Counting Elements in Data Items for Histogram Processing

Resolve Bottlenecks,
Find Innovative Solutions
Generate Solutions

Solution Overview

Problem

Data processing apparatuses face inefficiencies in counting elements within large input data items due to increased complexity and memory access requirements, particularly in histogram processing, where the number of permutations of comparisons becomes impractical.

Innovation Solution

The apparatus employs instruction decoder circuitry and data processing circuitry to perform count operations on subsets of elements within input data items, using control signals to determine the number of elements matching a reference value, and includes register scatter storage and gather index generation capabilities to optimize memory access and processing efficiency.

Engineering Contradictions & Design Principles

VSEngineering Contradiction Analysis

1Measurement precision

If conventional counting methods are used for large input data items, then counting accuracy is maintained, but processing complexity and memory access requirements increase significantly

Engineering Contradiction:
Improvecounting accuracyVSAvoidprocessing complexity
Core Design Contradiction:
Measurement precisionVSDevice complexity

Solution Approach 1:

The patent divides the input data item into multiple subsets of elements, allowing the counting operation to process smaller segments independently. This segmentation reduces the complexity of comparing each element against all possible reference values while maintaining accurate counting through aggregation of segment results.

Inventive Principle:
Principle #1Segmentation

Solution Approach 2:

The patent introduces a new dimension by adding a count value dimension to the data structure. Instead of only storing element values, each element now includes an associated count value that can be incremented, transforming the counting problem from a comparison-heavy operation to an accumulation operation that reduces processing complexity.

Inventive Principle:
Principle #17Another dimension (Dimensionality change)

2Measurement precision

If conventional counting methods are used for large input data items, then complete counting is achieved, but the number of memory accesses increases

Engineering Contradiction:
Improvecounting completenessVSAvoidmemory access time
Core Design Contradiction:
Measurement precisionVSLoss of time

Solution Approach 1:

The patent performs preliminary actions by pre-initializing count values to zero before processing elements and pre-allocating data structures to hold count values. This preliminary setup eliminates the need for repeated memory allocations and initializations during the counting process, reducing overall memory access time.

Inventive Principle:
Principle #10Preliminary action

Solution Approach 2:

The patent merges the counting operation with the data processing operation by maintaining count values within the same data structure as the element values. This merging eliminates separate memory accesses for storing and retrieving count information, reducing the total number of memory accesses required.

Inventive Principle:
Principle #5Merging (Combining)

3Loss of information

If histogram processing is performed on large datasets, then comprehensive statistical analysis is achieved, but processing burden increases

Engineering Contradiction:
Improvestatistical information completenessVSAvoidprocessing efficiency
Core Design Contradiction:
Loss of informationVSProductivity

Solution Approach 1:

The patent segments the histogram processing into independent counting operations for each element subset. This allows parallel processing of different segments and reduces the sequential processing burden, maintaining comprehensive statistical information while improving overall processing efficiency through divided computation.

Inventive Principle:
Principle #1Segmentation

Solution Approach 2:

The patent changes the processing parameter from comparing each element against all histogram bins to incrementing pre-identified bin counters. This parameter change transforms an O(n*m) operation into an O(n) operation where n is the number of elements and m is the number of bins, significantly improving productivity while maintaining statistical completeness.

Inventive Principle:
Principle #35Parameter changes

Data Source

PatentEP3662365B1Counting elements in data items in a data processing apparatus
Publication Date: 2021.09.22 ARM LTD
  • EP3662365B1 patent drawingFigure 1
  • EP3662365B1 patent drawingFigure 2
  • EP3662365B1 patent drawingFigure 3

AI summary

An apparatus and method of operating the apparatus are provided for performing a count operation. Instruction decoder circuitry is responsive to a count instruction specifying an input data item to generate control signals to control the data processing circuitry to perform a count operation. The count operation determines a count value indicative of a number of input elements of a subset of elements in the specified input data item which have a value which matches a reference value in a reference element in a reference data item. A plurality of count operations may be performed to determine a count data item corresponding to the input data item. A register scatter storage instruction, a gather index generation instruction, and respective apparatuses responsive to them, as well as simulator implementations are also provided.