Reduced Acceleration Structures for Ray Tracing Memory
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Current ray tracing systems face challenges in efficiently processing large 3D scenes due to high processing power and memory requirements, particularly on devices with tight constraints on silicon area, cost, and power consumption, such as mobile devices.
Innovation Solution
The generation of reduced acceleration structures, which represent subsets of the 3D scene, reduces the number of nodes and data required for intersection testing, thereby decreasing processing load and memory storage needs.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If a complete acceleration structure representing the entire 3D scene is used for intersection testing, then comprehensive scene coverage is achieved, but processing load and memory requirements increase significantly
Solution Approach 1:
The patent divides the complete acceleration structure into multiple reduced acceleration structures, each representing a subset of the 3D scene. This segmentation allows the system to process only relevant portions of the scene for each ray tracing operation, reducing the number of nodes that need to be traversed and tested for intersection, thereby lowering processing load while maintaining accurate intersection detection where needed.
Solution Approach 2:
The patent creates acceleration structures with varying levels of detail and completeness based on local requirements. Reduced acceleration structures contain only the necessary nodes and geometry data for specific regions of interest, allowing high-quality intersection testing in critical areas while using simplified representations in less important regions, thus optimizing the balance between accuracy and processing complexity.
2Measurement precision
If a complete acceleration structure representing the entire 3D scene is used for intersection testing, then comprehensive scene coverage is achieved, but memory storage requirements increase
Solution Approach 1:
The patent segments the large complete acceleration structure into multiple smaller reduced acceleration structures, each storing only the necessary nodes and geometry data for a specific subset of the scene. This reduces the total memory footprint by eliminating redundant data storage, as each reduced structure contains only the minimal required information for its designated region rather than duplicating the entire scene data.
Solution Approach 2:
The patent extracts only the essential nodes and geometry data needed for specific intersection testing scenarios from the complete acceleration structure. By taking out and storing only the relevant portions in reduced acceleration structures, the system maintains intersection testing accuracy for critical elements while significantly reducing overall memory storage requirements by excluding unnecessary data.
3Productivity
If traditional acceleration structures are used in ray tracing systems, then intersection testing can be performed, but latency and power consumption remain high on resource-constrained devices
Solution Approach 1:
The patent applies partial action by using reduced acceleration structures that contain only the necessary nodes and data for specific subsets of the scene rather than processing the complete structure. This allows the system to perform sufficient intersection testing for the required regions of interest without the excessive processing overhead of complete scene analysis, thereby reducing latency while maintaining adequate ray tracing functionality for real-time rendering.
4Productivity
If traditional acceleration structures are used in ray tracing systems, then intersection testing can be performed, but power consumption remains high on resource-constrained devices
Solution Approach 1:
The patent segments the processing workload by dividing the complete acceleration structure into multiple reduced acceleration structures. This segmentation enables the system to process only the necessary subsets of scene data for each ray tracing operation, reducing the total number of computational operations required. Consequently, power consumption is lowered while maintaining adequate ray tracing processing capability for real-time applications on resource-constrained devices.
Data Source
AI summary
Ray tracing units, processing modules and methods are described for generating one or more reduced acceleration structures to be used for intersection testing in a ray tracing system for processing a 3D scene. Nodes of the reduced acceleration structure(s) are determined, wherein a reduced acceleration structure represents a subset of the 3D scene. The reduced acceleration structure(s) are stored for use in intersection testing. Since the reduced acceleration structures represent a subset of the scene (rather than the whole scene) the memory usage for storing the acceleration structure is reduced, and the latency in the traversal of the acceleration structure is reduced.