered at the first GPU when the information is received after rendering of the first piece of geometry begins.

### Claim 4 (depends on 1)

The method of claim 1, wherein the plurality of screen regions is configured to reduce imbalance of rendering time between the plurality of GPUs.

### Claim 5 (depends on 1)

The method of claim 1, wherein each of the plurality of screen regions is not uniform in size.

### Claim 6 (depends on 1)

The method of claim 1, wherein the plurality of screen regions change dynamically.

### Claim 7 (depends on 1)

The method of claim 1, wherein the piece of geometry corresponds to a geometry used by or generated by a draw call.

### Claim 8 (depends on 1)

The method of claim 1,
wherein a geometry used by or generated by a draw call is subdivided into a plurality of pieces of geometry,
wherein corresponding information is generated at the second GPU regarding at least one of the plurality of pieces of geometry with respect to the plurality of screen regions.

### Claim 9 (depends on 1)

The method of claim 1, wherein the information regarding the piece of geometry is generated at the second GPU for individual primitives.

### Claim 10 (depends on 1)

The method of claim 1, wherein the information regarding the piece of geometry includes a vertex count or a primitive count.

### Claim 11 (depends on 1)

The method of claim 1, wherein the information generated regarding the piece of geometry includes a specific set of primitives for rendering or a specific set of vertices for rendering.

### Claim 12 (depends on 1)

The method of claim 1, further comprising:
using a common rendering command buffer for the first GPU and the second GPU; and
limiting execution of a command in the common rendering command buffer to one of the first GPU and the second GPU.

### Claim 13 (depends on 1)

The method of claim 1,
wherein the first screen region is part of a first set of screen regions for which the first GPU has the first division of responsibility,
wherein the second GPU has a second division of responsibility for a second set of screen regions including a second screen region.

### Claim 14 (depends on 1)

The method of claim 1,
wherein each of the plurality of GPUs is responsible for rendering geometry in,
one or more screen regions.

### Claim 15 (depends on 1)

The method of claim 1, wherein the first GPU and the second GPU are portions of a larger GPU that is configured to include the plurality of GPUs.

### Claim 16 (depends on 1)

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.

### Claim 17 (depends on 1)

The method of claim 1, wherein the information may be generated, or not generated depending on the properties of the piece of geometry.

### Claim 18 (depends on 1)

The method of claim 1, further comprising:
using a scan converter in a rasterization stage to generate the information.

### Claim 19 (depends on 1)

The method of claim 1, further comprising:
using at least one shader in a geometry processing stage to generate the information.

### Claim 20 (depends on 19)

The method of claim 19, wherein the at least one shader uses at least one dedicated instruction to accelerate the generation of the information.

### Claim 21 (depends on 1)

The method of claim 1, further comprising:
using the information generated in a first phase of rendering in a second phase of rendering.

### Claim 22

A computer system comprising:
a processor;
memory coupled to the processor and having stored therein instructions that, if executed by the computer system, cause the computer system to execute a method for implementing a graphics pipeline, comprising:
rendering graphics for an application using a plurality of graphics processing units (GPUs) for presentation on a display;
dividing responsibility for the rendering geometry of the graphics between the plurality of GPUs based on a plurality of screen regions corresponding to the display, each GPU having a corresponding division of the responsibility which is known to the plurality of GPUs;
while rendering a piece of geometry at a second GPU for an image, generating information regarding the piece of geometry with respect to a first screen region for which a first GPU has a first division of responsibility; and
rendering the piece of geometry for the image at the first GPU using the information.

### Claim 23 (depends on 22)

The computer system of claim 22,
wherein in the method the first screen region is part of a first set of screen regions for which the first GPU has the first division of responsibility,
wherein in the method the second GPU has a second division of responsibility for a second set of screen regions including a second screen region.

### Claim 24 (depends on 22)

The computer system of claim 22,
wherein in the method each of the plurality of GPUs is responsible for rendering geometry in a corresponding set of screen regions,
wherein in the method the corresponding set of screen regions incudes one or more screen regions.

### Claim 25 (depends on 22)

The computer system of claim 22,
wherein in the method the information indicates that the first GPU should skip rendering of the piece of geometry entirely.

### Claim 26 (depends on 22)

The computer system of claim 22,
wherein in the method the information is a hint to the first GPU, such that the information is considered if received before rendering of the first piece of geometry begins at the first GPU,
wherein in the method the first piece of geometry is fully rendered at the first GPU when the information is received after rendering of the first piece of geometry begins.
