Modular Space Division for Interprocessor Communication
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Current data-parallel ray tracing technologies face challenges in load balancing and efficiency due to variations in ray distribution and object density, leading to severe load imbalances and high processing costs, especially when dealing with large and dynamic scenes.
Innovation Solution
The method employs a modular mapping of scene data into a uniform grid, allowing for neighbor-based data exchange and dynamic load balancing between processors, reducing the need for global communication and leveraging shared cache-memory access for efficient data transfer.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Loss of energy
If data-parallel ray tracing is used with uniform grid distribution, then data locality is improved and massive data moves are reduced, but load balancing deteriorates due to variations in ray distribution and object density
Solution Approach 1:
The scene is divided into multiple subspace regions, each assigned to a different processor. This segmentation allows each processor to handle a specific portion of the scene data, improving data locality while enabling load balancing through region-based task distribution.
Solution Approach 2:
The system implements dynamic task assignment where processors can request and receive new ray-tracing tasks from a master processor. This dynamic approach allows the system to adapt to varying workloads and maintain load balancing efficiency despite variations in ray distribution and object density across different regions.
2Ease of operation
If demand driven approach is used with image space subdivision, then task distribution is simplified, but memory requirements worsen due to need for massive copies of geometric data
Solution Approach 1:
The scene is divided into multiple subspace regions, each assigned to a different processor. This segmentation allows each processor to handle a specific portion of the scene data, improving data locality while enabling load balancing through region-based task distribution.
Solution Approach 2:
The system uses a master processor to distribute tasks to worker processors. When a worker processor finishes its current task, it requests the next task from the master processor, which copies only the necessary task data rather than requiring all geometric data to be replicated across all processors.
Data Source
AI summary
Novel method and system for distributed database ray-tracing is presented, based on modular mapping of scene-data among processors. Its inherent properties include scattering data among processors for improved load balancing, and matching between geographical proximity in the scene with communication proximity between processors. High utilization is enabled by unique mechanism of cache sharing. The resulting improved performance enables deep level of ray tracing for real time applications.