Data-Based Power Mode Control for Multi-Core Processors
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Multi-core systems in mobile devices and embedded systems face challenges in power management, as they often consume more power than single processors, despite using multiple processors to reduce resource wastage, and existing power management techniques are not fine-grained enough to adapt to different data processing needs.
Innovation Solution
A processing apparatus that includes a power information manager to determine and provide predefined power modes for processors based on specific data, using a power information table generated from execution codes, allowing for selective power mode control to optimize power usage during data processing.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Productivity
If a single processor with excellent performance is used, then processing speed and performance are improved, but power consumption increases significantly
Solution Approach 1:
The patent segments the processor into multiple cores with different performance levels (high-performance cores and low-performance cores). This allows the system to divide processing tasks across multiple segments, using only the necessary performance level for each task, thereby reducing overall power consumption while maintaining productivity.
Solution Approach 2:
The patent applies local quality by assigning different power modes to different processor cores based on their specific functions and workload requirements. High-performance cores operate at higher power levels when needed, while low-performance cores handle routine tasks at lower power levels, optimizing the power-quality balance locally across the processor.
2Use of energy by moving object
If multiple processors are used in a multi-core system, then resource wastage is reduced and power consumption is lowered, but existing power management techniques are not fine-grained enough to adapt to different data processing needs
Solution Approach 1:
The patent implements dynamic power management by introducing a power information manager that continuously monitors data characteristics and dynamically adjusts processor power modes accordingly. The system can switch between different power modes (high-performance, low-performance, idle) based on real-time data analysis, making the power management adaptive and versatile for different processing needs.
Solution Approach 2:
The patent employs feedback mechanisms where the power information manager analyzes data characteristics and feeds this information back to the power controller, which then adjusts the processor's power mode. This closed-loop feedback system enables fine-grained adaptation to different data processing requirements while optimizing power consumption.
3Device complexity
If power management is implemented at a coarse level, then system complexity is reduced, but power optimization cannot be fine-grained enough for specific data processing tasks
Solution Approach 1:
The patent introduces a power information manager as an intermediary component that sits between the data processing unit and the power control system. This intermediary analyzes data characteristics and translates them into appropriate power mode selections, enabling fine-grained power optimization without requiring complex direct control between each data task and power management, thus balancing optimization precision with manageable complexity.
Data Source
AI summary
A processing apparatus for managing power based on data is provided. The processing apparatus may obtain, in response to an access request from a processor for particular data stored in a memory, existing power information having a predefined correspondence to the particular data, and control a power mode of the processor based on the existing power information.