Programmable Sample Pattern Tables for Graphics Compression

Resolve Bottlenecks,
Find Innovative SolutionsGenerate Solutions

Solution Overview

Problem

Existing graphics processing technologies face inefficiencies due to fixed sample patterns stored in SRAM, limiting the implementation of anti-aliasing algorithms and causing issues with data reconstruction when sample patterns change, leading to undesirable visual artifacts.

Innovation Solution

Implementing programmable sample pattern tables that allow for varying sample locations across pixels, enabling the storage and reconstruction of graphics data in a sample-specific compressed format, ensuring correct data reconstruction and reduced visual artifacts.

Engineering Contradictions & Design Principles

VSEngineering Contradiction Analysis

1Productivity

If fixed sample patterns are stored in SRAM, then processing efficiency is improved, but adaptability of anti-aliasing algorithms is limited

Engineering Contradiction:

Improveprocessing efficiencyVSAvoidadaptability of anti-aliasing algorithms

Core Design Contradiction:

ProductivityVSAdaptability or versatility

Solution Approach 1:

The patent implements dynamic sample patterns by storing multiple sample patterns in a sample pattern buffer and selectively loading them into the sample pattern table based on rendering requirements. This allows the system to transition from fixed to variable sample patterns, enabling different anti-aliasing algorithms (MSAA, SSAA, FXAA) to be implemented without sacrificing processing efficiency.

Inventive Principle:

Principle #15Dynamics

2Adaptability or versatility

If sample patterns are changed, then adaptability of anti-aliasing algorithms is improved, but data reconstruction accuracy deteriorates

Engineering Contradiction:

Improveadaptability of anti-aliasing algorithmsVSAvoiddata reconstruction accuracy

Core Design Contradiction:

Adaptability or versatilityVSManufacturing precision

Solution Approach 1:

The patent performs preliminary decomposition of compressed surface data using the original sample pattern before changing to a new sample pattern. The decompressed data is then re-compressed using the new sample pattern, ensuring data integrity is maintained during sample pattern transitions. This preliminary action prevents reconstruction errors that would occur if sample patterns were changed without proper data transformation.

Inventive Principle:

Principle #10Preliminary action

Solution Approach 2:

The patent introduces an intermediary decompression/compression process that acts as a mediator between different sample patterns. When changing sample patterns, the system uses the intermediary process to convert data from one sample pattern format to another, ensuring accurate reconstruction without direct incompatible access.

Inventive Principle:

Principle #24Intermediary (Mediator)

3Adaptability or versatility

If multiple sample patterns are supported, then adaptability of anti-aliasing algorithms is improved, but device complexity increases

Engineering Contradiction:

Improveadaptability of anti-aliasing algorithmsVSAvoiddevice complexity

Core Design Contradiction:

Adaptability or versatilityVSDevice complexity

Solution Approach 1:

The patent implements a universal sample pattern table and buffer structure that can store and manage multiple sample patterns (MSAA, SSAA, FXAA) using the same hardware resources. The sample pattern buffer acts as a multi-functional storage unit that can hold different pattern types, and the loading mechanism universally handles pattern switching without requiring separate dedicated hardware for each algorithm.

Inventive Principle:

Principle #6Universality (Multi-functionality)

4Loss of energy

If sample-specific compressed format is used, then memory bandwidth is reduced, but reconstruction complexity increases

Engineering Contradiction:

Improvememory bandwidth usageVSAvoidreconstruction complexity

Core Design Contradiction:

Loss of energyVSDevice complexity

Solution Approach 1:

The patent implements sample-specific compression where each sample location has its own compression parameters stored in the sample pattern table. During reconstruction, only the specific sample parameters needed for the current rendering operation are loaded from the buffer into the active table, allowing local optimization without requiring full decompression of all sample data.

Inventive Principle:

Principle #3Local quality

Data Source

PatentUS9230363B2System, method, and computer program product for using compression with programmable sample locations

Publication Date: 2016.01.05 NVIDIA CORP

AI summary

A system, method, and computer program product enable compression with programmable sample locations, where the compression is a function of the programmable sample locations. The method includes the steps of storing a first value specifying a programmed sample location within a pixel in a first sample pattern table that is associated with a first display surface and storing, in a memory, geometric surface parameters corresponding to a first attribute at the programmed sample location within a first pixel of the first display surface. A second value specifying the programmed sample location within the pixel in a second sample pattern table that is associated with a second display surface is also stored and the first attribute is reconstructed based on the geometric surface parameters and the first value.