Frequency Scaling for Mobile Processors

Resolve Bottlenecks,
Find Innovative Solutions
Generate Solutions

Solution Overview

Problem

Existing frequency scaling methods in mobile devices, such as smartphones, fail to timely respond to performance changes, leading to inefficient power consumption and battery life reduction.

Innovation Solution

A frequency scaling method that predicts energy efficiency parameters for modules like CPU, GPU, and memory, selecting optimal frequency sets based on these parameters to ensure timely response to load changes and reduce power consumption.

Engineering Contradictions & Design Principles

VSEngineering Contradiction Analysis

1Use of energy by moving object

If frequency scaling is performed level by level based on predicted CPU usage by using a window size as a period, then power consumption can be reduced, but the system cannot timely respond to a change of the performance requirement

Engineering Contradiction:
Improvepower consumptionVSAvoidresponse speed to performance change
Core Design Contradiction:
Use of energy by moving objectVSSpeed

Solution Approach 1:

The system performs preliminary frequency adjustment within the current window based on detected performance changes, rather than waiting for the window to complete. This allows the frequency to be adjusted in advance (preliminarily) when a performance change is detected, reducing the response delay while still maintaining the window-based power management approach

Inventive Principle:
Principle #10Preliminary action

Solution Approach 2:

The system dynamically adjusts the frequency scaling behavior based on whether a performance change is detected. When no change is detected, the system follows the original window-based level-by-level scaling. When a change is detected, the system dynamically switches to an accelerated adjustment mode, making the frequency scaling process adaptive and dynamic rather than static

Inventive Principle:
Principle #15Dynamics

2Use of energy by moving object

If the frequency of the CPU is scaled based on a load using traditional methods, then power consumption can be reduced, but the frequency cannot quickly respond to a load requirement

Engineering Contradiction:
Improvepower consumptionVSAvoidresponse time to load requirement
Core Design Contradiction:
Use of energy by moving objectVSLoss of time

Solution Approach 1:

The system continuously monitors performance indicators (such as frame rate, CPU usage, or other load metrics) and uses this feedback to detect when a performance change occurs. This feedback mechanism triggers the preliminary frequency adjustment, creating a closed-loop control system that responds to actual load conditions rather than relying solely on predetermined window-based timing

Inventive Principle:
Principle #23Feedback

Solution Approach 2:

Upon detecting a performance change through feedback, the system performs preliminary frequency adjustment within the current window, anticipating the need for frequency change before the window completes. This preliminary action reduces the time delay between load requirement and frequency response while maintaining power efficiency

Inventive Principle:
Principle #10Preliminary action

Data Source

PatentEP3819744B1Frequency adjustment method and apparatus
Publication Date: 2025.06.25 HUAWEI TECH CO LTD
  • EP3819744B1 patent drawingFigure 1~2
  • EP3819744B1 patent drawingFigure 3~4
  • EP3819744B1 patent drawingFigure 5

AI summary

This application provides a frequency scaling method. The method includes: predicting an energy efficiency parameter for processing a current frame of an image by at least one module; and selecting, from a plurality of frequency sets based on the predicted energy efficiency parameter, a first frequency set that meets an energy efficiency requirement, and scaling a working frequency of each of the at least one module for processing the current frame to a preset frequency corresponding to each of the at least one module. According to the technical solutions provided in the embodiments of this application, a load change requirement can be responded to in time in a frequency scaling process.