Volumetric Filtering Pipeline With Zero-Coefficient Bypass

Resolve Bottlenecks,
Find Innovative Solutions
Generate Solutions

Solution Overview

Problem

Existing texture filtering methods in hardware acceleration for 3D computer graphics are inefficient when multiple filtering modes are combined, leading to complex and computationally expensive operations.

Innovation Solution

A filtering unit with a computation pipeline that bypasses computations when a volumetric filter coefficient is zero, using a register to store input values and allowing the pipeline to perform other interpolations, combined with sequencers and an arbiter to manage access to computation pipelines, enabling efficient handling of multiple filtering modes.

Engineering Contradictions & Design Principles

VSEngineering Contradiction Analysis

1Speed

If hardware acceleration is used to perform texture filtering, then filtering speed is improved, but device complexity increases

Engineering Contradiction:
Improvefiltering speedVSAvoidhardware complexity
Core Design Contradiction:
SpeedVSDevice complexity

Solution Approach 1:

The filtering unit is divided into separate computation pipelines for different filtering modes (volumetric, anisotropic, trilinear). Each pipeline is specialized and can be independently controlled, allowing the system to activate only the necessary pipeline for the current filtering operation, thus reducing overall complexity while maintaining high speed

Inventive Principle:
Principle #1Segmentation

Solution Approach 2:

The filtering unit dynamically switches between different computation pipelines based on the filtering mode required. Control logic selectively enables or disables specific pipelines, allowing the hardware to adapt its complexity to the actual filtering needs, avoiding the permanent overhead of having all pipelines fully active

Inventive Principle:
Principle #15Dynamics

2Adaptability or versatility

If combinations of volumetric, anisotropic, and trilinear filtering are implemented, then filtering versatility is improved, but device complexity increases

Engineering Contradiction:
Improvefiltering mode combinationsVSAvoidhardware complexity
Core Design Contradiction:
Adaptability or versatilityVSDevice complexity

Solution Approach 1:

Different filtering modes are implemented as separate computation pipelines rather than a single complex unified pipeline. This segmentation allows each filtering mode to be independently optimized and controlled, making it easier to support multiple mode combinations without exponentially increasing complexity

Inventive Principle:
Principle #1Segmentation

Solution Approach 2:

The filtering unit uses a universal control structure that can direct input data to any combination of specialized pipelines. The same control logic and data paths can serve multiple filtering modes, allowing versatile filtering combinations while reusing hardware resources across different functions

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

3Reliability

If volumetric filtering with zero coefficient is processed through the computation pipeline, then processing consistency is maintained, but productivity decreases

Engineering Contradiction:
Improveprocessing consistencyVSAvoidthroughput
Core Design Contradiction:
ReliabilityVSProductivity

Solution Approach 1:

When a volumetric filtering operation has a zero coefficient, the system performs only the necessary partial computation (storing the input value directly) rather than completing the full interpolation pipeline. This partial action maintains consistency with the overall filtering process while avoiding unnecessary computational steps that would reduce throughput

Inventive Principle:
Principle #16Partial or excessive action

Solution Approach 2:

The system detects when the volumetric filter coefficient is zero and changes the processing parameter to bypass the interpolation computation. This parameter change allows the same hardware to handle both full interpolation operations and zero-coefficient cases efficiently, maintaining correctness while improving throughput for the latter case

Inventive Principle:
Principle #35Parameter changes

Data Source

PatentEP4654139A1Volumetric point filtering unit and method
Publication Date: 2025.11.26 IMAGINATION TECH LTD
  • EP4654139A1 patent drawingFigure 1
  • EP4654139A1 patent drawingFigure 2
  • EP4654139A1 patent drawingFigure 3~4

AI summary

A filtering unit implemented in hardware logic configured to apply one or more filtering methods to a plurality of input values to determine output values. The filtering unit comprises a plurality of inputs configured to receive one or more input values, one or more signal values, and one or more filter coefficients. The one or more signal values define a filtering mode and wherein each of the one or more filter coefficients corresponds to one of the one or more filtering methods. The filtering unit comprises a computation pipeline configured to receive two input values and a corresponding filter coefficient, and perform an interpolation using at least the two input values and the corresponding filter coefficient.. The filtering unit comprises a plurality of registers configured to store intermediate output values generated by the computation pipeline. The filtering unit is configured to receive one or more signal values, a volumetric filter coefficient, and an input value, and determine i) that a filtering mode defined by the received one or more signal values comprises volumetric filtering and at least one other filtering method and ii) that the volumetric filter coefficient is equal to zero. In response to determining i) and ii) the filtering unit is configured to store the received input value in a register of the plurality of registers.