Power Budget Allocation to Prevent Battery Overtaxing Shutdowns
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing power management systems in integrated circuits struggle to effectively manage power consumption, particularly with subsystems that require high peak power, sudden power changes, or steady-state power exceeding battery capabilities, leading to potential battery overtaxing and undesirable user experiences.
Innovation Solution
A combined closed-loop and open-loop power management system that integrates long-term power estimation and closed-loop feedback to preemptively adjust power budgets, preventing battery overtaxing while ensuring subsystems operate within sustainable power levels.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Power
If power management systems allow subsystems to draw high peak power or sudden power changes, then subsystem performance is improved, but battery capability is exceeded and system reliability deteriorates
Solution Approach 1:
The power management system performs preliminary estimation of upcoming power demands and battery capability before the actual power draw occurs. By predicting future power requirements and comparing them against estimated battery capability, the system proactively adjusts power budgets to prevent battery overtaxing, thereby maintaining system reliability while allowing subsystems to operate at high peak power when feasible.
Solution Approach 2:
The power management system dynamically adjusts power budgets for subsystems based on real-time estimates of battery capability and predicted power demands. Rather than using fixed power limits, the system continuously adapts power allocation to match changing battery conditions and subsystem requirements, enabling high peak power delivery when battery capability permits while preventing reliability issues when the battery is stressed.
2Use of energy by stationary object
If power management systems use closed-loop feedback control, then power consumption is regulated, but response time to sudden power changes increases
Solution Approach 1:
The system performs preliminary estimation of power demands and battery capability in advance, before sudden power changes occur. This open-loop prediction component prepares power budget adjustments proactively, so when sudden power changes are detected, the system can respond more quickly without waiting for closed-loop feedback to complete full cycles, thereby reducing response time while maintaining power consumption regulation.
Solution Approach 2:
The power management system merges open-loop predictive estimation with closed-loop feedback control into a hybrid approach. The open-loop component provides preliminary predictions that prepare the system for upcoming power demands, while the closed-loop component continuously monitors actual power consumption and makes real-time adjustments. This combination allows the system to respond faster to sudden power changes than closed-loop alone while maintaining effective power regulation.
3Reliability
If power management systems reduce power budgets preemptively, then battery overtaxing is prevented, but subsystem performance and user experience deteriorate
Solution Approach 1:
The system dynamically adjusts power budgets based on actual battery capability estimates and predicted power demands, rather than applying uniform preemptive reductions. When battery capability is sufficient to support high performance, the system allows subsystems to operate at full performance levels. When battery capability is limited, the system selectively reduces power budgets only for non-critical subsystems or during periods of low activity, thereby preventing battery overtaxing while minimizing impact on user experience.
Solution Approach 2:
The power management system applies different power budget adjustment strategies to different subsystems based on their criticality and power characteristics. Critical subsystems that require high performance maintain their power budgets even when battery capability is limited, while non-critical subsystems receive preemptive power reductions. This localized approach prevents battery overtaxing while preserving essential subsystem performance and maintaining acceptable user experience.
Data Source
AI summary
Systems and methods are disclosed for allocating and distributing power management budgets for subsystems of a computer system. For each of a plurality of subsystems, a power management system may generate a long-term power budget and a closed-loop power budget, and then determine a final power budget to provide to the subsystem, e.g., by applying a min function, a weighted sum, or some other function to the long-term power budget and the closed-loop power budget. The closed-loop power budget is determined based on observations of power draw over a past period of time, and therefore cannot respond immediately to large changes in power. The long-term power budget is generated based on a prediction of battery capability over an upcoming window of time, and may therefore provide a power cap before the system is under duress.