Early termination in bottom-up acceleration data structure refit

Patent metadata

Field	Value
Publication number	US-11158111-B2
Application number	US-202016915599-A
Country	US
Kind code	B2
Filing date	Jun 29, 2020
Priority date	Dec 28, 2018
Publication date	Oct 26, 2021
Grant date	Oct 26, 2021

Abstract

Apparatus and method for bottom-up BVH refit. For example, one embodiment of an apparatus comprises: a hierarchical acceleration data structure generator to construct an acceleration data structure comprising a plurality of hierarchically arranged nodes; traversal hardware logic to traverse one or more rays through the acceleration data structure; intersection hardware logic to determine intersections between the one or more rays and one or more primitives within the hierarchical acceleration data structure; a node unit comprising circuitry and/or logic to perform refit operations on nodes of the hierarchical acceleration data structure, the refit operations to adjust spatial dimensions of one or more of the nodes; and an early termination evaluator to determine whether to proceed with refit operations or to terminate refit operations for a current node based on refit data associated with one or more child nodes of the current node.

First claim

What is claimed is:

1. An apparatus comprising:
a hierarchical acceleration data structure generator to construct an acceleration data structure comprising a plurality of hierarchically arranged nodes;
traversal hardware logic to traverse one or more rays through the hierarchical acceleration data structure;
intersection hardware logic to determine intersections between the one or more rays and one or more primitives within the hierarchical acceleration data structure; and
a node refit unit comprising circuitry and/or logic to perform refit operations on nodes of the hierarchical acceleration data structure to adjust spatial dimensions of one or more of the nodes,
wherein the node refit unit is to conditionally refit a current node based on refit data associated with one or more child nodes of the current node,
wherein if the current node needs to be reduced, the node refit unit is to refit the current node only if the refit is justified based on benefits and/or costs associated therewith.

2. The apparatus of claim 1 wherein if the current node needs to be enlarged, the node refit unit is to always refit the current node.

3. The apparatus of claim 2 wherein if the current node needs to be reduced, the node refit unit is also to refit the current node if the refit is justified based on one or more threshold values being reached.

4. The apparatus of claim 3 wherein one of the thresholds comprises a ratio of volume occupancy of the one or more child nodes over the volume of the current node.

5. The apparatus of claim 3 wherein one of the thresholds comprises degraded performance over a threshold period of time or number of processing cycles.

6. The apparatus of claim 1 wherein the node refit unit is to perform a first sequence of operations to iterate over all leaf nodes which point to leaf data, the node refit unit to update bounding volumes associated with one or more of the leaf nodes.

7. The apparatus of claim 1 wherein the hierarchical acceleration data structure comprises a bounding volume hierarchy (BVH).

8. The apparatus of claim 1 wherein refit operations include adjusting dimensions of the current node based on changes to dimensions or locations of the one or more child nodes.

9. The apparatus of claim 1 , wherein the node refit unit is to terminate the refit operation for the current node based on refit data associated with one or more child nodes of the current node.

10. A method comprising:
constructing a hierarchical acceleration data structure comprising a plurality of hierarchically arranged nodes;
traversing one or more rays through the hierarchical acceleration data structure;
determining intersections between the one or more rays and one or more primitives within the hierarchical acceleration data structure;
performing refit operations on nodes of the hierarchical acceleration data structure to adjust spatial dimensions of one or more of the nodes; and
conditionally refitting a current node based on refit data associated with one or more child nodes of the current node,
wherein conditionally refitting the current node comprises that if the current node needs to be reduced, refitting the current node only if the refit is justified based on benefits and/or costs associated therewith.

11. The method of claim 10 wherein conditionally refitting the current node further comprises always refitting the current node if the current node needs to be enlarged.

12. The method of claim 11 wherein conditionally refitting the current node further comprises that if the current node needs to be reduced, refitting the current node if the refit is justified based on one or more threshold values being reached.

13. The method of claim 12 wherein one of the thresholds comprises a ratio of volume occupancy of the one or more child nodes over the volume of the current node.

14. The method of claim 12 wherein one of the thresholds comprises degraded performance over a threshold period of time or number of processing cycles.

15. The method of claim 11 further comprises performing a first sequence of operations to iterate over all leaf nodes which point to leaf data and updating bounding volumes associated with one or more of the leaf nodes.

16. The method of claim 11 wherein the hierarchical acceleration data structure comprises a bounding volume hierarchy (BVH).

17. The method of claim 11 further comprise adjusting dimensions of the current node based on changes to dimensions or locations of the one or more child nodes.

18. A non-transitory machine-readable medium having program code stored thereon which, when executed by a machine, causes the machine to perform the operations of:
constructing a hierarchical acceleration data structure comprising a plurality of hierarchically arranged nodes;
traversing one or more rays through the hierarchical acceleration data structure;
determining intersections between the one or more rays and one or more primitives within the hierarchical acceleration data structure;
performing refit operations on nodes of the hierarchical acceleration data structure to adjust spatial dimensions of one or more of the nodes; and
conditionally refitting a current node based on refit data associated with one or more child nodes of the current node,
wherein conditionally refitting the current node comprises that if the current node needs to be reduced, refitting the current node only if the refit is justified based on benefits and/or costs associated therewith.

19. The non-transitory machine-readable medium of claim 18 wherein conditionally refitting the current node further comprises always refitting the current node if the current node needs to be enlarged.

20. The non-transitory machine-readable medium of claim 19 wherein conditionally refitting the current node further comprises that if the current node needs to be reduced, refitting the current node if the refit is justified based on one or more threshold values being reached.

21. The non-transitory machine-readable medium of claim 20 wherein one of the thresholds comprises a ratio of volume occupancy of the one or more child nodes over the volume of the current node.

22. The non-transitory machine-readable medium of claim 20 wherein one of the thresholds comprises degraded performance over a threshold period of time or number of processing cycles.

23. The non-transitory machine-readable medium of claim 19 wherein the operations further comprise performing a first sequence of operations to iterate over all leaf nodes which point to leaf data and updating bounding volumes associated with one or more of the leaf nodes.

24. The non-transitory machine-readable medium of claim 19 wherein the hierarchical acceleration data structure comprises a bounding volume hierarchy (BVH).

25. The non-transitory machine-readable medium of claim 19 wherein the operations further comprise adjusting dimensions of the current node based on changes to dimensions or locations of the one or more child nodes.

Assignees

• Intel Corp

Inventors

• Xiao Kai
• Apodaca Michael
• Brownlee Carson
• Raoux Thomas
• Barczak Joshua
• Liktor Gabor

Classifications

• G06T2200/04

  involving 3D image data · CPC title

• G06T15/06 Primary

  Ray-tracing · CPC title

• G06T1/20

  Processor architectures; Processor configuration, e.g. pipelining · CPC title

• G06T15/205 Primary

  Image-based rendering · CPC title