Parallel Data Processing Resequencing to Reduce Power Spikes

Resolve Bottlenecks,
Find Innovative SolutionsGenerate Solutions

Solution Overview

Problem

Parallel processing in data systems often leads to disruptive power effects such as current spikes and voltage droops, which degrade system performance and reduce the benefits of parallel processing, especially in mobile devices where power considerations are critical.

Innovation Solution

The method involves generating data sub-blocks, determining data portions based on density ranges, and resequencing these portions to avoid concurrent peak processing demands across multiple processing units, thereby reducing power spikes and thermal loading through dynamic resequencing techniques like density binning and process time resequencing.

Engineering Contradictions & Design Principles

VSEngineering Contradiction Analysis

1Productivity

If parallel processing is implemented to improve performance, then processing speed is improved, but disruptive power effects such as current spikes and voltage droops occur

Engineering Contradiction:

Improveprocessing speedVSAvoidsystem stability

Core Design Contradiction:

ProductivityVSReliability

Solution Approach 1:

The system performs preliminary analysis of data density in each processing lane before execution and pre-resequences the data to distribute peak densities across different time periods. This preliminary resequencing prevents simultaneous peak power consumption across all processing units, thereby avoiding current spikes and voltage droops while maintaining parallel processing performance.

Inventive Principle:

Principle #10Preliminary action

2Productivity

If multiple processing units initiate intensive data processing simultaneously, then processing throughput is improved, but power consumption spikes and system degradation occurs

Engineering Contradiction:

Improveprocessing throughputVSAvoidpower consumption

Core Design Contradiction:

ProductivityVSUse of energy by moving object

Solution Approach 1:

The system analyzes data density distributions across all processing lanes and resequences data elements before they are processed. By distributing high-density data portions across different time periods rather than processing them simultaneously, the system maintains high throughput while avoiding peak power consumption spikes.

Inventive Principle:

Principle #10Preliminary action

Solution Approach 2:

The system dynamically adjusts the processing schedule by resequencing data based on density analysis. This creates a dynamic power consumption profile where peak demands are spread out over time, transforming the static simultaneous processing model into a dynamic time-distributed processing model that reduces overall power spikes.

Inventive Principle:

Principle #15Dynamics

Data Source

PatentUS11983567B2Processing data stream modification to reduce power effects during parallel processing

Publication Date: 2024.05.14 QUALCOMM INC

AI summary

Certain aspects of the present disclosure provide a method for performing parallel data processing, including: receiving data for parallel processing from a data processing requestor; generating a plurality of data sub-blocks; determining a plurality of data portions in each data sub-block of the plurality of data sub-blocks; changing an order of the plurality of data portions in at least one data sub-block of the plurality of data sub-blocks; providing the plurality of data sub-blocks, including the at least one data sub-block comprising the changed order of the plurality of data portions, to a plurality of processing units for parallel processing; and receiving processed data associated with the plurality of data sub-blocks from the plurality of processing units.