Application Processor Dynamic Power Management via Data Change Signals
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Conventional schedulers in systems with heterogeneous multi-processors do not effectively manage power consumption based on changes in user data, leading to inefficient performance and increased energy usage.
Innovation Solution
An application processor that includes a check manager to generate control signals indicating changes in user data, allowing for dynamic power management by adjusting the reference values for task transitions between high-performance and low-performance processors, and controlling clock signals and voltage/frequency scaling.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Productivity
If a conventional scheduler schedules tasks according to the amount of given jobs between heterogeneous multi-processors, then the system can show satisfactory performance when there are a large number of given jobs, but power consumption increases when there are many jobs requiring high-performance processors
Solution Approach 1:
The scheduler dynamically adjusts the reference value for task migration between high-performance and low-performance processors based on real-time conditions. When user data changes frequently, the reference value is maintained or decreased to favor low-power processors. When user data remains stable, the reference value is increased to allow migration to high-performance processors, optimizing the balance between performance and power consumption adaptively
Solution Approach 2:
The system changes the reference value parameter that controls task migration decisions between processors. This parameter is adjusted based on user data change detection: decreased when data changes frequently to maintain tasks on low-power processors, and increased when data is stable to enable migration to high-performance processors for improved system performance
2Use of energy by moving object
If tasks are migrated frequently between high-performance and low-performance processors to optimize power consumption, then energy usage decreases, but system performance may deteriorate due to excessive task transitions
Solution Approach 1:
The scheduler implements a feedback mechanism that monitors user data changes and adjusts the reference value accordingly. This feedback loop prevents unnecessary task migrations by only changing migration thresholds when actual user data changes are detected, thereby maintaining system performance while optimizing power consumption through informed migration decisions
3Productivity
If the reference value for task migration is increased to improve performance, then tasks can be executed on high-performance processors, but power consumption increases
Solution Approach 1:
The reference value is made dynamic rather than fixed, adjusting based on user data change patterns. The scheduler increases the reference value only when user data stability indicates that performance optimization is appropriate, and decreases it when data changes suggest power savings are more critical, creating a dynamic balance between performance and energy consumption
Data Source
AI summary
An application processor includes an application processor including a first processor configured to generate a control signal based on whether user data is changed, wherein the application processor is configured to implement a power manager which dynamically controls power provided to the first processor, in response to the control signal.