Power-Saving Circuitry for Dynamic AI Workload and Charging Control
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing computing devices struggle to efficiently manage power consumption during charging and AI inference operations, particularly in determining user presence and adjusting operations accordingly.
Innovation Solution
The implementation of power-saving circuitry that detects user presence using various sensors and adjusts the charging rate and AI workload distribution accordingly, either performing local AI inference when the user is present or offloading it to a remote device when the user is absent.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Speed
If the computing device performs local AI inference operations, then AI processing speed and responsiveness are improved, but power consumption increases
Solution Approach 1:
The system dynamically adjusts AI processing location between local and remote based on user presence detection. When a user is present, local AI inference is performed for fast response. When no user is present, AI workloads are offloaded to remote devices to conserve power, thus adapting the system behavior to current usage conditions.
Solution Approach 2:
A user presence detection system acts as an intermediary that determines whether to perform AI processing locally or remotely. The presence detector monitors user presence and controls the workload distributor to switch between local and remote AI inference, mediating the trade-off between speed and power consumption.
2Speed
If the computing device charges at a high rate, then charging speed is improved, but heat generation increases which may require cooling
Solution Approach 1:
The charging rate is dynamically adjusted based on detected user presence. When a user is present, the system charges at a first (lower) rate to maintain comfortable temperatures. When no user is present, the system charges at a second (higher) rate to maximize charging speed, thus adapting charging behavior to usage conditions.
Solution Approach 2:
The system performs preliminary user presence detection before initiating high-rate charging. By detecting user presence in advance, the system can prevent high-rate charging when a user is present, avoiding excessive heat generation before it occurs.
3Measurement precision
If the device monitors user presence continuously, then user presence detection accuracy is improved, but power consumption increases
Solution Approach 1:
Instead of continuous monitoring, the system uses periodic user presence detection at key decision points (before AI workload execution and before high-rate charging). This periodic checking provides sufficient accuracy for controlling power-intensive operations while significantly reducing power consumption compared to continuous monitoring.
Data Source
AI summary
Systems, apparatus, articles of manufacture, and methods are disclosed to perform power-saving based on user presence, including a network interface to communicate with a cloud device, user presence detector circuitry to determine if a user is present or not present; workload distributor circuitry to distribute an AI workload to either first AI inference circuitry or second AI inference circuitry; and power circuitry to charge a battery at either a first charge level or a second charge level.