GPU Texture Memory Allocation for Image Processing

Resolve Bottlenecks,
Find Innovative Solutions
Generate Solutions

Solution Overview

Problem

Conventional image display processes in mobile devices result in memory occupation waste and frame freezing due to the need for separate memory allocation for CPU and GPU, as well as time-consuming data conversion and uploading processes.

Innovation Solution

An image processing method that allocates memory directly to the GPU for storing texture data, eliminating the need for CPU memory allocation and performing parallel data conversion to generate texture data, thereby reducing memory waste and data uploading time.

Engineering Contradictions & Design Principles

VSEngineering Contradiction Analysis

1Reliability

If separate memory partitions are allocated for CPU and GPU, then each processor can access its own memory, but memory occupation waste occurs

Engineering Contradiction:
Improvememory access reliabilityVSAvoidmemory occupation
Core Design Contradiction:
ReliabilityVSQuantity of substance

Solution Approach 1:

The patent merges the memory allocation process by allowing the GPU to directly access the memory partition originally allocated for CPU bitmap data. This eliminates the need for separate memory partitions for CPU and GPU, reducing memory waste while maintaining access reliability for both processors.

Inventive Principle:
Principle #5Merging (Combining)

Solution Approach 2:

The memory partition allocated for CPU bitmap data is made universally accessible to both CPU and GPU. This multi-functional memory allocation allows the same memory space to serve both processors, eliminating redundant memory occupation while ensuring reliable access for both CPU decoding and GPU rendering operations.

Inventive Principle:
Principle #6Universality (Multi-functionality)

2Reliability

If bitmap data is converted to texture data and uploaded to GPU memory, then the GPU can perform drawing operations, but the process consumes significant time causing frame freezing

Engineering Contradiction:
Improvedrawing operation reliabilityVSAvoiddata conversion and uploading time
Core Design Contradiction:
ReliabilityVSLoss of time

Solution Approach 1:

The patent performs preliminary action by pre-converting bitmap data to texture data format in advance, before GPU rendering is needed. This allows the GPU to directly use the pre-converted texture data without waiting for conversion during the critical rendering path, significantly reducing frame freezing while ensuring reliable drawing operations.

Inventive Principle:
Principle #10Preliminary action

Solution Approach 2:

The patent maintains continuity of useful action by overlapping the bitmap-to-texture conversion process with other non-critical operations. The conversion continues in the background while other tasks execute, ensuring the GPU always has available texture data without interruption, thus eliminating frame freezing while maintaining continuous productive work.

Inventive Principle:
Principle #20Continuity of useful action

Data Source

PatentUS12175558B2Image processing method and apparatus
Publication Date: 2024.12.24 HUAWEI TECH CO LTD
  • US12175558B2 patent drawing
  • US12175558B2 patent drawing
  • US12175558B2 patent drawing

AI summary

A method includes: displaying an interaction interface of an application, and when a preset operation performed by a user on the interaction interface is detected, decoding to-be-displayed image data of the application into bitmap data, and encapsulating the bitmap data into texture data; storing the texture data in a memory partition accessible to a graphics processing unit GPU; triggering the GPU to read the texture data and perform drawing processing to obtain rendered data; and triggering a display to display an image based on the rendered data.