Adaptive Texture Filtering via Dynamic Mip-Map Sampling

Resolve Bottlenecks,
Find Innovative Solutions
Generate Solutions

Solution Overview

Problem

Current graphics processing units (GPUs) face inefficiencies in texture filtering, particularly in selecting the optimal number of samples from mip-map layers, leading to increased processing workloads and power consumption.

Innovation Solution

The GPU selectively determines a first quantity of samples from a lower mip-map layer and a second quantity from a higher mip-map layer based on a relative distance, with the second quantity being less than the first, thereby reducing the total number of samples and filter taps required for texture filtering.

Engineering Contradictions & Design Principles

VSEngineering Contradiction Analysis

1Measurement precision

If a fixed number of samples is used from both mip-map layers, then filtering quality is maintained, but processing workload and power consumption increase

Engineering Contradiction:
Improvefiltering qualityVSAvoidprocessing workload
Core Design Contradiction:
Measurement precisionVSProductivity

Solution Approach 1:

The patent applies dynamics by making the number of samples adaptive rather than fixed. The system dynamically adjusts the number of samples taken from each mip-map layer based on the relative distance calculation, allowing the sampling strategy to change according to rendering conditions while maintaining filtering quality and reducing processing workload when appropriate

Inventive Principle:
Principle #15Dynamics

Solution Approach 2:

The patent changes the parameter of sample quantity from a static value to a dynamic value determined by relative distance. By calculating relative distance between texture sizes and using this to determine sample counts, the system optimizes the balance between filtering quality and processing efficiency through parameter adaptation

Inventive Principle:
Principle #35Parameter changes

2Measurement precision

If more samples are taken from both mip-map layers, then filtering quality improves, but power consumption increases

Engineering Contradiction:
Improvefiltering qualityVSAvoidpower consumption
Core Design Contradiction:
Measurement precisionVSUse of energy by moving object

Solution Approach 1:

The system changes the sample quantity parameter based on relative distance calculations, reducing power consumption by taking fewer samples when high filtering quality is not critically needed while maintaining quality when the relative distance indicates it is necessary

Inventive Principle:
Principle #35Parameter changes

3Stability of the object's composition

If equal numbers of samples are taken from both mip-map layers, then filtering consistency is maintained, but processing efficiency decreases

Engineering Contradiction:
Improvefiltering consistencyVSAvoidprocessing efficiency
Core Design Contradiction:
Stability of the object's compositionVSProductivity

Solution Approach 1:

The patent applies local quality by allowing different numbers of samples to be taken from different mip-map layers based on their respective contributions. Instead of uniform sampling, the system locally adapts the sample count for each layer according to the relative distance metric, improving processing efficiency while maintaining filtering consistency through targeted sampling

Inventive Principle:
Principle #3Local quality

Data Source

PatentUS11257277B2Methods and apparatus to facilitate adaptive texture filtering
Publication Date: 2022.02.22 QUALCOMM INC
  • US11257277B2 patent drawing
  • US11257277B2 patent drawing
  • US11257277B2 patent drawing

AI summary

The present disclosure relates to methods and apparatus for graphics processing. In some aspects, the apparatus selects a first mip-map layer with a first texture size and a second mip-map layer with a second texture size based on a third texture size of an image. The apparatus also determines a relative distance associated with the texture sizes. Additionally, the apparatus determines a first quantity of samples to select from the first mip-map layer, and determines a second quantity of samples to select from the second mip-map layer, the second quantity of samples being less than the first quantity of samples, and a second quantity of filter taps being less than a first quantity of filter taps. Also, the apparatus generates the image at the third texture size through filtering based on the first quantity of samples and the second quantity of samples.