Graphics Processing Internal Vector Dynamic Memory Buffering

Resolve Bottlenecks,
Find Innovative Solutions
Generate Solutions

Solution Overview

Problem

Current graphics processing units (GPUs) face inefficiencies in capturing and processing image content due to repeated loading of overlapped pixel data, which increases system overhead and memory usage.

Innovation Solution

Implementing an internal vector dynamic memory with separate buffers for non-overlapped and overlapped pixel data, where non-overlapped data is not reused and overlapped data is reused across slice processing steps, reducing the need for repeated data retrieval from system memory.

Engineering Contradictions & Design Principles

VSEngineering Contradiction Analysis

1Productivity

If pixel data is loaded from system memory for each slice processing step, then processing can be performed, but system overhead increases and memory usage increases due to repeated loading of overlapped data

Engineering Contradiction:
Improvegraphics processing efficiencyVSAvoidsystem overhead
Core Design Contradiction:
ProductivityVSLoss of energy

Solution Approach 1:

The pixel data is segmented into non-overlapped pixel data and overlapped pixel data. Non-overlapped data is processed and discarded after use, while overlapped data is retained for reuse in subsequent slice processing steps, eliminating redundant memory loads and reducing system overhead

Inventive Principle:
Principle #1Segmentation

Solution Approach 2:

Overlapped pixel data is pre-loaded and stored in the internal vector dynamic memory before being needed in subsequent processing steps. This preliminary action allows the data to be readily available when required, avoiding repeated loading from system memory and reducing processing delays

Inventive Principle:
Principle #10Preliminary action

2Productivity

If overlapped pixel data is reloaded from system memory for each slice, then processing continues, but memory usage increases due to repeated data retrieval

Engineering Contradiction:
Improveprocessing continuityVSAvoidmemory usage
Core Design Contradiction:
ProductivityVSQuantity of substance

Solution Approach 1:

An internal vector dynamic memory structure is introduced as an intermediary between system memory and the processing unit. This intermediary buffers overlapped pixel data, allowing it to be reused across multiple slice processing steps without requiring repeated retrieval from system memory, thus reducing overall memory usage

Inventive Principle:
Principle #24Intermediary (Mediator)

Solution Approach 2:

Overlapped pixel data is copied from system memory into the internal vector dynamic memory buffer once, and then this copy is reused multiple times across different slice processing steps. This single copy approach eliminates the need for repeated full data retrieval, reducing memory bandwidth consumption and usage

Inventive Principle:
Principle #26Copying

Data Source

PatentEP3474224B1Graphics processing method and device
Publication Date: 2020.08.26 VIA ALLIANCE SEMICON CO LTD
  • EP3474224B1 patent drawingFigure 1
  • EP3474224B1 patent drawingFigure 2
  • EP3474224B1 patent drawingFigure 3

AI summary

A technique for graphics processing, which completes graphics processing of an image loaded from a system memory by performing a series of slice processing steps. A device for graphics processing has an internal vector dynamic memory for buffering slices of pixel data loaded from the system memory. The internal vector dynamic memory has a first buffer for buffering non-overlapped pixel data, which is not reused in a next slice processing step and a second buffer for buffering overlapped pixel data, which is reused in the next slice processing step.