Dynamic Voltage Frequency Scaling Mobile Processor
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Modern mobile devices face a design trade-off between application performance and battery life due to increasing power consumption, as processor performance and transistor density increase, while battery technology has not kept pace, and conventional dynamic voltage and frequency scaling (DVFS) algorithms are not optimal for all interactive and unpredictable workloads.
Innovation Solution
A method for managing and exposing a set of performance scaling algorithms on a mobile device, which involves providing and associating these algorithms with parameters in non-volatile memory, identifying suitable programs for processor cores, and creating an interface to expose them, allowing for dynamic adjustment of performance and power consumption based on specific conditions and events.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Productivity
If processor clock frequency and transistor density are increased to improve application performance, then processing speed and computational capability are improved, but power consumption increases rapidly
Solution Approach 1:
The system dynamically adjusts processor operating parameters (voltage and frequency) based on actual workload demands and battery status. The DVFS controller continuously monitors system state and adjusts processor performance levels in real-time, transforming the static high-performance configuration into a dynamic system that adapts between high-performance and power-saving modes.
Solution Approach 2:
The invention changes the operating parameters of the processor (voltage and frequency) to optimize the balance between performance and power consumption. By adjusting these parameters based on workload characteristics and battery state, the system achieves different operating points that resolve the contradiction between high performance and low power consumption.
2Use of energy by moving object
If conventional DVFS algorithms use idle time measurement to reduce performance level, then power consumption is reduced, but the approach is not optimal for interactive and unpredictable workloads
Solution Approach 1:
The system implements feedback mechanisms where the DVFS controller continuously monitors multiple parameters including idle time, application behavior patterns, and battery status. This feedback loop enables the system to learn from actual workload characteristics and adjust DVFS strategies accordingly, making the power management adaptive to different workload types rather than relying on a single idle-time-based algorithm.
Solution Approach 2:
The invention creates a universal DVFS framework that can handle multiple types of workloads (interactive, batch, real-time, unpredictable) through a single adaptable system. The controller evaluates various workload characteristics and selects appropriate scaling strategies, making the power management system versatile across different application scenarios rather than specialized for one workload type.
3Speed
If processor performance is maintained at high levels, then application responsiveness is improved, but battery life is reduced
Solution Approach 1:
The system employs periodic monitoring and adjustment of processor performance levels based on battery status and workload patterns. Rather than maintaining constant high performance, the system periodically evaluates system state and adjusts performance levels appropriately, achieving a rhythm of high-performance bursts followed by lower-power operation that extends battery life while maintaining perceived responsiveness.
Data Source
AI summary
A mobile device, a method for managing and exposing a set of performance scaling algorithms on the device, and a computer program product are disclosed. The mobile device includes a multiple-core processor communicatively coupled to a non-volatile memory. The non-volatile memory includes a set of programs defined by a respective combination of a performance scaling algorithm and a set of parameters, a startup program that when executed by the multiple-core processor identifies at least one member of the set of programs suitable for monitoring operation of the mobile device and scaling the performance of an identified processor core and an application programming interface that exposes the set of programs.