# US 2026/0065589 A1 · Bounding volume hierarchy caching by storing topology

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

- Number: US 2026/0065589 A1
- Document ID: US20260065589A1
- Status: Published application
- Kind: A1
- Application number: 18/816,548
- Filed: 2024-08-27
- Published: 2026-03-05
- Earliest claimed: 2024-08-27
- Assignees: Advanced Micro Devices, Inc.
- Applicants: Advanced Micro Devices, Inc.
- Inventors: Andreas Leitner; David Ronald Oldcorn
- Web page: https://patentgrep.com/patent/US20260065589A1

## CPC

- G06T 17/005: {Tree description, e.g. octree, quadtree}
- G06T 1/60: Memory management
- G06T 15/005: {General purpose rendering architectures}
- G06T 15/06: Ray-tracing
- G06T 2210/12: Bounding box
- G06T: IMAGE DATA PROCESSING OR GENERATION, IN GENERAL

## Abstract

In order to perform ray tracing operations, a ray tracing pipeline uses one or more bounding volume hierarchies (“BVHs”) that act as an acceleration structure for accessing the geometry of a scene. In applications such as video games, simulations, or other real-time applications, geometry of the scene changes frequently such as at every frame. Thus, to have appropriate information for ray tracing, a BVH is built quite frequently. Such an operation is an expensive one. Thus, efficient techniques for BVH construction are desirable. Thus, techniques are provided herein for BVH caching BVH information. In general, these caching operations store portions of a BVH in a cache when such portions are initially encountered. Later, when that geometry is re-accessed, the caching operations search the cache for the BVH portions in the cache and uses such BVH portions if found. In some examples, the BVH cache persists between application executions.

## Claims

### Claim 1

A method comprising:
storing acceleration structure topology corresponding to vertex information into a cache;
retrieving the acceleration structure topology from the cache based on the vertex information; and
rebuilding an acceleration structure based on the acceleration structure topology and the vertex information.

### Claim 2 (depends on 1)

The method of claim 1, further comprising requesting a driver to generate the acceleration structure topology based on the vertex information.

### Claim 3 (depends on 2)

The method of claim 2, wherein the requesting and the storing are performed by an application via an application programming interface.

### Claim 4 (depends on 1)

The method of claim 1, wherein the retrieving is performed in a subsequent frame as the storing.

### Claim 5 (depends on 1)

The method of claim 1, wherein the retrieving is performed using a key generated from the vertex information.

### Claim 6 (depends on 1)

The method of claim 1, wherein the rebuilding comprises inserting the vertex information into the vertex information.

### Claim 7 (depends on 6)

The method of claim 6, wherein the rebuilding also comprises performing a refit information.

### Claim 8 (depends on 1)

The method of claim 1, wherein the acceleration structure topology stores oriented bounding box information.

### Claim 9 (depends on 1)

The method of claim 1, wherein the storing is performed in response to the acceleration structure topology not being stored in the cache.

### Claim 10

A system comprising:
a memory; and
a processor configured to perform operations comprising:
storing acceleration structure topology corresponding to vertex information into a cache of the memory;
retrieving the acceleration structure topology from the cache based on the vertex information; and
rebuilding a acceleration structure based on the acceleration structure topology and the vertex information.

### Claim 11 (depends on 10)

The system of claim 10, wherein the operations further comprise requesting a driver to generate the acceleration structure topology based on the vertex information.

### Claim 12 (depends on 11)

The system of claim 11, wherein the requesting and the storing are performed by an application via an application programming interface.

### Claim 13 (depends on 10)

The system of claim 10, wherein the retrieving is performed in a subsequent frame as the storing.

### Claim 14 (depends on 10)

The system of claim 10, wherein the retrieving is performed using a key generated from the vertex information.

### Claim 15 (depends on 10)

The system of claim 10, wherein the rebuilding comprises inserting the vertex information into the vertex information.

### Claim 16 (depends on 15)

The system of claim 15, wherein the rebuilding also comprises performing a refit information.

### Claim 17 (depends on 10)

The system of claim 10, wherein the acceleration structure topology stores oriented bounding box information.

### Claim 18 (depends on 10)

The system of claim 10, wherein the storing is performed in response to the acceleration structure topology not being stored in the cache.

### Claim 19

A non-transitory computer-readable medium storing instructions that, when executed by a processor, cause the processor to perform operations comprising:
storing acceleration structure topology corresponding to vertex information into a cache;
retrieving the acceleration structure topology from the cache based on the vertex information; and
rebuilding an acceleration structure based on the acceleration structure topology and the vertex information.

### Claim 20 (depends on 19)

The non-transitory computer-readable medium of claim 19, wherein the operations further comprise requesting a driver to generate the acceleration structure topology based on the vertex information.
