Multi-GPU Rendering via Dynamic Screen Region Assignment
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing technologies face challenges in efficiently utilizing multiple GPUs to render complex images, as they struggle to equally distribute work among GPUs, leading to inefficiencies in handling increased pixel counts and geometric complexity.
Innovation Solution
The method involves using multiple GPUs in collaboration to render an image frame by performing geometry analysis during or prior to rendering, and dynamically assigning screen regions to GPUs based on this analysis, as well as subdividing geometry into smaller portions for more efficient processing.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Productivity
If multiple GPUs are used to render complex scenes, then rendering capability and pixel count are improved, but workload distribution among GPUs becomes unbalanced
Solution Approach 1:
The patent divides the screen into multiple screen regions and assigns different regions to different GPUs. This segmentation approach allows each GPU to handle a specific portion of the rendering task, thereby improving workload distribution while maintaining enhanced rendering capability through multi-GPU collaboration.
2Power
If traditional multi-GPU rendering is used, then processing power is increased, but duplication of effort occurs during rendering
Solution Approach 1:
The patent performs geometry analysis to determine which geometry pieces are visible in which screen regions, then assigns each screen region to the GPU best suited for rendering it. This local optimization ensures that each GPU processes only the geometry necessary for its assigned regions, eliminating duplication of effort while maintaining high processing power.
3Ease of manufacture
If screen regions are statically assigned to GPUs, then setup is simplified, but adaptability to different geometry distributions is reduced
Solution Approach 1:
The patent implements dynamic screen region assignment by performing geometry analysis during or prior to rendering to generate information about geometry visibility, then assigning screen regions to GPUs based on this analyzed information. This dynamic approach allows the system to adapt to different geometry distributions while maintaining reasonable setup complexity through automated assignment.
Data Source
AI summary
A method for graphics processing. The method including rendering graphics for an application using a plurality of graphics processing units (GPUs). The method including using the plurality of GPUs in collaboration to render an image frame including a plurality of pieces of geometry. The method including during a pre-pass phase of rendering, generating information at the GPUs regarding the plurality of pieces of geometry and their relation to a plurality of screen regions. The method including assigning the plurality of screen regions to the plurality of GPUs based on the information for purposes of rendering the plurality of pieces of geometry in a subsequent phase of rendering.