Tile Binning for Path Rendering via Shape-Based Edge Grouping

Resolve Bottlenecks,
Find Innovative Solutions
Generate Solutions

Solution Overview

Problem

Enhancing the acceleration performance of graphic processing units (GPUs) during path rendering is complex due to the dependency between edges in traditional tile binning methods, which limits parallel processing and increases rendering time and power consumption.

Innovation Solution

A method of tile binning that classifies paths into groups based on their shapes, allowing for parallel processing by assigning identification values to tiles containing path start or end points, and calculating winding numbers to enable efficient shading of pixels based on path directions, thereby reducing dependency between edges and enabling simultaneous processing.

Engineering Contradictions & Design Principles

VSEngineering Contradiction Analysis

1Productivity

If traditional tile binning methods are used for path rendering, then edges are processed sequentially, but this creates dependency between edges that limits parallel processing and increases rendering time

Engineering Contradiction:
Improverendering speedVSAvoidprocessing complexity
Core Design Contradiction:
ProductivityVSDevice complexity

Solution Approach 1:

The patent segments the path rendering process by classifying edges into different groups based on their spatial relationships. Specifically, edges are divided into groups where each group contains edges that do not have spatial overlap or dependency relationships. This segmentation allows independent parallel processing of multiple edge groups simultaneously, breaking the sequential dependency chain while maintaining rendering accuracy.

Inventive Principle:
Principle #1Segmentation

Solution Approach 2:

The patent applies preliminary action by performing edge classification and grouping before the actual rendering process. The GPU determines spatial relationships between edges in advance, categorizes them into independent groups, and prepares parallel processing schedules before rendering begins. This preliminary organization enables subsequent parallel execution without runtime dependencies.

Inventive Principle:
Principle #10Preliminary action

2Measurement precision

If edges are processed sequentially to maintain accuracy, then rendering precision is preserved, but rendering time and power consumption increase

Engineering Contradiction:
Improverendering accuracyVSAvoidrendering time
Core Design Contradiction:
Measurement precisionVSLoss of time

Solution Approach 1:

The patent segments edges into multiple independent groups based on spatial classification. Each group is guaranteed to have no dependency relationships between its edges, allowing parallel processing. The segmentation ensures that within each group, edges can be rendered simultaneously without affecting accuracy, while the grouping strategy maintains overall rendering precision by preserving spatial relationships between groups.

Inventive Principle:
Principle #1Segmentation

Solution Approach 2:

The patent changes the processing parameter from sequential to parallel execution by modifying the dependency structure. Through spatial classification and grouping, the patent transforms the original sequential processing model into a parallel processing model where multiple edge groups are rendered simultaneously, reducing rendering time while maintaining accuracy through careful group construction.

Inventive Principle:
Principle #35Parameter changes

3Productivity

If parallel processing is implemented without classification, then rendering speed increases, but errors occur due to unhandled dependencies between edges

Engineering Contradiction:
Improveprocessing throughputVSAvoidrendering correctness
Core Design Contradiction:
ProductivityVSReliability

Solution Approach 1:

The patent segments edges into dependency-free groups through spatial classification. Each group is carefully constructed to contain only edges that can be safely processed in parallel without interfering with each other. This segmentation guarantees both high parallel throughput and rendering correctness by eliminating problematic dependencies within groups while maintaining necessary relationships between groups.

Inventive Principle:
Principle #1Segmentation

Solution Approach 2:

The patent introduces an intermediary classification mechanism that acts as a mediator between sequential accuracy requirements and parallel speed requirements. The spatial classification system serves as an intermediary layer that organizes edges into groups, enabling parallel processing while preserving correctness. This intermediary structure translates the complex dependency relationships into manageable parallel tasks.

Inventive Principle:
Principle #24Intermediary (Mediator)

Data Source

PatentEP3082108B1Method and apparatus for performing tile binning for path rendering
Publication Date: 2021.06.09 SAMSUNG ELECTRONICS CO LTD
  • EP3082108B1 patent drawingFigure 1
  • EP3082108B1 patent drawingFigure 2A
  • EP3082108B1 patent drawingFigure 2B

AI summary

A method and apparatus to perform tile binning for tile-based rendering include obtaining information about paths defining an object to be rendered. The method classifies one of the paths into a first group and another of the paths into a second group based on shapes of the paths, and performs the tile binning based on the classification.