Dynamic Voltage-Frequency Scaling for Variable Real-Time Workloads

Resolve Bottlenecks,
Find Innovative Solutions
Generate Solutions

Solution Overview

Problem

Existing methods for managing processor energy consumption through Dynamic Voltage and Frequency Scaling (DVFS) are inadequate for applications with variable real-time criticality and unpredictable deadlines, as they require fixed frequency settings, detailed knowledge of processing operations, or integrated scheduler management, and fail to account for criticality requirements.

Innovation Solution

A method that introduces a calculation capacity controller to dynamically adjust voltage-frequency levels based on trigger factors, deadlines, and criticality, allowing for automatic power monitoring without prior knowledge of the scheduler, enabling efficient energy management across variable workloads.

Engineering Contradictions & Design Principles

VSEngineering Contradiction Analysis

1Device complexity

If fixed frequency settings are used for processor energy management, then device complexity is reduced, but energy consumption cannot be optimized for variable workloads

Engineering Contradiction:
Improvefrequency management complexityVSAvoidprocessor energy consumption
Core Design Contradiction:
Device complexityVSUse of energy by moving object

Solution Approach 1:

The patent implements dynamic frequency adjustment by introducing a frequency manager that continuously monitors workload characteristics and adapts processor frequency settings in real-time. This transforms the static frequency management system into a dynamic one that can respond to varying computational demands, optimizing energy consumption without requiring complex manual configuration.

Inventive Principle:
Principle #15Dynamics

Solution Approach 2:

The system changes the frequency parameter of the processor based on detected workload patterns. By monitoring execution time, instruction mix, and other performance metrics, the frequency manager adjusts the operating frequency to match actual computational needs, thereby reducing energy consumption during low-utilization periods while maintaining performance during high-demand tasks.

Inventive Principle:
Principle #35Parameter changes

2Measurement precision

If detailed knowledge of processing operations is required for frequency management, then energy optimization precision is improved, but ease of operation deteriorates

Engineering Contradiction:
Improveenergy optimization precisionVSAvoidfrequency management ease
Core Design Contradiction:
Measurement precisionVSEase of operation

Solution Approach 1:

The frequency manager operates autonomously by automatically monitoring processor workload characteristics and adjusting frequency settings without requiring external intervention or detailed knowledge of processing operations. The system self-services by collecting performance metrics, analyzing workload patterns, and making frequency adjustments based on predefined policies, thereby maintaining high optimization precision while preserving ease of operation.

Inventive Principle:
Principle #25Self-service

Solution Approach 2:

The system implements a feedback mechanism where the frequency manager continuously monitors processor performance metrics such as execution time, instruction mix, and utilization patterns. This feedback information is used to dynamically adjust frequency settings, creating a closed-loop control system that achieves precise energy optimization without requiring manual configuration or deep knowledge of processing operations.

Inventive Principle:
Principle #23Feedback

3Productivity

If integrated scheduler management is used for frequency control, then productivity is improved, but device complexity increases

Engineering Contradiction:
Improveprocessing throughputVSAvoidscheduler integration complexity
Core Design Contradiction:
ProductivityVSDevice complexity

Solution Approach 1:

The patent introduces a frequency manager as an intermediary component that bridges the gap between the scheduler and processor frequency control. This intermediary monitors workload characteristics from the scheduler and translates them into appropriate frequency adjustments, thereby maintaining high processing throughput while avoiding the complexity of direct scheduler-integrated frequency management.

Inventive Principle:
Principle #24Intermediary (Mediator)

Solution Approach 2:

The system segments frequency management into a separate, independent module (frequency manager) that operates independently from the scheduler. This segmentation allows the scheduler to focus on task allocation and scheduling decisions while the frequency manager handles frequency optimization, thereby maintaining productivity without requiring complex integrated scheduler management.

Inventive Principle:
Principle #1Segmentation

4Device complexity

If criticality requirements are not considered in frequency management, then device complexity is reduced, but reliability deteriorates

Engineering Contradiction:
Improvecriticality management complexityVSAvoiddeadline meeting guarantee
Core Design Contradiction:
Device complexityVSReliability

Solution Approach 1:

The frequency manager implements local quality by applying different frequency management strategies to tasks with different criticality levels. Critical tasks receive prioritized frequency allocation and more aggressive frequency boosting to ensure deadline compliance, while non-critical tasks use standard energy-saving policies. This differentiated approach ensures reliability for time-sensitive operations without requiring complex system-wide criticality management.

Inventive Principle:
Principle #3Local quality

Solution Approach 2:

The system changes frequency parameters dynamically based on task criticality. When a critical task is detected, the frequency manager applies parameter changes to boost frequency specifically for that task's execution context, thereby guaranteeing deadline compliance for time-sensitive operations while maintaining lower frequencies for non-critical tasks to optimize energy consumption.

Inventive Principle:
Principle #35Parameter changes

Data Source

PatentEP2881839B1Method of managing processing resources for software applications
Publication Date: 2022.03.23 THALES SA
  • EP2881839B1 patent drawingFigure 1
  • EP2881839B1 patent drawingFigure 2

AI summary

The invention relates to a method for monitoring the level of computing capacity allocated to a hardware platform for the execution of a software application by identifying situations in which it is possible to limit the energy consumption of a processor during application execution.