# US 2017/0287202 A1 · Methods and apparatus for more efficient ray tracing of instanced geometry

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

- Number: US 2017/0287202 A1
- Document ID: US20170287202A1
- Status: Published application
- Kind: A1
- Application number: 15/086,033
- Filed: 2016-03-30
- Published: 2017-10-05
- Earliest claimed: 2016-03-30
- Assignees: Intel IP Corporation
- Applicants: Intel IP Corporation
- Inventors: Ingo Wald; Sven Woop; Carsten Benthin
- Web page: https://patentgrep.com/patent/US20170287202A1

## CPC

- G06T 15/06: Ray-tracing
- G06T 15/005: {General purpose rendering architectures}
- G06T: IMAGE DATA PROCESSING OR GENERATION, IN GENERAL

## Abstract

Methods and apparatus relating to more efficient ray tracing of instanced geometry are described. In an embodiment, overlapping instances are unbraided, by not instantiating the entire objects, but instantiating multiple sub-BVH nodes of the objects, which improves rendering performance by reducing overlap of BVH nodes. Other embodiments are also disclosed and claimed.

## Claims

### Claim 1

An apparatus comprising:
memory to store information corresponding to a two-level Bounding Volume Hierarchy (BVH) structure; and
a processor, coupled to the memory, to build and traverse the two-level BVH structure over instanced geometries,
wherein each leaf node of a top-level of the two-level BVH structure is to store a transformation and a reference to a node of a geometry BVH, wherein the processor is to add one or more nodes to the BVH structure, wherein the one or more nodes are capable to directly point to one or more locations of an instanced object to allow for a more efficient traversal of the BVH structure.

### Claim 2 (depends on 1)

The apparatus of claim 1, wherein the processor is to perform an operation to reduce spatial overlap in a list of instanced BVH nodes by repeatedly replacing one or more instanced nodes with an instancing of their child nodes until some termination criteria is fulfilled.

### Claim 3 (depends on 2)

The apparatus of claim 2, wherein the processor is to perform the operation as a preprocessing pass to generate the list of instanced BVH nodes as input to a BVH building procedure.

### Claim 4 (depends on 2)

The apparatus of claim 2, wherein the processor is to perform the operation to split leaf nodes of an existing top-level BVH, followed by a pass to improve the top-level of the BVH.

### Claim 5 (depends on 2)

The apparatus of claim 2, wherein the processor is to apply the operation to a list of build primitives during a top-down BVH build procedure.

### Claim 6 (depends on 5)

The apparatus of claim 5, wherein the processor is to apply the operation to build primitives when generation of the primitives improves a cost of splitting heuristic used during the generation of primitives.

### Claim 7 (depends on 1)

The apparatus of claim 1, wherein the processor is to modify the BVH structure to allow one or more leaf nodes in the top-level of the BVH structure to reference one or more BVH sub-trees, wherein each leaf node of the top-level BVH is to correspond to a pair of instance identifier and a sub-tree identifier.

### Claim 8 (depends on 7)

The apparatus of claim 7, wherein the instance identifier is to correspond to a transformation operation.

### Claim 9 (depends on 8)

The apparatus of claim 8, wherein information corresponding to the transformation operation is to be cached.

### Claim 10

(canceled)

### Claim 11 (depends on 1)

The apparatus of claim 1, wherein the processor is to add one or more nodes to a top level of the BVH structure, wherein the one or more nodes of the tope level are capable to directly point to one or more locations of the instanced object.

### Claim 12 (depends on 1)

The apparatus of claim 1, wherein the processor is to comprise one or more of:
a Graphics Processing Unit (GPU) and a processor core.

### Claim 13 (depends on 12)

The apparatus of claim 12, wherein the GPU is to comprise one or more graphics processing cores.

### Claim 14 (depends on 1)

The apparatus of claim 1, wherein the processor is to comprise one or more processor cores.

### Claim 15 (depends on 1)

The apparatus of claim 1, wherein the processor is to comprise at least a portion of the memory.

### Claim 16 (depends on 1)

The apparatus of claim 1, wherein the processor and the memory are on a single integrated circuit die.

### Claim 17

A method comprising:
storing information corresponding to a two-level Bounding Volume Hierarchy (BVH) structure in memory; and
building and traversing, at a processor, the two-level BVH structure over instanced geometries,
wherein each leaf node of a top-level of the two-level BVH structure stores a transformation and a reference to a node of a geometry BVH, wherein one or more nodes are added to the BVH structure, wherein the one or more nodes are capable to directly point to one or more locations of an instanced object to allow for a more efficient traversal of the BVH structure.

### Claim 18 (depends on 17)

The method of claim 17, further comprising the processor performing an operation to reduce spatial overlap in a list of instanced BVH nodes by repeatedly replacing one or more instanced nodes with an instancing of their child nodes until some termination criteria is fulfilled.

### Claim 19 (depends on 18)

The method of claim 18, further comprising the processor performing the operation as a preprocessing pass to generate the list of instanced BVH nodes as input to a BVH building procedure.

### Claim 20 (depends on 18)

The method of claim 18, further comprising the processor performing the operation to split leaf nodes of an existing top-level BVH, followed by a pass to improve the top-level of the BVH.

### Claim 21 (depends on 18)

The method of claim 18, further comprising the processor performing the operation during a top-down BVH build procedure to a list of build primitives that is recursively handled.

### Claim 22 (depends on 21)

The method of claim 21, further comprising the processor performing the operation to build primitives when generation of the primitives improves a cost of splitting heuristic used during the generation of the primitives.

### Claim 23

One or more non-transitory computer-readable medium comprising one or more instructions that when executed on at least one processor configure the at least one processor to perform one or more operations to:
store information corresponding to a two-level Bounding Volume Hierarchy (BVH) structure in memory; and
building and traversing the two-level BVH structure over instanced geometries,
wherein each leaf node of a top-level of the two-level BVH structure stores a transformation and a reference to a node of a geometry BVH, wherein the processor is to add one or more nodes to the BVH structure, wherein the one or more nodes are capable to directly point to one or more locations of an instanced object to allow for a more efficient traversal of the BVH structure.

### Claim 24 (depends on 23)

The one or more non-transitory computer-readable medium of claim 23, further comprising one or more instructions that when executed on the at least one processor configure the at least one processor to perform one or more operations to cause the processor to perform an operation to reduce spatial overlap in a list of instanced BVH nodes by repeatedly replacing one or more instanced nodes with an instancing of their child nodes until some termination criteria is fulfilled.

### Claim 25 (depends on 24)

The one or more non-transitory computer-readable medium of claim 24, further comprising one or more instructions that when executed on the at least one processor configure the at least one processor to perform one or more operations to cause the processor to perform the operation performing the operation as a preprocessing pass to generate the list of instanced BVH nodes as input to a BVH building procedure.
