Cull streams for fine-grained rendering predication

• US 9,489,767 B1
• Filed: 12/13/2007
• Issued: 11/08/2016
• Est. Priority Date: 12/13/2007
• Status: Active Grant

• Current Assignee: NVIDIA Corporation
• Original Assignee: NVIDIA Corporation
• Inventors: Everitt, Cass W., Diard, Franck R.
• Primary Examiner(s): McDowell, Jr., Maurice L
• Application Number:
  US11/956,306
• Time in Patent Office: 3,253 Days
• Field of Search: None
• US Class Current: 1/1
• CPC Class Codes: G06T 15/005 General purpose rendering architectures
• CPC Class Codes: G06T 15/40 Hidden part removal
• CPC Class Codes: G06T 15/405 using Z-buffer
• CPC Class Codes: G06T 15/80 Shading

Abstract

One embodiment of the present invention sets forth a technique to perform fine-grained rendering predication using an IGPU and a DGPU. A graphics driver divides a 3D object into batches of triangles. The IGPU processes each batch of triangles through a modified rendering pipeline to determine if the batch is culled. The IGPU writes bits into a bitstream corresponding to the visibility of the batches. The DGPU reads bits from the bitstream and performs full-blown rendering, including shading, but only on the batches of triangles whose bit indicates that the batch is visible. Advantageously, this approach to rendering predication provides fine-grained culling without adding unnecessary overhead, thereby optimizing both hardware resources and performance.

Claims

1. A computer-implemented method for fine-grained rendering predication, the method comprising:

receiving a batch of triangles associated with a graphics object;

processing the batch of triangles to determine only position information in a modified graphics rendering pipeline;

determining, based on the position information, if the batch of triangles is to be culled; and

writing a bit into a bitstream residing in a memory indicating whether the batch of triangles is to be culled.

2. The method of claim 1, wherein processing the batch of triangles comprises determining if the batch of triangles is outside a viewport.

3. The method of claim 1, wherein processing the batch of triangles comprises performing limited fragment shading operations to determine whether the batch of triangles fails a depth test.

4. The method of claim 1, wherein processing the batch of triangles comprises performing limited raster operations to determine whether the batch of triangles fails a stencil test.

5. The method of claim 1, wherein if the bit written into the bitstream indicates that the batch of triangles is to be culled, then further processing of the batch of triangles is bypassed.

6. The method of claim 5, wherein the further processing is performed in a discrete graphics processing unit that is configured to perform color-based and texture-based shading operations.

7. The method of claim 6, wherein the discrete graphics processing unit is further configured to render the batch of triangles for a frame even when the modified graphics rendering pipeline has not finished processing the batch of triangles for the frame.

8. The method of claim 1, wherein the batch of triangles is included in a second frame and a first batch of triangles included in a first frame is rendered by a graphics rendering pipeline implemented in a discrete graphics processor that performs color-based and texture-based shading operations while the batch of triangles is processed in the modified graphics rendering pipeline.

9. The method of claim 1, further comprising, prior to processing the batch of triangles, the bitstream is initialized to indicate that the batch of triangles is visible.

10. The method of claim 1, wherein a single bit is written to the bitstream for the batch of triangles and the batch of triangles includes at least two triangles.

11. A system, comprising:

an interface that receives a batch of triangles associated with a graphics object; and

an integrated graphics processor configured to;

process the batch of triangles to determine only position information;

determine, based on the position information, if the batch of triangles is to be culled, and

write a bit into a bitstream residing in a memory indicating whether the batch of triangles is to be culled.

12. The system of claim 11, wherein the integrated graphics processor is further configured to:

generate modified graphics primitives based on the vertex data; and

determine that the batch of triangles is to be culled if no modified graphics primitives are generated.

13. The system of claim 11, wherein the integrated graphics processor processes the batch of triangles by determining if the batch of triangles is outside a viewport.

14. The system of claim 11, wherein the integrated graphics processor processes the batch of triangles by performing limited fragment shading operations to determine whether the batch of triangles fails a depth test.

15. The system of claim 11, wherein the integrated graphics processor processes the batch of triangles by performing limited raster operations to determine whether the batch of triangles fails a stencil test.

16. The system of claim 11, wherein if the bit written into the bitstream indicates that the batch of triangles is to be culled, then further processing of the batch of triangles is bypassed.

17. The system of claim 16, wherein further processing is performed in a discrete graphics processor that is configured to perform color-based and texture-based shading operations.

18. The system of claim 17, wherein the discrete graphics processor is further configured to render the batch of triangles for a frame even when the integrated graphics processor has not finished processing the batch of triangles for the frame.

19. The system of claim 11, wherein the batch of triangles is included in a second frame, and further comprising a discreet graphics processor configured to perform color-based and texture-based shading operations to render a first batch of triangles that is included in a first frame while the batch of triangles is processed by the integrated graphics processor.

20. The system of claim 11, wherein a single bit is written to the bitstream for the batch of triangles, and the batch of triangles includes at least two triangles.