Frame Generation Power Scheduling for Inference and Non-Inference Units
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Mobile electronic devices experience non-optimal power consumption due to third-party applications instructing graphics-related devices to operate with inappropriate parameters, leading to inefficient power usage.
Innovation Solution
A power consumption management method and device that optimize power consumption by determining operation times and configurations for non-inference and inference units within a frame generation procedure, considering historical data and target performance time, to ensure efficient power usage while meeting performance requirements.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Ease of operation
If third-party applications instruct graphics-related devices to operate with inappropriate parameters, then the application can run with simple control logic, but power consumption becomes non-optimal
Solution Approach 1:
The system introduces an intermediary power consumption management mechanism that sits between the third-party application and the graphics-related devices. This intermediary layer receives instructions from the application, determines appropriate operation parameters based on device capabilities and historical data, and translates them into suitable commands for the graphics devices, thereby optimizing power consumption without complicating the application's control logic
Solution Approach 2:
The system dynamically changes operation parameters of graphics-related devices based on historical data and current performance requirements. By adjusting parameters such as operation time, resolution, and frame rate according to the specific needs of each application and the capabilities of the device, the system achieves optimal power consumption while maintaining appropriate performance levels
2Productivity
If non-inference units operate at high power consumption rates to complete frame generation tasks quickly, then frame generation speed increases, but overall power consumption increases
Solution Approach 1:
The system segments the frame generation process into two distinct parts: non-inference stages handled by non-inference units and inference stages handled by inference units. By dividing the workload and assigning tasks to appropriate units based on their power consumption characteristics and capabilities, the system optimizes the balance between frame generation speed and power consumption
Solution Approach 2:
The system dynamically determines the operation time of non-inference units based on historical data and current performance requirements, rather than using fixed high-power settings. This dynamic adjustment allows the system to use high power consumption rates only when necessary to meet performance targets, while using lower power consumption rates when performance targets can be met with less power
3Use of energy by moving object
If the system uses historical data and dynamic parameter adjustment to optimize power consumption, then power efficiency improves, but system complexity increases
Solution Approach 1:
The system performs preliminary actions by collecting and storing historical data about device performance and power consumption characteristics before actual frame generation tasks. This pre-processing of data allows the system to make informed decisions about parameter adjustment without adding complexity to the real-time frame generation process, as the decision-making framework is already in place
Data Source
AI summary
A power consumption management method, for managing power consumptions of units for performing a frame generation procedure to generate frames required by an application, comprising: (a) deciding a first operation time of at least one non-inference unit for performing non-inference stages of the frame generation procedure according to a target performance time determined by the application and historical data relevant to the non-inference stages; (b) computing a second operation time of an inference unit for performing an inference stage of the frame generation procedure, according to the target performance time and the first operation time; and (c) setting configurations of the inference unit according to the second operation time; wherein the at least one non-inference unit has a first power consumption rate higher than a second power consumption rate of the inference unit.