Graphics Texture Processing with Neural Decompression Circuits

Resolve Bottlenecks,
Find Innovative SolutionsGenerate Solutions

Solution Overview

Problem

Existing graphics processors are inefficient in supporting neural network based texture compression/decompression schemes, leading to suboptimal performance and increased silicon area and energy consumption.

Innovation Solution

Integrate a neural network processing circuit within the graphics processor to perform neural network based texture decompression, facilitating efficient decompression of graphics texture data directly from a compressed format to an uncompressed format for use by the graphics processor.

Engineering Contradictions & Design Principles

VSEngineering Contradiction Analysis

1Ease of manufacture

If traditional texture decompression hardware circuits are used, then texture data can be decompressed, but silicon area and energy consumption increase

Engineering Contradiction:

Improvetexture decompression capabilityVSAvoidsilicon area

Core Design Contradiction:

Ease of manufactureVSArea of stationary object

Solution Approach 1:

The neural network processing circuit is designed to perform multiple functions: it can decompress texture data in addition to its primary neural network processing tasks. This multi-functionality allows the same hardware circuit to serve both AI processing and texture decompression purposes, thereby reducing the need for dedicated texture decompression hardware and minimizing silicon area consumption.

Inventive Principle:

Principle #6Universality (Multi-functionality)

2Ease of manufacture

If traditional texture decompression hardware circuits are used, then texture data can be decompressed, but energy consumption increases

Engineering Contradiction:

Improvetexture decompression capabilityVSAvoidenergy consumption

Core Design Contradiction:

Ease of manufactureVSUse of energy by stationary object

Solution Approach 1:

The neural network processing circuit performs texture decompression as one of its multiple functions, eliminating the need for separate dedicated decompression hardware. By reusing the same circuit for both neural network operations and texture decompression, the system reduces overall energy consumption while maintaining full texture decompression capability.

Inventive Principle:

Principle #6Universality (Multi-functionality)

3Adaptability or versatility

If neural network based texture decompression is implemented, then flexibility and configurability are enhanced, but device complexity increases

Engineering Contradiction:

Improveflexibility and configurabilityVSAvoiddevice complexity

Core Design Contradiction:

Adaptability or versatilityVSDevice complexity

Solution Approach 1:

The neural network processing circuit is designed to autonomously handle texture decompression operations without requiring complex external control mechanisms. The circuit self-configures and self-manages the decompression process, reducing the need for additional complex control logic and interface circuits, thereby minimizing device complexity while maintaining high flexibility and configurability.

Inventive Principle:

Principle #25Self-service

Data Source

PatentUS20250278885A1Graphics texture processing

Publication Date: 2025.09.04 ARM LTD

AI summary

Disclosed are graphics processor arrangements in which the graphics processor is configured to support neural network based texture compression schemes. In particular, a graphics processing system is provided including a neural network processing circuit that is operable and configured to execute one or more neural networks to process graphics texture data that has been compressed using a neural network based texture compression scheme into a suitable (decompressed) format for use by the graphics processor. Also disclosed are compiler operations for generating shader programs to control such texturing operations.