Wide BVH Node Prefetching for Real-Time Ray Traversal

Resolve Bottlenecks,
Find Innovative Solutions
Generate Solutions

Solution Overview

Problem

Existing graphics processing technologies face challenges in efficiently traversing and intersecting nodes in a bounding volume hierarchy (BVH) during ray tracing, which is resource-intensive and limits real-time performance.

Innovation Solution

Implementing node prefetching techniques in a wide BVH traversal with a stack to optimize the traversal process, utilizing a compressed traversal stack and on-demand builder for efficient BVH processing.

Engineering Contradictions & Design Principles

VSEngineering Contradiction Analysis

1Productivity

If traditional BVH traversal methods are used, then the traversal process is simple to implement, but resource consumption is high and real-time performance is limited

Engineering Contradiction:
Improvereal-time performanceVSAvoidresource consumption
Core Design Contradiction:
ProductivityVSUse of energy by moving object

Solution Approach 1:

The patent implements node prefetching by predicting future BVH nodes that will be accessed during ray traversal and loading them into cache memory in advance. This preliminary action reduces wait time during traversal and improves real-time performance while optimizing resource utilization by pre-loading only the necessary nodes.

Inventive Principle:
Principle #10Preliminary action

Solution Approach 2:

The patent divides the BVH traversal process into multiple independent stages using a stack-based approach, where nodes are processed in discrete segments. This segmentation allows for efficient parallel processing and cache utilization, reducing resource consumption while maintaining high productivity in real-time rendering.

Inventive Principle:
Principle #1Segmentation

2Productivity

If node prefetching is implemented, then traversal efficiency is improved, but device complexity increases

Engineering Contradiction:
Improvetraversal efficiencyVSAvoidsystem complexity
Core Design Contradiction:
ProductivityVSDevice complexity

Solution Approach 1:

The patent implements an on-demand builder that automatically constructs BVH structures during runtime based on the actual traversal needs. This self-service mechanism eliminates the need for manual BVH construction and optimization, improving traversal efficiency while keeping the system complexity manageable through automated adaptation.

Inventive Principle:
Principle #25Self-service

Solution Approach 2:

The patent uses a dynamic stack-based traversal approach where the traversal depth and node loading are adapted in real-time based on the ray traversal state. This dynamic behavior optimizes traversal efficiency by loading only necessary nodes while maintaining relatively simple device architecture through flexible runtime adaptation.

Inventive Principle:
Principle #15Dynamics

Data Source

PatentUS12561753B2Node prefetching in a wide BVH traversal with a stack
Publication Date: 2026.02.24 INTEL CORP
  • US12561753B2 patent drawing
  • US12561753B2 patent drawing
  • US12561753B2 patent drawing

AI summary

Apparatus and method for prefetching node data. For example, one embodiment of an apparatus comprises: ray tracing acceleration hardware to manage active ray tracing stack allocations within a traversal stack; and stack management hardware logic to issue a prefetch operation comprising an indication of bounding volume hierarchy (BVH) node data to be prefetched and an indication of a cache level, wherein responsive to performing the prefetch operation, the BVH node data is to be prefetched to the indicated cache level.