Dynamic Task Migration for Multi-Processor Power Optimization
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
In multi-processor systems, existing methods control task operations using a fixed value regardless of task situations, types, and attributes, leading to unnecessary power consumption and performance degradation.
Innovation Solution
A task migration method and apparatus that dynamically configure critical conditions based on task types, attributes, and situations, allowing for optimized scheduling by migrating tasks between different operating devices.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Device complexity
If a fixed critical value is used to control task operations in a multi-processor system, then the control mechanism is simple and easy to implement, but unnecessary power consumption occurs and system performance is degraded
Solution Approach 1:
The patent applies dynamics by transitioning from a fixed critical value to a dynamic critical value that changes based on processor states. The critical value is adjusted according to whether the processor is in active or idle mode, allowing the control mechanism to adapt to different operational conditions and avoid unnecessary power consumption while maintaining simplicity.
Solution Approach 2:
The patent changes the parameter of critical value from a static fixed number to a dynamic parameter that varies based on processor state. By modifying the critical value parameter according to active/idle states and task characteristics, the system achieves better power efficiency and performance without significantly increasing complexity.
2Device complexity
If a fixed critical value is used to control task operations, then the control logic is simple, but system performance is degraded due to inability to adapt to different task situations
Solution Approach 1:
The control logic becomes dynamic by adjusting the critical value based on processor state (active or idle) and task characteristics. This allows the system to optimize performance for different situations while maintaining relatively simple control structures, resolving the contradiction between simplicity and adaptability.
Solution Approach 2:
The patent applies local quality by using different critical values for different processor states and different task types. Instead of a single global critical value, the system tailors the critical value to local conditions (processor state and task characteristics), improving overall system performance without requiring complex global optimization.
3Adaptability or versatility
If tasks are migrated frequently to optimize performance, then system adaptability improves, but overhead increases and may degrade overall efficiency
Solution Approach 1:
The patent implements feedback by continuously monitoring processor state and task characteristics, then using this information to dynamically adjust the critical value and determine whether migration is necessary. This feedback mechanism ensures tasks are migrated only when beneficial, balancing adaptability with efficiency by avoiding unnecessary migrations.
Solution Approach 2:
The system applies partial action by migrating tasks selectively rather than continuously. By using a dynamically adjusted critical value, the system performs migration only when the condition is met, avoiding excessive migration operations and their associated overhead while still achieving necessary adaptability.
Data Source
AI summary
A method for recognizing an image and an apparatus using the same are provided. The method includes receiving image information including a first object and a second object, recognizing position information of the first object indicated by the second information in the received image information, extracting effective information included in the first object of the received image information in response to the recognized position information, and outputting related information corresponding to the recognized effective information.