Z-Order Multidimensional Parsing for Skipping Invalid Coordinates

Resolve Bottlenecks,
Find Innovative Solutions
Generate Solutions

Solution Overview

Problem

The use of Z-ordering functions to map multidimensional scene data into one-dimensional sequences can result in inefficiencies when the valid region of the grid is not an exact power-of-2, leading to the generation of invalid coordinates and wasted cycles due to inefficient scanning.

Innovation Solution

A method is provided to determine a sequence of valid positions by calculating an overflow value that flips the least significant bit from '1' to '0' in the binary representation of the one-dimensional data, allowing efficient skipping over invalid coordinates using bitwise operations.

Engineering Contradictions & Design Principles

VSEngineering Contradiction Analysis

1Reliability

If Z-ordering function is used to map multidimensional data into one-dimensional sequence, then spatial locality of reference is improved and cache misses are reduced, but invalid coordinates are generated when valid region is not an exact power-of-2

Engineering Contradiction:
Improvespatial locality of referenceVSAvoidinvalid coordinates
Core Design Contradiction:
ReliabilityVSObject-generated harmful factors

Solution Approach 1:

The patent extracts and removes the harmful invalid coordinates from the Z-order sequence by implementing a validation mechanism that identifies and skips over coordinates that fall outside the valid region boundaries, thereby maintaining the spatial locality benefits while eliminating the harmful effect of invalid coordinates

Inventive Principle:
Principle #2Taking out (Extraction)

Solution Approach 2:

The patent changes the parameter of coordinate validation by introducing boundary checking mechanisms that modify the standard Z-order generation process. When a coordinate exceeds the valid region boundaries, the system adjusts the sequence generation to skip invalid coordinates, effectively changing how the Z-order function handles non-power-of-2 dimensions

Inventive Principle:
Principle #35Parameter changes

2Productivity

If Z-ordering function is used to map multidimensional data, then cache line efficiency is improved, but processing time increases due to wasted cycles from invalid grid coordinates

Engineering Contradiction:
Improvecache line efficiencyVSAvoidwasted cycles
Core Design Contradiction:
ProductivityVSLoss of time

Solution Approach 1:

The patent implements a skipping mechanism that rapidly advances through the Z-order sequence by calculating jump values that skip over blocks of invalid coordinates. This allows the system to rush through the invalid regions without processing them individually, thereby recovering processing time while maintaining cache line efficiency benefits

Inventive Principle:
Principle #21Skipping (Rushing through)

Solution Approach 2:

The patent performs preliminary validation of coordinates before processing them, using boundary checking to pre-identify invalid coordinates and skip them in advance. This preliminary action prevents wasted cycles by avoiding the generation and processing of invalid coordinates in the first place

Inventive Principle:
Principle #10Preliminary action

3Ease of manufacture

If strided order scanning is used, then implementation simplicity is maintained, but cache misses increase and spatial locality is reduced

Engineering Contradiction:
Improveimplementation simplicityVSAvoidspatial locality of reference
Core Design Contradiction:
Ease of manufactureVSReliability

Solution Approach 1:

The patent merges the simplicity of strided order with the benefits of Z-ordering by combining boundary checking logic with the sequence generation process. This integration allows the system to maintain ease of implementation while achieving improved spatial locality through the hybrid approach of validating and skipping invalid coordinates

Inventive Principle:
Principle #5Merging (Combining)

Data Source

PatentEP4664408A1Methods and systems for parsing multidimensional data
Publication Date: 2025.12.17 IMAGINATION TECH LTD
  • EP4664408A1 patent drawingFigure 1
  • EP4664408A1 patent drawingFigure 2A
  • EP4664408A1 patent drawingFigure 2B

AI summary

There is provided a computer-implemented method of parsing multidimensional data to identify work packets to be processed, wherein the multidimensional data is mapped into one-dimensional data using a Z-ordering function. The method comprises: obtaining dimensions of the multidimensional data and obtaining information identifying a valid region of the multidimensional data, where the valid region identifies work packets. The method comprises determining at least part of a sequence, within the one-dimensional data, corresponding to positions within the multidimensional data lying in the valid region, by: determining, in dependence on the Z-ordering function, whether a position defined by a first value of the one-dimensional data lies within the valid region; and in response to determining that the position defined by the first value does not lie within the valid region: calculating an overflow value which, when added to the first value, causes a least significant bit that is a 1 in a binary representation of the first value to flip to a 0. The method further comprises summing the first value and the overflow value to obtain a trial value and determining whether the trial value lies within the valid region to thereby determine whether the trial value forms part of the sequence, wherein the sequence identifies work packets to be processed.