Tiled Deferred Shading for GPU Photon Mapping Scalability

Resolve Bottlenecks,
Find Innovative Solutions
Generate Solutions

Solution Overview

Problem

Current graphics processing systems face challenges in maintaining quality while efficiently scaling photon mapping for real-time rendering on modern GPUs, particularly in the process of photon shading.

Innovation Solution

The implementation of tiled deferred shading, which identifies and processes photons in screen-space tiles, selecting a subset of photons for shading computation, and utilizing parallel processing units to efficiently compute shading across these tiles.

Engineering Contradictions & Design Principles

VSEngineering Contradiction Analysis

1Manufacturing precision

If all photons are processed for shading computation, then shading quality is improved, but processing time and computational load increase

Engineering Contradiction:
Improveshading qualityVSAvoidprocessing time
Core Design Contradiction:
Manufacturing precisionVSLoss of time

Solution Approach 1:

The patent divides the screen space into multiple tiles and processes photons in a deferred manner, organizing photons by their spatial distribution across tiles. This segmentation allows the system to process photons in manageable groups rather than all at once, reducing immediate computational load while maintaining overall shading quality through systematic processing of all photons across frames.

Inventive Principle:
Principle #1Segmentation

Solution Approach 2:

The patent implements a hybrid approach where a subset of photons is processed stochastically for efficiency, while ensuring that sufficient photon sampling occurs to maintain quality. The system dynamically adjusts the number of photons processed based on scene complexity and performance requirements, processing more photons when quality is prioritized and fewer when real-time performance is critical.

Inventive Principle:
Principle #16Partial or excessive action

2Manufacturing precision

If photon mapping is implemented for real-time rendering, then lighting quality is improved, but scalability and performance on modern GPUs deteriorate

Engineering Contradiction:
Improvelighting qualityVSAvoidscalability and performance
Core Design Contradiction:
Manufacturing precisionVSProductivity

Solution Approach 1:

The patent segments the photon mapping process into distinct phases: photon tracing, photon storage in tiles, and deferred shading computation. This segmentation allows each phase to be optimized independently for GPU execution, with photon tracing handled by ray tracing hardware and shading computation distributed across GPU cores processing different tiles in parallel, thereby improving scalability.

Inventive Principle:
Principle #1Segmentation

Solution Approach 2:

The patent transitions from traditional per-pixel photon gathering to a tile-based approach, adding a spatial dimension to the organization of photon data. By grouping photons and pixels into tiles, the system enables efficient parallel processing across multiple GPU streaming multiprocessors, improving throughput and scalability while maintaining lighting quality through systematic coverage of all screen space tiles.

Inventive Principle:
Principle #17Another dimension (Dimensionality change)

3Speed

If photon shading is optimized for speed, then processing time is reduced, but shading quality deteriorates

Engineering Contradiction:
Improveprocessing speedVSAvoidshading quality
Core Design Contradiction:
SpeedVSManufacturing precision

Solution Approach 1:

The patent employs stochastic sampling to process a representative subset of photons for each pixel, rather than exhaustively processing all photons. This partial action approach reduces processing time significantly while maintaining statistically accurate shading results. The system adjusts the sampling rate to balance speed and quality, processing more photons in complex scenes and fewer in simpler scenes.

Inventive Principle:
Principle #16Partial or excessive action

Solution Approach 2:

The patent performs preliminary organization of photons into spatial tiles and pre-computes photon properties during the tracing phase. This preliminary action prepares photon data in an optimized format that enables rapid shading computation in the deferred phase, reducing the computational burden during actual rendering while maintaining accuracy through complete photon utilization when needed.

Inventive Principle:
Principle #10Preliminary action

Data Source

PatentUS9305324B2System, method, and computer program product for tiled deferred shading
Publication Date: 2016.04.05 NVIDIA CORP
  • US9305324B2 patent drawing
  • US9305324B2 patent drawing
  • US9305324B2 patent drawing

AI summary

A system, method, and computer program product are provided for tiled deferred shading. In operation, a plurality of photons associated with at least one scene are identified. Further, a plurality of screen-space tiles associated with the at least one scene are identified. Additionally, each of the plurality of screen-space tiles capable of being affected by a projection of an effect sphere for each of the plurality of photons are identified. Furthermore, at least a subset of photons associated with each of the screen-space tiles from which to compute shading are selected. Moreover, shading for the at least one scene is computed utilizing the selected at least a subset of photons.