Graphics Edge Coverage Caching for Rasterization Power Reduction
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Highly parallel rasterizers consume excessive power when rendering graphics constructs that share edges, such as triangle meshes, due to redundant sampling of shared edges.
Innovation Solution
A method for caching graphics primitive edge coverage, where a cache read tag is computed based on a grid identifier and edge identifier, allowing stored coverage information to be read from a cache line when edges are shared between graphics primitives, thereby minimizing redundant sampling.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Speed
If highly parallel rasterization techniques are used to increase rasterization parallelism, then processing speed is improved, but power consumption increases due to redundant sampling of shared edges
Solution Approach 1:
The patent applies preliminary action by computing and storing coverage information for shared edges in a cache before they are needed by multiple triangles. When the first triangle is processed, its edge coverage is calculated and cached. When the second triangle shares this edge, the cached coverage is retrieved rather than recomputed, preventing redundant processing and reducing power consumption while maintaining high parallelism
Solution Approach 2:
The patent uses copying by creating a cache copy of the coverage information for shared edges. Instead of having each triangle independently sample and compute coverage for shared edges, the system creates a cached copy of the coverage data that can be reused by multiple triangles, eliminating redundant computations and reducing energy usage
2Productivity
If every sample within a screen-space region is tested in parallel against each triangle, then rasterization parallelism is increased, but the same edge is sampled multiple times for adjacent triangles
Solution Approach 1:
The patent applies merging by combining the coverage computation for adjacent triangles that share edges. Instead of treating each triangle independently and sampling shared edges multiple times, the system merges the coverage computation by caching the result from the first triangle and reusing it for the second triangle, eliminating redundant sampling operations and reducing energy waste
Data Source
AI summary
A technique for caching coverage information for edges that are shared between adjacent graphics primitives may reduce the number of times a shared edge is rasterized. Consequently, power consumed during rasterization may be reduced. During rasterization of a first graphics primitive coverage information is generated that (1) indicates cells within a sampling grid that are entirely outside an edge of the first graphics primitive and (2) indicates cells within the sampling grid that are intersected by the edge and are only partially covered by the first graphics primitive. The coverage information for the edge is stored in a cache. When a second graphics primitive is rasterized that shares the edge with the first graphics primitive, the coverage information is read from the cache instead of being recomputed.