GPU Sprite Processing for Structured Data Queries

Resolve Bottlenecks,
Find Innovative SolutionsGenerate Solutions

Solution Overview

Problem

The processing of multi-dimensional data in modern computing systems is computationally intensive, time-consuming, and energy-consuming, particularly for tasks involving structured data, which existing CPU-based systems struggle to handle efficiently.

Innovation Solution

Converting structured data into image data format suitable for GPU processing, utilizing GPUs' parallel processing capabilities to perform data analytics, thereby reducing computational complexity and resource requirements.

Engineering Contradictions & Design Principles

VSEngineering Contradiction Analysis

1Productivity

If CPU-based systems are used to process multi-dimensional structured data, then data processing can be performed, but processing time and energy consumption increase significantly

Engineering Contradiction:

Improvedata processing speedVSAvoidprocessing time

Core Design Contradiction:

ProductivityVSLoss of time

Solution Approach 1:

The patent replaces CPU-based sequential processing with GPU-based parallel processing. The structured data is transformed into image data format, allowing GPUs to leverage their parallel architecture for simultaneous processing of multiple data records, dramatically reducing processing time while maintaining accuracy.

Inventive Principle:

Principle #28Mechanics substitution (Replace mechanical system)

Solution Approach 2:

The patent changes the data representation parameter from traditional structured formats to image data format. This transformation enables the data to be processed by GPU architectures which are optimized for image processing operations, thereby improving processing speed and efficiency.

Inventive Principle:

Principle #35Parameter changes

2Productivity

If CPU-based systems process multi-dimensional data, then data analysis can be completed, but energy consumption becomes excessive

Engineering Contradiction:

Improvedata analysis capabilityVSAvoidenergy consumption

Core Design Contradiction:

ProductivityVSUse of energy by stationary object

Solution Approach 1:

The patent substitutes CPU processing with GPU processing for data analysis tasks. GPUs are more energy-efficient for parallel workloads typical in data analysis, reducing overall energy consumption while maintaining or improving analytical capability.

Inventive Principle:

Principle #28Mechanics substitution (Replace mechanical system)

Solution Approach 2:

The patent transforms structured data into image data, adding a visual dimension to the data representation. This allows exploitation of GPU's specialized image processing capabilities, which are more energy-efficient for handling multi-dimensional data patterns.

Inventive Principle:

Principle #17Another dimension (Dimensionality change)

3Ease of operation

If traditional structured data processing is used, then data queries can be executed, but computational complexity increases

Engineering Contradiction:

Improvequery executionVSAvoidcomputational complexity

Core Design Contradiction:

Ease of operationVSDevice complexity

Solution Approach 1:

The patent replaces complex CPU-based query processing with simpler GPU-based image processing operations. By transforming data into image format, complex relational queries become simpler pixel-level operations that GPUs can execute efficiently with reduced computational complexity.

Inventive Principle:

Principle #28Mechanics substitution (Replace mechanical system)

Data Source

PatentUS20250315473A1System and method for GPU processing

Publication Date: 2025.10.09 ROYAL BANK OF CANADA

AI summary

There is provided a system and method for sprite-based Graphics Processing Unit (GPU) processing of structured data records. One or more structured databases may include one or more data records, each of said data records having one or more attributes having values. In response to receiving a query for the databases, for each data record, an image map comprising one or more pixels may be generated. The colour of each respective pixel may correspond to a value of each respective attribute for that data record. A plurality of image maps may be processed by a GPU to determine a response to the query. The result of the query may be displayed on a dashboard.